{
  "best_global_step": 25200,
  "best_metric": 0.2112353891134262,
  "best_model_checkpoint": "saves/prompt-tuning/gemma-3-1b-it/train_boolq_1745950272/checkpoint-25200",
  "epoch": 18.85902876001886,
  "eval_steps": 200,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0023573785950023575,
      "grad_norm": 31.988269805908203,
      "learning_rate": 0.29999999259779675,
      "loss": 10.4129,
      "num_input_tokens_seen": 4304,
      "step": 5
    },
    {
      "epoch": 0.004714757190004715,
      "grad_norm": 20.223493576049805,
      "learning_rate": 0.29999996252634736,
      "loss": 12.1618,
      "num_input_tokens_seen": 8432,
      "step": 10
    },
    {
      "epoch": 0.007072135785007072,
      "grad_norm": 4.1196746826171875,
      "learning_rate": 0.2999999093230187,
      "loss": 9.0938,
      "num_input_tokens_seen": 12688,
      "step": 15
    },
    {
      "epoch": 0.00942951438000943,
      "grad_norm": 3.138713836669922,
      "learning_rate": 0.299999832987819,
      "loss": 4.8869,
      "num_input_tokens_seen": 17360,
      "step": 20
    },
    {
      "epoch": 0.011786892975011787,
      "grad_norm": 1.5348763465881348,
      "learning_rate": 0.29999973352076004,
      "loss": 2.1921,
      "num_input_tokens_seen": 21808,
      "step": 25
    },
    {
      "epoch": 0.014144271570014143,
      "grad_norm": 1.0330103635787964,
      "learning_rate": 0.2999996109218572,
      "loss": 1.3614,
      "num_input_tokens_seen": 25952,
      "step": 30
    },
    {
      "epoch": 0.0165016501650165,
      "grad_norm": 0.6571065783500671,
      "learning_rate": 0.2999994651911293,
      "loss": 1.008,
      "num_input_tokens_seen": 29568,
      "step": 35
    },
    {
      "epoch": 0.01885902876001886,
      "grad_norm": 1.2071079015731812,
      "learning_rate": 0.2999992963285989,
      "loss": 0.599,
      "num_input_tokens_seen": 33712,
      "step": 40
    },
    {
      "epoch": 0.021216407355021217,
      "grad_norm": 0.5068820118904114,
      "learning_rate": 0.29999910433429194,
      "loss": 0.8035,
      "num_input_tokens_seen": 37856,
      "step": 45
    },
    {
      "epoch": 0.023573785950023574,
      "grad_norm": 0.16513298451900482,
      "learning_rate": 0.29999888920823814,
      "loss": 0.3537,
      "num_input_tokens_seen": 41936,
      "step": 50
    },
    {
      "epoch": 0.02593116454502593,
      "grad_norm": 0.5797112584114075,
      "learning_rate": 0.29999865095047057,
      "loss": 0.4111,
      "num_input_tokens_seen": 46448,
      "step": 55
    },
    {
      "epoch": 0.028288543140028287,
      "grad_norm": 0.3267705738544464,
      "learning_rate": 0.29999838956102604,
      "loss": 0.4687,
      "num_input_tokens_seen": 51120,
      "step": 60
    },
    {
      "epoch": 0.030645921735030647,
      "grad_norm": 0.09841182082891464,
      "learning_rate": 0.29999810503994484,
      "loss": 0.2947,
      "num_input_tokens_seen": 55520,
      "step": 65
    },
    {
      "epoch": 0.033003300330033,
      "grad_norm": 0.10266968607902527,
      "learning_rate": 0.29999779738727084,
      "loss": 0.3085,
      "num_input_tokens_seen": 60496,
      "step": 70
    },
    {
      "epoch": 0.03536067892503536,
      "grad_norm": 0.20409666001796722,
      "learning_rate": 0.29999746660305154,
      "loss": 0.2472,
      "num_input_tokens_seen": 64368,
      "step": 75
    },
    {
      "epoch": 0.03771805752003772,
      "grad_norm": 0.17530688643455505,
      "learning_rate": 0.2999971126873379,
      "loss": 0.5316,
      "num_input_tokens_seen": 68720,
      "step": 80
    },
    {
      "epoch": 0.040075436115040074,
      "grad_norm": 0.1986703872680664,
      "learning_rate": 0.2999967356401845,
      "loss": 0.3228,
      "num_input_tokens_seen": 72592,
      "step": 85
    },
    {
      "epoch": 0.042432814710042434,
      "grad_norm": 0.10162245482206345,
      "learning_rate": 0.29999633546164944,
      "loss": 0.3781,
      "num_input_tokens_seen": 76384,
      "step": 90
    },
    {
      "epoch": 0.04479019330504479,
      "grad_norm": 0.2918356657028198,
      "learning_rate": 0.29999591215179444,
      "loss": 0.3153,
      "num_input_tokens_seen": 80224,
      "step": 95
    },
    {
      "epoch": 0.04714757190004715,
      "grad_norm": 0.11471439898014069,
      "learning_rate": 0.2999954657106849,
      "loss": 0.2848,
      "num_input_tokens_seen": 84704,
      "step": 100
    },
    {
      "epoch": 0.04950495049504951,
      "grad_norm": 0.16378742456436157,
      "learning_rate": 0.2999949961383896,
      "loss": 0.2694,
      "num_input_tokens_seen": 88688,
      "step": 105
    },
    {
      "epoch": 0.05186232909005186,
      "grad_norm": 0.31704339385032654,
      "learning_rate": 0.2999945034349809,
      "loss": 0.6022,
      "num_input_tokens_seen": 93184,
      "step": 110
    },
    {
      "epoch": 0.05421970768505422,
      "grad_norm": 0.08817286789417267,
      "learning_rate": 0.2999939876005348,
      "loss": 0.3188,
      "num_input_tokens_seen": 98192,
      "step": 115
    },
    {
      "epoch": 0.056577086280056574,
      "grad_norm": 0.43094635009765625,
      "learning_rate": 0.29999344863513094,
      "loss": 0.4523,
      "num_input_tokens_seen": 102912,
      "step": 120
    },
    {
      "epoch": 0.058934464875058934,
      "grad_norm": 0.22856155037879944,
      "learning_rate": 0.2999928865388523,
      "loss": 0.4735,
      "num_input_tokens_seen": 108384,
      "step": 125
    },
    {
      "epoch": 0.061291843470061294,
      "grad_norm": 0.20460647344589233,
      "learning_rate": 0.29999230131178567,
      "loss": 0.3447,
      "num_input_tokens_seen": 113248,
      "step": 130
    },
    {
      "epoch": 0.06364922206506365,
      "grad_norm": 0.13002483546733856,
      "learning_rate": 0.2999916929540212,
      "loss": 0.4521,
      "num_input_tokens_seen": 117872,
      "step": 135
    },
    {
      "epoch": 0.066006600660066,
      "grad_norm": 0.1392366588115692,
      "learning_rate": 0.29999106146565285,
      "loss": 0.3571,
      "num_input_tokens_seen": 122176,
      "step": 140
    },
    {
      "epoch": 0.06836397925506836,
      "grad_norm": 0.13612762093544006,
      "learning_rate": 0.29999040684677786,
      "loss": 0.2907,
      "num_input_tokens_seen": 126432,
      "step": 145
    },
    {
      "epoch": 0.07072135785007072,
      "grad_norm": 0.08337651938199997,
      "learning_rate": 0.2999897290974972,
      "loss": 0.2691,
      "num_input_tokens_seen": 130576,
      "step": 150
    },
    {
      "epoch": 0.07307873644507308,
      "grad_norm": 0.046880874782800674,
      "learning_rate": 0.2999890282179155,
      "loss": 0.2996,
      "num_input_tokens_seen": 135184,
      "step": 155
    },
    {
      "epoch": 0.07543611504007544,
      "grad_norm": 0.1292787343263626,
      "learning_rate": 0.29998830420814077,
      "loss": 0.2249,
      "num_input_tokens_seen": 139248,
      "step": 160
    },
    {
      "epoch": 0.07779349363507779,
      "grad_norm": 0.07504863291978836,
      "learning_rate": 0.2999875570682846,
      "loss": 0.328,
      "num_input_tokens_seen": 143984,
      "step": 165
    },
    {
      "epoch": 0.08015087223008015,
      "grad_norm": 0.10210473835468292,
      "learning_rate": 0.2999867867984623,
      "loss": 0.3487,
      "num_input_tokens_seen": 148704,
      "step": 170
    },
    {
      "epoch": 0.08250825082508251,
      "grad_norm": 0.10031376034021378,
      "learning_rate": 0.29998599339879267,
      "loss": 0.2901,
      "num_input_tokens_seen": 153440,
      "step": 175
    },
    {
      "epoch": 0.08486562942008487,
      "grad_norm": 0.10620614141225815,
      "learning_rate": 0.29998517686939796,
      "loss": 0.3292,
      "num_input_tokens_seen": 157264,
      "step": 180
    },
    {
      "epoch": 0.08722300801508723,
      "grad_norm": 0.06989361345767975,
      "learning_rate": 0.29998433721040413,
      "loss": 0.3229,
      "num_input_tokens_seen": 161840,
      "step": 185
    },
    {
      "epoch": 0.08958038661008957,
      "grad_norm": 0.037209998816251755,
      "learning_rate": 0.29998347442194073,
      "loss": 0.2519,
      "num_input_tokens_seen": 165616,
      "step": 190
    },
    {
      "epoch": 0.09193776520509193,
      "grad_norm": 0.03803081810474396,
      "learning_rate": 0.2999825885041407,
      "loss": 0.2571,
      "num_input_tokens_seen": 169760,
      "step": 195
    },
    {
      "epoch": 0.0942951438000943,
      "grad_norm": 0.08795905113220215,
      "learning_rate": 0.29998167945714077,
      "loss": 0.2964,
      "num_input_tokens_seen": 174096,
      "step": 200
    },
    {
      "epoch": 0.0942951438000943,
      "eval_loss": 0.3283520042896271,
      "eval_runtime": 21.8999,
      "eval_samples_per_second": 43.06,
      "eval_steps_per_second": 21.553,
      "num_input_tokens_seen": 174096,
      "step": 200
    },
    {
      "epoch": 0.09665252239509665,
      "grad_norm": 0.03700771927833557,
      "learning_rate": 0.2999807472810811,
      "loss": 0.3905,
      "num_input_tokens_seen": 178704,
      "step": 205
    },
    {
      "epoch": 0.09900990099009901,
      "grad_norm": 0.032493144273757935,
      "learning_rate": 0.29997979197610536,
      "loss": 0.26,
      "num_input_tokens_seen": 182912,
      "step": 210
    },
    {
      "epoch": 0.10136727958510136,
      "grad_norm": 0.05959407612681389,
      "learning_rate": 0.299978813542361,
      "loss": 0.2928,
      "num_input_tokens_seen": 186880,
      "step": 215
    },
    {
      "epoch": 0.10372465818010372,
      "grad_norm": 0.08966787159442902,
      "learning_rate": 0.2999778119799988,
      "loss": 0.2034,
      "num_input_tokens_seen": 190976,
      "step": 220
    },
    {
      "epoch": 0.10608203677510608,
      "grad_norm": 0.06561319530010223,
      "learning_rate": 0.29997678728917326,
      "loss": 0.2247,
      "num_input_tokens_seen": 195136,
      "step": 225
    },
    {
      "epoch": 0.10843941537010844,
      "grad_norm": 0.06829624623060226,
      "learning_rate": 0.2999757394700424,
      "loss": 0.2597,
      "num_input_tokens_seen": 199376,
      "step": 230
    },
    {
      "epoch": 0.1107967939651108,
      "grad_norm": 0.33134883642196655,
      "learning_rate": 0.29997466852276783,
      "loss": 0.4023,
      "num_input_tokens_seen": 203104,
      "step": 235
    },
    {
      "epoch": 0.11315417256011315,
      "grad_norm": 0.1378246247768402,
      "learning_rate": 0.29997357444751466,
      "loss": 0.3438,
      "num_input_tokens_seen": 207440,
      "step": 240
    },
    {
      "epoch": 0.11551155115511551,
      "grad_norm": 0.1637970358133316,
      "learning_rate": 0.2999724572444516,
      "loss": 0.2485,
      "num_input_tokens_seen": 211520,
      "step": 245
    },
    {
      "epoch": 0.11786892975011787,
      "grad_norm": 0.33689868450164795,
      "learning_rate": 0.29997131691375095,
      "loss": 0.4969,
      "num_input_tokens_seen": 216224,
      "step": 250
    },
    {
      "epoch": 0.12022630834512023,
      "grad_norm": 0.17855975031852722,
      "learning_rate": 0.2999701534555886,
      "loss": 0.4601,
      "num_input_tokens_seen": 220352,
      "step": 255
    },
    {
      "epoch": 0.12258368694012259,
      "grad_norm": 0.12806305289268494,
      "learning_rate": 0.2999689668701439,
      "loss": 0.3606,
      "num_input_tokens_seen": 224352,
      "step": 260
    },
    {
      "epoch": 0.12494106553512493,
      "grad_norm": 0.04102272540330887,
      "learning_rate": 0.29996775715759993,
      "loss": 0.2911,
      "num_input_tokens_seen": 228240,
      "step": 265
    },
    {
      "epoch": 0.1272984441301273,
      "grad_norm": 0.029556559398770332,
      "learning_rate": 0.2999665243181432,
      "loss": 0.2242,
      "num_input_tokens_seen": 232304,
      "step": 270
    },
    {
      "epoch": 0.12965582272512965,
      "grad_norm": 0.04120897129178047,
      "learning_rate": 0.2999652683519638,
      "loss": 0.1807,
      "num_input_tokens_seen": 236464,
      "step": 275
    },
    {
      "epoch": 0.132013201320132,
      "grad_norm": 0.07952621579170227,
      "learning_rate": 0.29996398925925544,
      "loss": 0.3414,
      "num_input_tokens_seen": 241440,
      "step": 280
    },
    {
      "epoch": 0.13437057991513437,
      "grad_norm": 0.10197393596172333,
      "learning_rate": 0.2999626870402154,
      "loss": 0.3377,
      "num_input_tokens_seen": 245104,
      "step": 285
    },
    {
      "epoch": 0.13672795851013672,
      "grad_norm": 0.07283695042133331,
      "learning_rate": 0.29996136169504445,
      "loss": 0.3599,
      "num_input_tokens_seen": 249552,
      "step": 290
    },
    {
      "epoch": 0.1390853371051391,
      "grad_norm": 0.036499980837106705,
      "learning_rate": 0.29996001322394694,
      "loss": 0.2603,
      "num_input_tokens_seen": 253232,
      "step": 295
    },
    {
      "epoch": 0.14144271570014144,
      "grad_norm": 0.07302112132310867,
      "learning_rate": 0.29995864162713093,
      "loss": 0.283,
      "num_input_tokens_seen": 257456,
      "step": 300
    },
    {
      "epoch": 0.1438000942951438,
      "grad_norm": 0.02899056300520897,
      "learning_rate": 0.2999572469048079,
      "loss": 0.4889,
      "num_input_tokens_seen": 262112,
      "step": 305
    },
    {
      "epoch": 0.14615747289014616,
      "grad_norm": 0.04443933069705963,
      "learning_rate": 0.29995582905719287,
      "loss": 0.2237,
      "num_input_tokens_seen": 266208,
      "step": 310
    },
    {
      "epoch": 0.1485148514851485,
      "grad_norm": 0.05909077078104019,
      "learning_rate": 0.2999543880845046,
      "loss": 0.2465,
      "num_input_tokens_seen": 269792,
      "step": 315
    },
    {
      "epoch": 0.15087223008015088,
      "grad_norm": 0.03209182620048523,
      "learning_rate": 0.2999529239869652,
      "loss": 0.2853,
      "num_input_tokens_seen": 274000,
      "step": 320
    },
    {
      "epoch": 0.15322960867515323,
      "grad_norm": 0.04247352480888367,
      "learning_rate": 0.2999514367648005,
      "loss": 0.2556,
      "num_input_tokens_seen": 278304,
      "step": 325
    },
    {
      "epoch": 0.15558698727015557,
      "grad_norm": 0.1382414549589157,
      "learning_rate": 0.29994992641823987,
      "loss": 0.2522,
      "num_input_tokens_seen": 282976,
      "step": 330
    },
    {
      "epoch": 0.15794436586515795,
      "grad_norm": 0.1388586014509201,
      "learning_rate": 0.29994839294751613,
      "loss": 0.2722,
      "num_input_tokens_seen": 287584,
      "step": 335
    },
    {
      "epoch": 0.1603017444601603,
      "grad_norm": 0.06847704946994781,
      "learning_rate": 0.29994683635286584,
      "loss": 0.254,
      "num_input_tokens_seen": 293632,
      "step": 340
    },
    {
      "epoch": 0.16265912305516267,
      "grad_norm": 0.15747271478176117,
      "learning_rate": 0.2999452566345291,
      "loss": 0.2665,
      "num_input_tokens_seen": 297712,
      "step": 345
    },
    {
      "epoch": 0.16501650165016502,
      "grad_norm": 0.45045727491378784,
      "learning_rate": 0.2999436537927494,
      "loss": 0.6767,
      "num_input_tokens_seen": 302240,
      "step": 350
    },
    {
      "epoch": 0.16737388024516736,
      "grad_norm": 0.05649186298251152,
      "learning_rate": 0.299942027827774,
      "loss": 0.3598,
      "num_input_tokens_seen": 306336,
      "step": 355
    },
    {
      "epoch": 0.16973125884016974,
      "grad_norm": 0.06870386004447937,
      "learning_rate": 0.29994037873985363,
      "loss": 0.3187,
      "num_input_tokens_seen": 310864,
      "step": 360
    },
    {
      "epoch": 0.17208863743517208,
      "grad_norm": 0.08070796728134155,
      "learning_rate": 0.29993870652924254,
      "loss": 0.2599,
      "num_input_tokens_seen": 315072,
      "step": 365
    },
    {
      "epoch": 0.17444601603017446,
      "grad_norm": 0.09333045780658722,
      "learning_rate": 0.29993701119619876,
      "loss": 0.3068,
      "num_input_tokens_seen": 319552,
      "step": 370
    },
    {
      "epoch": 0.1768033946251768,
      "grad_norm": 0.05553876981139183,
      "learning_rate": 0.2999352927409835,
      "loss": 0.3616,
      "num_input_tokens_seen": 323952,
      "step": 375
    },
    {
      "epoch": 0.17916077322017915,
      "grad_norm": 0.04436880350112915,
      "learning_rate": 0.29993355116386194,
      "loss": 0.2284,
      "num_input_tokens_seen": 327856,
      "step": 380
    },
    {
      "epoch": 0.18151815181518152,
      "grad_norm": 0.011247720569372177,
      "learning_rate": 0.29993178646510266,
      "loss": 0.1212,
      "num_input_tokens_seen": 332384,
      "step": 385
    },
    {
      "epoch": 0.18387553041018387,
      "grad_norm": 0.05439724773168564,
      "learning_rate": 0.2999299986449777,
      "loss": 0.353,
      "num_input_tokens_seen": 336992,
      "step": 390
    },
    {
      "epoch": 0.18623290900518624,
      "grad_norm": 0.06764350831508636,
      "learning_rate": 0.29992818770376284,
      "loss": 0.2543,
      "num_input_tokens_seen": 340592,
      "step": 395
    },
    {
      "epoch": 0.1885902876001886,
      "grad_norm": 0.055654413998126984,
      "learning_rate": 0.29992635364173725,
      "loss": 0.3882,
      "num_input_tokens_seen": 344560,
      "step": 400
    },
    {
      "epoch": 0.1885902876001886,
      "eval_loss": 0.25987884402275085,
      "eval_runtime": 21.9212,
      "eval_samples_per_second": 43.018,
      "eval_steps_per_second": 21.532,
      "num_input_tokens_seen": 344560,
      "step": 400
    },
    {
      "epoch": 0.19094766619519093,
      "grad_norm": 0.0670490488409996,
      "learning_rate": 0.2999244964591839,
      "loss": 0.3973,
      "num_input_tokens_seen": 349088,
      "step": 405
    },
    {
      "epoch": 0.1933050447901933,
      "grad_norm": 0.03140193969011307,
      "learning_rate": 0.2999226161563891,
      "loss": 0.3203,
      "num_input_tokens_seen": 353472,
      "step": 410
    },
    {
      "epoch": 0.19566242338519566,
      "grad_norm": 0.02117753215134144,
      "learning_rate": 0.2999207127336429,
      "loss": 0.2315,
      "num_input_tokens_seen": 357904,
      "step": 415
    },
    {
      "epoch": 0.19801980198019803,
      "grad_norm": 0.037303805351257324,
      "learning_rate": 0.2999187861912387,
      "loss": 0.2653,
      "num_input_tokens_seen": 362448,
      "step": 420
    },
    {
      "epoch": 0.20037718057520038,
      "grad_norm": 0.018823646008968353,
      "learning_rate": 0.2999168365294737,
      "loss": 0.2847,
      "num_input_tokens_seen": 367744,
      "step": 425
    },
    {
      "epoch": 0.20273455917020272,
      "grad_norm": 0.028078552335500717,
      "learning_rate": 0.29991486374864856,
      "loss": 0.2498,
      "num_input_tokens_seen": 371920,
      "step": 430
    },
    {
      "epoch": 0.2050919377652051,
      "grad_norm": 0.017890460789203644,
      "learning_rate": 0.29991286784906745,
      "loss": 0.2245,
      "num_input_tokens_seen": 376832,
      "step": 435
    },
    {
      "epoch": 0.20744931636020744,
      "grad_norm": 0.03922800347208977,
      "learning_rate": 0.2999108488310382,
      "loss": 0.2555,
      "num_input_tokens_seen": 380800,
      "step": 440
    },
    {
      "epoch": 0.20980669495520982,
      "grad_norm": 0.01561557874083519,
      "learning_rate": 0.29990880669487213,
      "loss": 0.3052,
      "num_input_tokens_seen": 384736,
      "step": 445
    },
    {
      "epoch": 0.21216407355021216,
      "grad_norm": 0.03607628121972084,
      "learning_rate": 0.29990674144088425,
      "loss": 0.2627,
      "num_input_tokens_seen": 389840,
      "step": 450
    },
    {
      "epoch": 0.2145214521452145,
      "grad_norm": 0.012515510432422161,
      "learning_rate": 0.299904653069393,
      "loss": 0.2101,
      "num_input_tokens_seen": 393680,
      "step": 455
    },
    {
      "epoch": 0.21687883074021688,
      "grad_norm": 0.021081510931253433,
      "learning_rate": 0.29990254158072044,
      "loss": 0.2589,
      "num_input_tokens_seen": 397328,
      "step": 460
    },
    {
      "epoch": 0.21923620933521923,
      "grad_norm": 0.01580131985247135,
      "learning_rate": 0.2999004069751921,
      "loss": 0.2122,
      "num_input_tokens_seen": 401648,
      "step": 465
    },
    {
      "epoch": 0.2215935879302216,
      "grad_norm": 0.050282612442970276,
      "learning_rate": 0.2998982492531373,
      "loss": 0.3045,
      "num_input_tokens_seen": 405744,
      "step": 470
    },
    {
      "epoch": 0.22395096652522395,
      "grad_norm": 0.01552183460444212,
      "learning_rate": 0.2998960684148887,
      "loss": 0.2318,
      "num_input_tokens_seen": 410016,
      "step": 475
    },
    {
      "epoch": 0.2263083451202263,
      "grad_norm": 0.027648374438285828,
      "learning_rate": 0.29989386446078264,
      "loss": 0.2629,
      "num_input_tokens_seen": 413680,
      "step": 480
    },
    {
      "epoch": 0.22866572371522867,
      "grad_norm": 0.030563373118638992,
      "learning_rate": 0.299891637391159,
      "loss": 0.2688,
      "num_input_tokens_seen": 418064,
      "step": 485
    },
    {
      "epoch": 0.23102310231023102,
      "grad_norm": 0.03588169813156128,
      "learning_rate": 0.2998893872063612,
      "loss": 0.2644,
      "num_input_tokens_seen": 422400,
      "step": 490
    },
    {
      "epoch": 0.2333804809052334,
      "grad_norm": 0.012404967099428177,
      "learning_rate": 0.2998871139067363,
      "loss": 0.2515,
      "num_input_tokens_seen": 426688,
      "step": 495
    },
    {
      "epoch": 0.23573785950023574,
      "grad_norm": 0.0213252454996109,
      "learning_rate": 0.2998848174926348,
      "loss": 0.231,
      "num_input_tokens_seen": 431264,
      "step": 500
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 0.018889490514993668,
      "learning_rate": 0.2998824979644109,
      "loss": 0.2347,
      "num_input_tokens_seen": 435184,
      "step": 505
    },
    {
      "epoch": 0.24045261669024046,
      "grad_norm": 0.03740219399333,
      "learning_rate": 0.29988015532242224,
      "loss": 0.2503,
      "num_input_tokens_seen": 439376,
      "step": 510
    },
    {
      "epoch": 0.2428099952852428,
      "grad_norm": 0.010630019009113312,
      "learning_rate": 0.29987778956703015,
      "loss": 0.2142,
      "num_input_tokens_seen": 443472,
      "step": 515
    },
    {
      "epoch": 0.24516737388024518,
      "grad_norm": 0.014807060360908508,
      "learning_rate": 0.2998754006985994,
      "loss": 0.2627,
      "num_input_tokens_seen": 447904,
      "step": 520
    },
    {
      "epoch": 0.24752475247524752,
      "grad_norm": 0.11478953063488007,
      "learning_rate": 0.29987298871749846,
      "loss": 0.2388,
      "num_input_tokens_seen": 452928,
      "step": 525
    },
    {
      "epoch": 0.24988213107024987,
      "grad_norm": 0.018875492736697197,
      "learning_rate": 0.2998705536240992,
      "loss": 0.2299,
      "num_input_tokens_seen": 456768,
      "step": 530
    },
    {
      "epoch": 0.2522395096652522,
      "grad_norm": 0.01869705691933632,
      "learning_rate": 0.2998680954187772,
      "loss": 0.1783,
      "num_input_tokens_seen": 461040,
      "step": 535
    },
    {
      "epoch": 0.2545968882602546,
      "grad_norm": 0.05137253180146217,
      "learning_rate": 0.2998656141019115,
      "loss": 0.3477,
      "num_input_tokens_seen": 465088,
      "step": 540
    },
    {
      "epoch": 0.25695426685525696,
      "grad_norm": 0.03432634472846985,
      "learning_rate": 0.2998631096738848,
      "loss": 0.1988,
      "num_input_tokens_seen": 469568,
      "step": 545
    },
    {
      "epoch": 0.2593116454502593,
      "grad_norm": 0.024540293961763382,
      "learning_rate": 0.29986058213508326,
      "loss": 0.2772,
      "num_input_tokens_seen": 473264,
      "step": 550
    },
    {
      "epoch": 0.26166902404526166,
      "grad_norm": 0.02988324500620365,
      "learning_rate": 0.29985803148589674,
      "loss": 0.2602,
      "num_input_tokens_seen": 477792,
      "step": 555
    },
    {
      "epoch": 0.264026402640264,
      "grad_norm": 0.02994399517774582,
      "learning_rate": 0.2998554577267185,
      "loss": 0.252,
      "num_input_tokens_seen": 481856,
      "step": 560
    },
    {
      "epoch": 0.2663837812352664,
      "grad_norm": 0.03604767099022865,
      "learning_rate": 0.2998528608579455,
      "loss": 0.2275,
      "num_input_tokens_seen": 487024,
      "step": 565
    },
    {
      "epoch": 0.26874115983026875,
      "grad_norm": 0.054496679455041885,
      "learning_rate": 0.2998502408799781,
      "loss": 0.2532,
      "num_input_tokens_seen": 490816,
      "step": 570
    },
    {
      "epoch": 0.2710985384252711,
      "grad_norm": 0.05150502547621727,
      "learning_rate": 0.2998475977932205,
      "loss": 0.2135,
      "num_input_tokens_seen": 495696,
      "step": 575
    },
    {
      "epoch": 0.27345591702027344,
      "grad_norm": 0.06451847404241562,
      "learning_rate": 0.29984493159808023,
      "loss": 0.2666,
      "num_input_tokens_seen": 500800,
      "step": 580
    },
    {
      "epoch": 0.2758132956152758,
      "grad_norm": 0.07649912685155869,
      "learning_rate": 0.29984224229496836,
      "loss": 0.2301,
      "num_input_tokens_seen": 504864,
      "step": 585
    },
    {
      "epoch": 0.2781706742102782,
      "grad_norm": 0.0168441291898489,
      "learning_rate": 0.2998395298842998,
      "loss": 0.1898,
      "num_input_tokens_seen": 509104,
      "step": 590
    },
    {
      "epoch": 0.28052805280528054,
      "grad_norm": 0.015987364575266838,
      "learning_rate": 0.29983679436649263,
      "loss": 0.3044,
      "num_input_tokens_seen": 512720,
      "step": 595
    },
    {
      "epoch": 0.2828854314002829,
      "grad_norm": 0.027446391060948372,
      "learning_rate": 0.2998340357419689,
      "loss": 0.262,
      "num_input_tokens_seen": 517536,
      "step": 600
    },
    {
      "epoch": 0.2828854314002829,
      "eval_loss": 0.3626938462257385,
      "eval_runtime": 21.8854,
      "eval_samples_per_second": 43.088,
      "eval_steps_per_second": 21.567,
      "num_input_tokens_seen": 517536,
      "step": 600
    },
    {
      "epoch": 0.28524280999528523,
      "grad_norm": 0.020620601251721382,
      "learning_rate": 0.29983125401115385,
      "loss": 0.3096,
      "num_input_tokens_seen": 521488,
      "step": 605
    },
    {
      "epoch": 0.2876001885902876,
      "grad_norm": 0.009467436000704765,
      "learning_rate": 0.29982844917447654,
      "loss": 0.1952,
      "num_input_tokens_seen": 525808,
      "step": 610
    },
    {
      "epoch": 0.28995756718529,
      "grad_norm": 0.022830162197351456,
      "learning_rate": 0.2998256212323695,
      "loss": 0.2815,
      "num_input_tokens_seen": 529840,
      "step": 615
    },
    {
      "epoch": 0.2923149457802923,
      "grad_norm": 0.011697076261043549,
      "learning_rate": 0.29982277018526887,
      "loss": 0.2466,
      "num_input_tokens_seen": 534096,
      "step": 620
    },
    {
      "epoch": 0.29467232437529467,
      "grad_norm": 0.023526204749941826,
      "learning_rate": 0.2998198960336143,
      "loss": 0.2343,
      "num_input_tokens_seen": 538752,
      "step": 625
    },
    {
      "epoch": 0.297029702970297,
      "grad_norm": 0.016746602952480316,
      "learning_rate": 0.299816998777849,
      "loss": 0.2687,
      "num_input_tokens_seen": 543968,
      "step": 630
    },
    {
      "epoch": 0.29938708156529936,
      "grad_norm": 0.027757268399000168,
      "learning_rate": 0.2998140784184197,
      "loss": 0.205,
      "num_input_tokens_seen": 548784,
      "step": 635
    },
    {
      "epoch": 0.30174446016030176,
      "grad_norm": 0.23165562748908997,
      "learning_rate": 0.2998111349557769,
      "loss": 0.456,
      "num_input_tokens_seen": 552880,
      "step": 640
    },
    {
      "epoch": 0.3041018387553041,
      "grad_norm": 0.016647884622216225,
      "learning_rate": 0.29980816839037444,
      "loss": 0.2253,
      "num_input_tokens_seen": 557184,
      "step": 645
    },
    {
      "epoch": 0.30645921735030646,
      "grad_norm": 0.019891247153282166,
      "learning_rate": 0.2998051787226698,
      "loss": 0.2597,
      "num_input_tokens_seen": 561840,
      "step": 650
    },
    {
      "epoch": 0.3088165959453088,
      "grad_norm": 0.028244683519005775,
      "learning_rate": 0.29980216595312403,
      "loss": 0.2352,
      "num_input_tokens_seen": 566736,
      "step": 655
    },
    {
      "epoch": 0.31117397454031115,
      "grad_norm": 0.02917248010635376,
      "learning_rate": 0.29979913008220177,
      "loss": 0.2471,
      "num_input_tokens_seen": 570528,
      "step": 660
    },
    {
      "epoch": 0.31353135313531355,
      "grad_norm": 0.04081382229924202,
      "learning_rate": 0.2997960711103711,
      "loss": 0.2323,
      "num_input_tokens_seen": 574720,
      "step": 665
    },
    {
      "epoch": 0.3158887317303159,
      "grad_norm": 0.016577957198023796,
      "learning_rate": 0.29979298903810386,
      "loss": 0.2201,
      "num_input_tokens_seen": 578912,
      "step": 670
    },
    {
      "epoch": 0.31824611032531824,
      "grad_norm": 0.014963102526962757,
      "learning_rate": 0.29978988386587524,
      "loss": 0.2517,
      "num_input_tokens_seen": 584544,
      "step": 675
    },
    {
      "epoch": 0.3206034889203206,
      "grad_norm": 0.012951698154211044,
      "learning_rate": 0.2997867555941642,
      "loss": 0.1883,
      "num_input_tokens_seen": 588752,
      "step": 680
    },
    {
      "epoch": 0.32296086751532294,
      "grad_norm": 0.023902473971247673,
      "learning_rate": 0.299783604223453,
      "loss": 0.2882,
      "num_input_tokens_seen": 593024,
      "step": 685
    },
    {
      "epoch": 0.32531824611032534,
      "grad_norm": 0.022149108350276947,
      "learning_rate": 0.29978042975422786,
      "loss": 0.2218,
      "num_input_tokens_seen": 597280,
      "step": 690
    },
    {
      "epoch": 0.3276756247053277,
      "grad_norm": 0.02305581234395504,
      "learning_rate": 0.29977723218697816,
      "loss": 0.2336,
      "num_input_tokens_seen": 601904,
      "step": 695
    },
    {
      "epoch": 0.33003300330033003,
      "grad_norm": 0.011331078596413136,
      "learning_rate": 0.299774011522197,
      "loss": 0.2282,
      "num_input_tokens_seen": 606368,
      "step": 700
    },
    {
      "epoch": 0.3323903818953324,
      "grad_norm": 0.019083211198449135,
      "learning_rate": 0.29977076776038114,
      "loss": 0.2236,
      "num_input_tokens_seen": 610096,
      "step": 705
    },
    {
      "epoch": 0.3347477604903347,
      "grad_norm": 0.047596994787454605,
      "learning_rate": 0.2997675009020307,
      "loss": 0.2257,
      "num_input_tokens_seen": 614288,
      "step": 710
    },
    {
      "epoch": 0.3371051390853371,
      "grad_norm": 0.019462116062641144,
      "learning_rate": 0.2997642109476496,
      "loss": 0.2999,
      "num_input_tokens_seen": 619344,
      "step": 715
    },
    {
      "epoch": 0.33946251768033947,
      "grad_norm": 0.032761137932538986,
      "learning_rate": 0.299760897897745,
      "loss": 0.2423,
      "num_input_tokens_seen": 624240,
      "step": 720
    },
    {
      "epoch": 0.3418198962753418,
      "grad_norm": 0.024460729211568832,
      "learning_rate": 0.29975756175282803,
      "loss": 0.2327,
      "num_input_tokens_seen": 629312,
      "step": 725
    },
    {
      "epoch": 0.34417727487034416,
      "grad_norm": 0.008055005222558975,
      "learning_rate": 0.29975420251341306,
      "loss": 0.1856,
      "num_input_tokens_seen": 633952,
      "step": 730
    },
    {
      "epoch": 0.3465346534653465,
      "grad_norm": 0.009399060159921646,
      "learning_rate": 0.29975082018001814,
      "loss": 0.282,
      "num_input_tokens_seen": 638352,
      "step": 735
    },
    {
      "epoch": 0.3488920320603489,
      "grad_norm": 0.013479001820087433,
      "learning_rate": 0.2997474147531648,
      "loss": 0.2542,
      "num_input_tokens_seen": 642656,
      "step": 740
    },
    {
      "epoch": 0.35124941065535126,
      "grad_norm": 0.013776581734418869,
      "learning_rate": 0.29974398623337833,
      "loss": 0.2361,
      "num_input_tokens_seen": 647696,
      "step": 745
    },
    {
      "epoch": 0.3536067892503536,
      "grad_norm": 0.03046252578496933,
      "learning_rate": 0.2997405346211873,
      "loss": 0.3057,
      "num_input_tokens_seen": 652832,
      "step": 750
    },
    {
      "epoch": 0.35596416784535595,
      "grad_norm": 0.024462277069687843,
      "learning_rate": 0.2997370599171241,
      "loss": 0.2421,
      "num_input_tokens_seen": 656768,
      "step": 755
    },
    {
      "epoch": 0.3583215464403583,
      "grad_norm": 0.012806993909180164,
      "learning_rate": 0.2997335621217246,
      "loss": 0.2402,
      "num_input_tokens_seen": 660624,
      "step": 760
    },
    {
      "epoch": 0.3606789250353607,
      "grad_norm": 0.01392971258610487,
      "learning_rate": 0.29973004123552816,
      "loss": 0.1834,
      "num_input_tokens_seen": 665488,
      "step": 765
    },
    {
      "epoch": 0.36303630363036304,
      "grad_norm": 0.014303174801170826,
      "learning_rate": 0.2997264972590777,
      "loss": 0.2897,
      "num_input_tokens_seen": 669504,
      "step": 770
    },
    {
      "epoch": 0.3653936822253654,
      "grad_norm": 0.024460889399051666,
      "learning_rate": 0.29972293019291973,
      "loss": 0.3049,
      "num_input_tokens_seen": 674032,
      "step": 775
    },
    {
      "epoch": 0.36775106082036774,
      "grad_norm": 0.03441541641950607,
      "learning_rate": 0.2997193400376045,
      "loss": 0.2487,
      "num_input_tokens_seen": 678384,
      "step": 780
    },
    {
      "epoch": 0.3701084394153701,
      "grad_norm": 0.014286152087152004,
      "learning_rate": 0.2997157267936854,
      "loss": 0.2959,
      "num_input_tokens_seen": 682944,
      "step": 785
    },
    {
      "epoch": 0.3724658180103725,
      "grad_norm": 0.015395323745906353,
      "learning_rate": 0.2997120904617199,
      "loss": 0.2552,
      "num_input_tokens_seen": 687072,
      "step": 790
    },
    {
      "epoch": 0.37482319660537483,
      "grad_norm": 0.030791128054261208,
      "learning_rate": 0.29970843104226863,
      "loss": 0.2168,
      "num_input_tokens_seen": 691632,
      "step": 795
    },
    {
      "epoch": 0.3771805752003772,
      "grad_norm": 0.01721709966659546,
      "learning_rate": 0.2997047485358959,
      "loss": 0.1964,
      "num_input_tokens_seen": 696016,
      "step": 800
    },
    {
      "epoch": 0.3771805752003772,
      "eval_loss": 0.22650428116321564,
      "eval_runtime": 21.9206,
      "eval_samples_per_second": 43.019,
      "eval_steps_per_second": 21.532,
      "num_input_tokens_seen": 696016,
      "step": 800
    },
    {
      "epoch": 0.3795379537953795,
      "grad_norm": 0.032669808715581894,
      "learning_rate": 0.2997010429431697,
      "loss": 0.2377,
      "num_input_tokens_seen": 700240,
      "step": 805
    },
    {
      "epoch": 0.38189533239038187,
      "grad_norm": 0.014330493286252022,
      "learning_rate": 0.29969731426466134,
      "loss": 0.2736,
      "num_input_tokens_seen": 704448,
      "step": 810
    },
    {
      "epoch": 0.38425271098538427,
      "grad_norm": 0.01586211286485195,
      "learning_rate": 0.299693562500946,
      "loss": 0.2426,
      "num_input_tokens_seen": 707968,
      "step": 815
    },
    {
      "epoch": 0.3866100895803866,
      "grad_norm": 0.027338912710547447,
      "learning_rate": 0.29968978765260207,
      "loss": 0.2485,
      "num_input_tokens_seen": 712272,
      "step": 820
    },
    {
      "epoch": 0.38896746817538896,
      "grad_norm": 0.012939891777932644,
      "learning_rate": 0.2996859897202118,
      "loss": 0.1615,
      "num_input_tokens_seen": 716784,
      "step": 825
    },
    {
      "epoch": 0.3913248467703913,
      "grad_norm": 0.02211010456085205,
      "learning_rate": 0.2996821687043609,
      "loss": 0.2933,
      "num_input_tokens_seen": 721296,
      "step": 830
    },
    {
      "epoch": 0.39368222536539366,
      "grad_norm": 0.015524206683039665,
      "learning_rate": 0.2996783246056384,
      "loss": 0.2197,
      "num_input_tokens_seen": 725536,
      "step": 835
    },
    {
      "epoch": 0.39603960396039606,
      "grad_norm": 0.02145790494978428,
      "learning_rate": 0.29967445742463744,
      "loss": 0.2467,
      "num_input_tokens_seen": 730384,
      "step": 840
    },
    {
      "epoch": 0.3983969825553984,
      "grad_norm": 0.01908414252102375,
      "learning_rate": 0.29967056716195417,
      "loss": 0.2577,
      "num_input_tokens_seen": 734960,
      "step": 845
    },
    {
      "epoch": 0.40075436115040075,
      "grad_norm": 0.01674271561205387,
      "learning_rate": 0.2996666538181885,
      "loss": 0.2306,
      "num_input_tokens_seen": 739040,
      "step": 850
    },
    {
      "epoch": 0.4031117397454031,
      "grad_norm": 0.017238173633813858,
      "learning_rate": 0.29966271739394407,
      "loss": 0.2691,
      "num_input_tokens_seen": 743184,
      "step": 855
    },
    {
      "epoch": 0.40546911834040544,
      "grad_norm": 0.009943000972270966,
      "learning_rate": 0.29965875788982776,
      "loss": 0.2181,
      "num_input_tokens_seen": 747056,
      "step": 860
    },
    {
      "epoch": 0.40782649693540785,
      "grad_norm": 0.06390685588121414,
      "learning_rate": 0.2996547753064503,
      "loss": 0.243,
      "num_input_tokens_seen": 751184,
      "step": 865
    },
    {
      "epoch": 0.4101838755304102,
      "grad_norm": 0.019871389493346214,
      "learning_rate": 0.29965076964442583,
      "loss": 0.2412,
      "num_input_tokens_seen": 755392,
      "step": 870
    },
    {
      "epoch": 0.41254125412541254,
      "grad_norm": 0.01780155673623085,
      "learning_rate": 0.299646740904372,
      "loss": 0.2466,
      "num_input_tokens_seen": 759648,
      "step": 875
    },
    {
      "epoch": 0.4148986327204149,
      "grad_norm": 0.009347254410386086,
      "learning_rate": 0.29964268908691016,
      "loss": 0.2328,
      "num_input_tokens_seen": 763792,
      "step": 880
    },
    {
      "epoch": 0.41725601131541723,
      "grad_norm": 0.010657994076609612,
      "learning_rate": 0.29963861419266513,
      "loss": 0.2629,
      "num_input_tokens_seen": 768944,
      "step": 885
    },
    {
      "epoch": 0.41961338991041963,
      "grad_norm": 0.010821755044162273,
      "learning_rate": 0.29963451622226533,
      "loss": 0.2335,
      "num_input_tokens_seen": 773360,
      "step": 890
    },
    {
      "epoch": 0.421970768505422,
      "grad_norm": 0.21754032373428345,
      "learning_rate": 0.29963039517634277,
      "loss": 0.2682,
      "num_input_tokens_seen": 777808,
      "step": 895
    },
    {
      "epoch": 0.4243281471004243,
      "grad_norm": 0.1522066444158554,
      "learning_rate": 0.2996262510555328,
      "loss": 0.304,
      "num_input_tokens_seen": 784416,
      "step": 900
    },
    {
      "epoch": 0.42668552569542667,
      "grad_norm": 0.04216880723834038,
      "learning_rate": 0.2996220838604746,
      "loss": 0.2466,
      "num_input_tokens_seen": 788464,
      "step": 905
    },
    {
      "epoch": 0.429042904290429,
      "grad_norm": 0.010158684104681015,
      "learning_rate": 0.29961789359181085,
      "loss": 0.2013,
      "num_input_tokens_seen": 793232,
      "step": 910
    },
    {
      "epoch": 0.4314002828854314,
      "grad_norm": 0.013883883133530617,
      "learning_rate": 0.29961368025018764,
      "loss": 0.2184,
      "num_input_tokens_seen": 797872,
      "step": 915
    },
    {
      "epoch": 0.43375766148043376,
      "grad_norm": 0.02833622694015503,
      "learning_rate": 0.2996094438362548,
      "loss": 0.2429,
      "num_input_tokens_seen": 802000,
      "step": 920
    },
    {
      "epoch": 0.4361150400754361,
      "grad_norm": 0.007342576049268246,
      "learning_rate": 0.2996051843506657,
      "loss": 0.2581,
      "num_input_tokens_seen": 806192,
      "step": 925
    },
    {
      "epoch": 0.43847241867043846,
      "grad_norm": 0.010723326355218887,
      "learning_rate": 0.299600901794077,
      "loss": 0.259,
      "num_input_tokens_seen": 810032,
      "step": 930
    },
    {
      "epoch": 0.4408297972654408,
      "grad_norm": 0.009109522216022015,
      "learning_rate": 0.29959659616714923,
      "loss": 0.2424,
      "num_input_tokens_seen": 813776,
      "step": 935
    },
    {
      "epoch": 0.4431871758604432,
      "grad_norm": 0.023440705612301826,
      "learning_rate": 0.2995922674705464,
      "loss": 0.2327,
      "num_input_tokens_seen": 818272,
      "step": 940
    },
    {
      "epoch": 0.44554455445544555,
      "grad_norm": 0.006356361787766218,
      "learning_rate": 0.2995879157049361,
      "loss": 0.1758,
      "num_input_tokens_seen": 823504,
      "step": 945
    },
    {
      "epoch": 0.4479019330504479,
      "grad_norm": 0.004717394709587097,
      "learning_rate": 0.2995835408709893,
      "loss": 0.4254,
      "num_input_tokens_seen": 827280,
      "step": 950
    },
    {
      "epoch": 0.45025931164545024,
      "grad_norm": 0.024635834619402885,
      "learning_rate": 0.29957914296938076,
      "loss": 0.2381,
      "num_input_tokens_seen": 831072,
      "step": 955
    },
    {
      "epoch": 0.4526166902404526,
      "grad_norm": 0.018351038917899132,
      "learning_rate": 0.2995747220007886,
      "loss": 0.2596,
      "num_input_tokens_seen": 834928,
      "step": 960
    },
    {
      "epoch": 0.454974068835455,
      "grad_norm": 0.004620480351150036,
      "learning_rate": 0.2995702779658947,
      "loss": 0.2103,
      "num_input_tokens_seen": 838800,
      "step": 965
    },
    {
      "epoch": 0.45733144743045734,
      "grad_norm": 0.006804523058235645,
      "learning_rate": 0.29956581086538425,
      "loss": 0.247,
      "num_input_tokens_seen": 842992,
      "step": 970
    },
    {
      "epoch": 0.4596888260254597,
      "grad_norm": 0.02102152444422245,
      "learning_rate": 0.2995613206999462,
      "loss": 0.2697,
      "num_input_tokens_seen": 847216,
      "step": 975
    },
    {
      "epoch": 0.46204620462046203,
      "grad_norm": 0.02126171998679638,
      "learning_rate": 0.29955680747027297,
      "loss": 0.2492,
      "num_input_tokens_seen": 851152,
      "step": 980
    },
    {
      "epoch": 0.4644035832154644,
      "grad_norm": 0.022664852440357208,
      "learning_rate": 0.2995522711770607,
      "loss": 0.26,
      "num_input_tokens_seen": 855520,
      "step": 985
    },
    {
      "epoch": 0.4667609618104668,
      "grad_norm": 0.010885710828006268,
      "learning_rate": 0.2995477118210087,
      "loss": 0.2049,
      "num_input_tokens_seen": 860288,
      "step": 990
    },
    {
      "epoch": 0.4691183404054691,
      "grad_norm": 0.00547809200361371,
      "learning_rate": 0.29954312940282024,
      "loss": 0.209,
      "num_input_tokens_seen": 864848,
      "step": 995
    },
    {
      "epoch": 0.47147571900047147,
      "grad_norm": 0.0102836973965168,
      "learning_rate": 0.29953852392320196,
      "loss": 0.2543,
      "num_input_tokens_seen": 868992,
      "step": 1000
    },
    {
      "epoch": 0.47147571900047147,
      "eval_loss": 0.22157074511051178,
      "eval_runtime": 21.9029,
      "eval_samples_per_second": 43.054,
      "eval_steps_per_second": 21.55,
      "num_input_tokens_seen": 868992,
      "step": 1000
    },
    {
      "epoch": 0.4738330975954738,
      "grad_norm": 0.011545664630830288,
      "learning_rate": 0.2995338953828641,
      "loss": 0.2249,
      "num_input_tokens_seen": 872576,
      "step": 1005
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 0.008610946126282215,
      "learning_rate": 0.2995292437825204,
      "loss": 0.2423,
      "num_input_tokens_seen": 876752,
      "step": 1010
    },
    {
      "epoch": 0.47854785478547857,
      "grad_norm": 0.008063350804150105,
      "learning_rate": 0.29952456912288816,
      "loss": 0.2478,
      "num_input_tokens_seen": 880720,
      "step": 1015
    },
    {
      "epoch": 0.4809052333804809,
      "grad_norm": 0.010864391922950745,
      "learning_rate": 0.2995198714046884,
      "loss": 0.2403,
      "num_input_tokens_seen": 884528,
      "step": 1020
    },
    {
      "epoch": 0.48326261197548326,
      "grad_norm": 0.00809762068092823,
      "learning_rate": 0.2995151506286454,
      "loss": 0.234,
      "num_input_tokens_seen": 889360,
      "step": 1025
    },
    {
      "epoch": 0.4856199905704856,
      "grad_norm": 0.01353034283965826,
      "learning_rate": 0.2995104067954873,
      "loss": 0.2257,
      "num_input_tokens_seen": 893504,
      "step": 1030
    },
    {
      "epoch": 0.48797736916548795,
      "grad_norm": 0.006150421686470509,
      "learning_rate": 0.2995056399059456,
      "loss": 0.2528,
      "num_input_tokens_seen": 897392,
      "step": 1035
    },
    {
      "epoch": 0.49033474776049035,
      "grad_norm": 0.008271138183772564,
      "learning_rate": 0.2995008499607554,
      "loss": 0.2202,
      "num_input_tokens_seen": 901952,
      "step": 1040
    },
    {
      "epoch": 0.4926921263554927,
      "grad_norm": 0.010571779683232307,
      "learning_rate": 0.2994960369606554,
      "loss": 0.2182,
      "num_input_tokens_seen": 906720,
      "step": 1045
    },
    {
      "epoch": 0.49504950495049505,
      "grad_norm": 0.013821846805512905,
      "learning_rate": 0.2994912009063878,
      "loss": 0.2535,
      "num_input_tokens_seen": 911456,
      "step": 1050
    },
    {
      "epoch": 0.4974068835454974,
      "grad_norm": 0.015538020059466362,
      "learning_rate": 0.29948634179869843,
      "loss": 0.2251,
      "num_input_tokens_seen": 915760,
      "step": 1055
    },
    {
      "epoch": 0.49976426214049974,
      "grad_norm": 0.022615710273385048,
      "learning_rate": 0.29948145963833656,
      "loss": 0.2744,
      "num_input_tokens_seen": 920016,
      "step": 1060
    },
    {
      "epoch": 0.5021216407355021,
      "grad_norm": 0.009418646804988384,
      "learning_rate": 0.29947655442605514,
      "loss": 0.2259,
      "num_input_tokens_seen": 924720,
      "step": 1065
    },
    {
      "epoch": 0.5044790193305044,
      "grad_norm": 0.014893700368702412,
      "learning_rate": 0.2994716261626106,
      "loss": 0.2194,
      "num_input_tokens_seen": 928208,
      "step": 1070
    },
    {
      "epoch": 0.5068363979255068,
      "grad_norm": 0.010897199623286724,
      "learning_rate": 0.2994666748487629,
      "loss": 0.1953,
      "num_input_tokens_seen": 932064,
      "step": 1075
    },
    {
      "epoch": 0.5091937765205092,
      "grad_norm": 0.011832601390779018,
      "learning_rate": 0.2994617004852756,
      "loss": 0.3198,
      "num_input_tokens_seen": 935712,
      "step": 1080
    },
    {
      "epoch": 0.5115511551155115,
      "grad_norm": 0.00727560929954052,
      "learning_rate": 0.2994567030729159,
      "loss": 0.1944,
      "num_input_tokens_seen": 940096,
      "step": 1085
    },
    {
      "epoch": 0.5139085337105139,
      "grad_norm": 0.010667894966900349,
      "learning_rate": 0.29945168261245436,
      "loss": 0.2348,
      "num_input_tokens_seen": 943872,
      "step": 1090
    },
    {
      "epoch": 0.5162659123055162,
      "grad_norm": 0.01917083188891411,
      "learning_rate": 0.29944663910466524,
      "loss": 0.2307,
      "num_input_tokens_seen": 949296,
      "step": 1095
    },
    {
      "epoch": 0.5186232909005186,
      "grad_norm": 0.01053266879171133,
      "learning_rate": 0.2994415725503263,
      "loss": 0.1967,
      "num_input_tokens_seen": 953040,
      "step": 1100
    },
    {
      "epoch": 0.520980669495521,
      "grad_norm": 0.02649979293346405,
      "learning_rate": 0.29943648295021885,
      "loss": 0.2724,
      "num_input_tokens_seen": 957600,
      "step": 1105
    },
    {
      "epoch": 0.5233380480905233,
      "grad_norm": 0.004224368836730719,
      "learning_rate": 0.2994313703051278,
      "loss": 0.2845,
      "num_input_tokens_seen": 962496,
      "step": 1110
    },
    {
      "epoch": 0.5256954266855257,
      "grad_norm": 0.00910792127251625,
      "learning_rate": 0.29942623461584156,
      "loss": 0.2659,
      "num_input_tokens_seen": 966688,
      "step": 1115
    },
    {
      "epoch": 0.528052805280528,
      "grad_norm": 0.006304461974650621,
      "learning_rate": 0.29942107588315214,
      "loss": 0.2732,
      "num_input_tokens_seen": 969984,
      "step": 1120
    },
    {
      "epoch": 0.5304101838755304,
      "grad_norm": 0.0066589065827429295,
      "learning_rate": 0.29941589410785513,
      "loss": 0.2305,
      "num_input_tokens_seen": 975280,
      "step": 1125
    },
    {
      "epoch": 0.5327675624705328,
      "grad_norm": 0.02922910265624523,
      "learning_rate": 0.29941068929074954,
      "loss": 0.208,
      "num_input_tokens_seen": 980064,
      "step": 1130
    },
    {
      "epoch": 0.5351249410655351,
      "grad_norm": 0.011774340644478798,
      "learning_rate": 0.2994054614326381,
      "loss": 0.254,
      "num_input_tokens_seen": 985024,
      "step": 1135
    },
    {
      "epoch": 0.5374823196605375,
      "grad_norm": 0.011308026500046253,
      "learning_rate": 0.29940021053432686,
      "loss": 0.2832,
      "num_input_tokens_seen": 989504,
      "step": 1140
    },
    {
      "epoch": 0.5398396982555398,
      "grad_norm": 0.0053458791226148605,
      "learning_rate": 0.29939493659662575,
      "loss": 0.2444,
      "num_input_tokens_seen": 994448,
      "step": 1145
    },
    {
      "epoch": 0.5421970768505422,
      "grad_norm": 0.013303283601999283,
      "learning_rate": 0.299389639620348,
      "loss": 0.2416,
      "num_input_tokens_seen": 998976,
      "step": 1150
    },
    {
      "epoch": 0.5445544554455446,
      "grad_norm": 0.004726475104689598,
      "learning_rate": 0.29938431960631046,
      "loss": 0.2453,
      "num_input_tokens_seen": 1003136,
      "step": 1155
    },
    {
      "epoch": 0.5469118340405469,
      "grad_norm": 0.003945779521018267,
      "learning_rate": 0.2993789765553335,
      "loss": 0.2349,
      "num_input_tokens_seen": 1006512,
      "step": 1160
    },
    {
      "epoch": 0.5492692126355493,
      "grad_norm": 0.0076730637811124325,
      "learning_rate": 0.2993736104682412,
      "loss": 0.2334,
      "num_input_tokens_seen": 1010576,
      "step": 1165
    },
    {
      "epoch": 0.5516265912305516,
      "grad_norm": 0.011619127355515957,
      "learning_rate": 0.299368221345861,
      "loss": 0.2155,
      "num_input_tokens_seen": 1014928,
      "step": 1170
    },
    {
      "epoch": 0.553983969825554,
      "grad_norm": 0.007023296318948269,
      "learning_rate": 0.29936280918902397,
      "loss": 0.1792,
      "num_input_tokens_seen": 1019680,
      "step": 1175
    },
    {
      "epoch": 0.5563413484205564,
      "grad_norm": 0.01198938675224781,
      "learning_rate": 0.2993573739985648,
      "loss": 0.2527,
      "num_input_tokens_seen": 1024112,
      "step": 1180
    },
    {
      "epoch": 0.5586987270155587,
      "grad_norm": 0.017957186326384544,
      "learning_rate": 0.2993519157753216,
      "loss": 0.2582,
      "num_input_tokens_seen": 1028768,
      "step": 1185
    },
    {
      "epoch": 0.5610561056105611,
      "grad_norm": 0.0077983723022043705,
      "learning_rate": 0.2993464345201361,
      "loss": 0.2404,
      "num_input_tokens_seen": 1032288,
      "step": 1190
    },
    {
      "epoch": 0.5634134842055634,
      "grad_norm": 0.019709035754203796,
      "learning_rate": 0.2993409302338536,
      "loss": 0.2598,
      "num_input_tokens_seen": 1036368,
      "step": 1195
    },
    {
      "epoch": 0.5657708628005658,
      "grad_norm": 0.007961167953908443,
      "learning_rate": 0.2993354029173229,
      "loss": 0.2859,
      "num_input_tokens_seen": 1040544,
      "step": 1200
    },
    {
      "epoch": 0.5657708628005658,
      "eval_loss": 0.22787168622016907,
      "eval_runtime": 21.921,
      "eval_samples_per_second": 43.018,
      "eval_steps_per_second": 21.532,
      "num_input_tokens_seen": 1040544,
      "step": 1200
    },
    {
      "epoch": 0.5681282413955682,
      "grad_norm": 0.01525899674743414,
      "learning_rate": 0.2993298525713965,
      "loss": 0.3122,
      "num_input_tokens_seen": 1044576,
      "step": 1205
    },
    {
      "epoch": 0.5704856199905705,
      "grad_norm": 0.003580045886337757,
      "learning_rate": 0.29932427919693017,
      "loss": 0.2322,
      "num_input_tokens_seen": 1049408,
      "step": 1210
    },
    {
      "epoch": 0.5728429985855729,
      "grad_norm": 0.00673142122104764,
      "learning_rate": 0.2993186827947834,
      "loss": 0.2054,
      "num_input_tokens_seen": 1053424,
      "step": 1215
    },
    {
      "epoch": 0.5752003771805752,
      "grad_norm": 0.006981938146054745,
      "learning_rate": 0.2993130633658194,
      "loss": 0.2239,
      "num_input_tokens_seen": 1058128,
      "step": 1220
    },
    {
      "epoch": 0.5775577557755776,
      "grad_norm": 0.005521902348846197,
      "learning_rate": 0.29930742091090456,
      "loss": 0.2125,
      "num_input_tokens_seen": 1062112,
      "step": 1225
    },
    {
      "epoch": 0.57991513437058,
      "grad_norm": 0.006411140784621239,
      "learning_rate": 0.29930175543090914,
      "loss": 0.2233,
      "num_input_tokens_seen": 1066544,
      "step": 1230
    },
    {
      "epoch": 0.5822725129655822,
      "grad_norm": 0.0065961722284555435,
      "learning_rate": 0.2992960669267068,
      "loss": 0.1323,
      "num_input_tokens_seen": 1071280,
      "step": 1235
    },
    {
      "epoch": 0.5846298915605846,
      "grad_norm": 0.010791966691613197,
      "learning_rate": 0.29929035539917476,
      "loss": 0.3092,
      "num_input_tokens_seen": 1075568,
      "step": 1240
    },
    {
      "epoch": 0.5869872701555869,
      "grad_norm": 0.004317036829888821,
      "learning_rate": 0.2992846208491938,
      "loss": 0.2783,
      "num_input_tokens_seen": 1079648,
      "step": 1245
    },
    {
      "epoch": 0.5893446487505893,
      "grad_norm": 0.010308889672160149,
      "learning_rate": 0.2992788632776483,
      "loss": 0.2229,
      "num_input_tokens_seen": 1084080,
      "step": 1250
    },
    {
      "epoch": 0.5917020273455917,
      "grad_norm": 0.00871394481509924,
      "learning_rate": 0.29927308268542613,
      "loss": 0.2297,
      "num_input_tokens_seen": 1087680,
      "step": 1255
    },
    {
      "epoch": 0.594059405940594,
      "grad_norm": 0.010654519312083721,
      "learning_rate": 0.2992672790734187,
      "loss": 0.2159,
      "num_input_tokens_seen": 1092080,
      "step": 1260
    },
    {
      "epoch": 0.5964167845355964,
      "grad_norm": 0.012754036113619804,
      "learning_rate": 0.299261452442521,
      "loss": 0.2184,
      "num_input_tokens_seen": 1096288,
      "step": 1265
    },
    {
      "epoch": 0.5987741631305987,
      "grad_norm": 0.004326207097619772,
      "learning_rate": 0.29925560279363167,
      "loss": 0.2332,
      "num_input_tokens_seen": 1099776,
      "step": 1270
    },
    {
      "epoch": 0.6011315417256011,
      "grad_norm": 0.011387339793145657,
      "learning_rate": 0.29924973012765266,
      "loss": 0.2524,
      "num_input_tokens_seen": 1103440,
      "step": 1275
    },
    {
      "epoch": 0.6034889203206035,
      "grad_norm": 0.00819116085767746,
      "learning_rate": 0.29924383444548974,
      "loss": 0.2292,
      "num_input_tokens_seen": 1107344,
      "step": 1280
    },
    {
      "epoch": 0.6058462989156058,
      "grad_norm": 0.021335134282708168,
      "learning_rate": 0.299237915748052,
      "loss": 0.2507,
      "num_input_tokens_seen": 1111728,
      "step": 1285
    },
    {
      "epoch": 0.6082036775106082,
      "grad_norm": 0.007584770210087299,
      "learning_rate": 0.2992319740362522,
      "loss": 0.2722,
      "num_input_tokens_seen": 1115376,
      "step": 1290
    },
    {
      "epoch": 0.6105610561056105,
      "grad_norm": 0.008759505115449429,
      "learning_rate": 0.2992260093110066,
      "loss": 0.2248,
      "num_input_tokens_seen": 1119184,
      "step": 1295
    },
    {
      "epoch": 0.6129184347006129,
      "grad_norm": 0.007545670494437218,
      "learning_rate": 0.2992200215732352,
      "loss": 0.2334,
      "num_input_tokens_seen": 1123600,
      "step": 1300
    },
    {
      "epoch": 0.6152758132956153,
      "grad_norm": 0.006894233636558056,
      "learning_rate": 0.2992140108238611,
      "loss": 0.236,
      "num_input_tokens_seen": 1127408,
      "step": 1305
    },
    {
      "epoch": 0.6176331918906176,
      "grad_norm": 0.012917310930788517,
      "learning_rate": 0.2992079770638115,
      "loss": 0.2289,
      "num_input_tokens_seen": 1131872,
      "step": 1310
    },
    {
      "epoch": 0.61999057048562,
      "grad_norm": 0.01972266100347042,
      "learning_rate": 0.29920192029401677,
      "loss": 0.2335,
      "num_input_tokens_seen": 1138416,
      "step": 1315
    },
    {
      "epoch": 0.6223479490806223,
      "grad_norm": 0.005994094535708427,
      "learning_rate": 0.2991958405154109,
      "loss": 0.2291,
      "num_input_tokens_seen": 1142144,
      "step": 1320
    },
    {
      "epoch": 0.6247053276756247,
      "grad_norm": 0.00641050236299634,
      "learning_rate": 0.29918973772893154,
      "loss": 0.2399,
      "num_input_tokens_seen": 1146544,
      "step": 1325
    },
    {
      "epoch": 0.6270627062706271,
      "grad_norm": 0.010463126935064793,
      "learning_rate": 0.29918361193551973,
      "loss": 0.2255,
      "num_input_tokens_seen": 1150784,
      "step": 1330
    },
    {
      "epoch": 0.6294200848656294,
      "grad_norm": 0.005755925551056862,
      "learning_rate": 0.29917746313612026,
      "loss": 0.222,
      "num_input_tokens_seen": 1155120,
      "step": 1335
    },
    {
      "epoch": 0.6317774634606318,
      "grad_norm": 0.005667397752404213,
      "learning_rate": 0.29917129133168124,
      "loss": 0.2233,
      "num_input_tokens_seen": 1160256,
      "step": 1340
    },
    {
      "epoch": 0.6341348420556341,
      "grad_norm": 0.004874077159911394,
      "learning_rate": 0.2991650965231546,
      "loss": 0.1927,
      "num_input_tokens_seen": 1164544,
      "step": 1345
    },
    {
      "epoch": 0.6364922206506365,
      "grad_norm": 0.010840351693332195,
      "learning_rate": 0.29915887871149544,
      "loss": 0.2075,
      "num_input_tokens_seen": 1168464,
      "step": 1350
    },
    {
      "epoch": 0.6388495992456389,
      "grad_norm": 0.009263534098863602,
      "learning_rate": 0.2991526378976628,
      "loss": 0.2149,
      "num_input_tokens_seen": 1171888,
      "step": 1355
    },
    {
      "epoch": 0.6412069778406412,
      "grad_norm": 0.005568221677094698,
      "learning_rate": 0.29914637408261896,
      "loss": 0.2563,
      "num_input_tokens_seen": 1175920,
      "step": 1360
    },
    {
      "epoch": 0.6435643564356436,
      "grad_norm": 0.004826083779335022,
      "learning_rate": 0.29914008726733,
      "loss": 0.3011,
      "num_input_tokens_seen": 1179792,
      "step": 1365
    },
    {
      "epoch": 0.6459217350306459,
      "grad_norm": 0.0208759605884552,
      "learning_rate": 0.2991337774527653,
      "loss": 0.2366,
      "num_input_tokens_seen": 1185264,
      "step": 1370
    },
    {
      "epoch": 0.6482791136256483,
      "grad_norm": 0.0211018193513155,
      "learning_rate": 0.2991274446398981,
      "loss": 0.2531,
      "num_input_tokens_seen": 1189472,
      "step": 1375
    },
    {
      "epoch": 0.6506364922206507,
      "grad_norm": 0.017729585990309715,
      "learning_rate": 0.29912108882970484,
      "loss": 0.2272,
      "num_input_tokens_seen": 1193104,
      "step": 1380
    },
    {
      "epoch": 0.652993870815653,
      "grad_norm": 0.02269987016916275,
      "learning_rate": 0.2991147100231657,
      "loss": 0.245,
      "num_input_tokens_seen": 1197856,
      "step": 1385
    },
    {
      "epoch": 0.6553512494106554,
      "grad_norm": 0.012412120588123798,
      "learning_rate": 0.2991083082212644,
      "loss": 0.2131,
      "num_input_tokens_seen": 1203328,
      "step": 1390
    },
    {
      "epoch": 0.6577086280056577,
      "grad_norm": 0.014815357513725758,
      "learning_rate": 0.2991018834249881,
      "loss": 0.3288,
      "num_input_tokens_seen": 1207376,
      "step": 1395
    },
    {
      "epoch": 0.6600660066006601,
      "grad_norm": 0.00789864081889391,
      "learning_rate": 0.29909543563532764,
      "loss": 0.2272,
      "num_input_tokens_seen": 1211680,
      "step": 1400
    },
    {
      "epoch": 0.6600660066006601,
      "eval_loss": 0.3682122528553009,
      "eval_runtime": 21.9505,
      "eval_samples_per_second": 42.96,
      "eval_steps_per_second": 21.503,
      "num_input_tokens_seen": 1211680,
      "step": 1400
    },
    {
      "epoch": 0.6624233851956625,
      "grad_norm": 0.007755184546113014,
      "learning_rate": 0.29908896485327746,
      "loss": 0.2966,
      "num_input_tokens_seen": 1215600,
      "step": 1405
    },
    {
      "epoch": 0.6647807637906648,
      "grad_norm": 0.009833133779466152,
      "learning_rate": 0.29908247107983527,
      "loss": 0.258,
      "num_input_tokens_seen": 1220256,
      "step": 1410
    },
    {
      "epoch": 0.6671381423856672,
      "grad_norm": 0.011940604075789452,
      "learning_rate": 0.29907595431600253,
      "loss": 0.2248,
      "num_input_tokens_seen": 1224464,
      "step": 1415
    },
    {
      "epoch": 0.6694955209806694,
      "grad_norm": 0.012441758066415787,
      "learning_rate": 0.29906941456278424,
      "loss": 0.246,
      "num_input_tokens_seen": 1228464,
      "step": 1420
    },
    {
      "epoch": 0.6718528995756718,
      "grad_norm": 0.015507173724472523,
      "learning_rate": 0.2990628518211889,
      "loss": 0.2411,
      "num_input_tokens_seen": 1232640,
      "step": 1425
    },
    {
      "epoch": 0.6742102781706742,
      "grad_norm": 0.0043488843366503716,
      "learning_rate": 0.2990562660922286,
      "loss": 0.2264,
      "num_input_tokens_seen": 1237248,
      "step": 1430
    },
    {
      "epoch": 0.6765676567656765,
      "grad_norm": 0.008267401717603207,
      "learning_rate": 0.2990496573769189,
      "loss": 0.222,
      "num_input_tokens_seen": 1241056,
      "step": 1435
    },
    {
      "epoch": 0.6789250353606789,
      "grad_norm": 0.014699293300509453,
      "learning_rate": 0.29904302567627894,
      "loss": 0.1974,
      "num_input_tokens_seen": 1245296,
      "step": 1440
    },
    {
      "epoch": 0.6812824139556812,
      "grad_norm": 0.027788080275058746,
      "learning_rate": 0.2990363709913314,
      "loss": 0.1928,
      "num_input_tokens_seen": 1249600,
      "step": 1445
    },
    {
      "epoch": 0.6836397925506836,
      "grad_norm": 0.005524135194718838,
      "learning_rate": 0.29902969332310264,
      "loss": 0.2234,
      "num_input_tokens_seen": 1253616,
      "step": 1450
    },
    {
      "epoch": 0.685997171145686,
      "grad_norm": 0.014030787162482738,
      "learning_rate": 0.2990229926726223,
      "loss": 0.2734,
      "num_input_tokens_seen": 1257392,
      "step": 1455
    },
    {
      "epoch": 0.6883545497406883,
      "grad_norm": 0.009144607000052929,
      "learning_rate": 0.29901626904092365,
      "loss": 0.2246,
      "num_input_tokens_seen": 1261776,
      "step": 1460
    },
    {
      "epoch": 0.6907119283356907,
      "grad_norm": 0.016202040016651154,
      "learning_rate": 0.2990095224290438,
      "loss": 0.2583,
      "num_input_tokens_seen": 1265888,
      "step": 1465
    },
    {
      "epoch": 0.693069306930693,
      "grad_norm": 0.005822889506816864,
      "learning_rate": 0.29900275283802297,
      "loss": 0.2408,
      "num_input_tokens_seen": 1269712,
      "step": 1470
    },
    {
      "epoch": 0.6954266855256954,
      "grad_norm": 0.00962250679731369,
      "learning_rate": 0.2989959602689051,
      "loss": 0.2166,
      "num_input_tokens_seen": 1274464,
      "step": 1475
    },
    {
      "epoch": 0.6977840641206978,
      "grad_norm": 0.006010857410728931,
      "learning_rate": 0.2989891447227379,
      "loss": 0.1959,
      "num_input_tokens_seen": 1278944,
      "step": 1480
    },
    {
      "epoch": 0.7001414427157001,
      "grad_norm": 0.016445614397525787,
      "learning_rate": 0.29898230620057215,
      "loss": 0.2415,
      "num_input_tokens_seen": 1282608,
      "step": 1485
    },
    {
      "epoch": 0.7024988213107025,
      "grad_norm": 0.004994721617549658,
      "learning_rate": 0.2989754447034626,
      "loss": 0.2072,
      "num_input_tokens_seen": 1286480,
      "step": 1490
    },
    {
      "epoch": 0.7048561999057048,
      "grad_norm": 0.01699483022093773,
      "learning_rate": 0.2989685602324673,
      "loss": 0.2332,
      "num_input_tokens_seen": 1290560,
      "step": 1495
    },
    {
      "epoch": 0.7072135785007072,
      "grad_norm": 0.004955081734806299,
      "learning_rate": 0.298961652788648,
      "loss": 0.2175,
      "num_input_tokens_seen": 1294240,
      "step": 1500
    },
    {
      "epoch": 0.7095709570957096,
      "grad_norm": 0.004256641026586294,
      "learning_rate": 0.29895472237306986,
      "loss": 0.2333,
      "num_input_tokens_seen": 1298512,
      "step": 1505
    },
    {
      "epoch": 0.7119283356907119,
      "grad_norm": 0.007530726958066225,
      "learning_rate": 0.29894776898680164,
      "loss": 0.2444,
      "num_input_tokens_seen": 1302736,
      "step": 1510
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.012377893552184105,
      "learning_rate": 0.29894079263091566,
      "loss": 0.225,
      "num_input_tokens_seen": 1307376,
      "step": 1515
    },
    {
      "epoch": 0.7166430928807166,
      "grad_norm": 0.019442012533545494,
      "learning_rate": 0.2989337933064877,
      "loss": 0.1936,
      "num_input_tokens_seen": 1311760,
      "step": 1520
    },
    {
      "epoch": 0.719000471475719,
      "grad_norm": 0.008796768262982368,
      "learning_rate": 0.29892677101459725,
      "loss": 0.2063,
      "num_input_tokens_seen": 1315200,
      "step": 1525
    },
    {
      "epoch": 0.7213578500707214,
      "grad_norm": 0.005080565810203552,
      "learning_rate": 0.2989197257563272,
      "loss": 0.2352,
      "num_input_tokens_seen": 1319008,
      "step": 1530
    },
    {
      "epoch": 0.7237152286657237,
      "grad_norm": 0.005311491433531046,
      "learning_rate": 0.2989126575327639,
      "loss": 0.2439,
      "num_input_tokens_seen": 1323696,
      "step": 1535
    },
    {
      "epoch": 0.7260726072607261,
      "grad_norm": 0.0038522803224623203,
      "learning_rate": 0.29890556634499754,
      "loss": 0.2362,
      "num_input_tokens_seen": 1327264,
      "step": 1540
    },
    {
      "epoch": 0.7284299858557284,
      "grad_norm": 0.007134600542485714,
      "learning_rate": 0.2988984521941216,
      "loss": 0.2052,
      "num_input_tokens_seen": 1331200,
      "step": 1545
    },
    {
      "epoch": 0.7307873644507308,
      "grad_norm": 0.1557055562734604,
      "learning_rate": 0.29889131508123307,
      "loss": 0.3011,
      "num_input_tokens_seen": 1336144,
      "step": 1550
    },
    {
      "epoch": 0.7331447430457332,
      "grad_norm": 0.010834681801497936,
      "learning_rate": 0.2988841550074327,
      "loss": 0.24,
      "num_input_tokens_seen": 1340688,
      "step": 1555
    },
    {
      "epoch": 0.7355021216407355,
      "grad_norm": 0.017351586371660233,
      "learning_rate": 0.2988769719738246,
      "loss": 0.2442,
      "num_input_tokens_seen": 1344960,
      "step": 1560
    },
    {
      "epoch": 0.7378595002357379,
      "grad_norm": 0.008527003228664398,
      "learning_rate": 0.29886976598151666,
      "loss": 0.2186,
      "num_input_tokens_seen": 1349968,
      "step": 1565
    },
    {
      "epoch": 0.7402168788307402,
      "grad_norm": 0.014329357072710991,
      "learning_rate": 0.29886253703161986,
      "loss": 0.3611,
      "num_input_tokens_seen": 1354240,
      "step": 1570
    },
    {
      "epoch": 0.7425742574257426,
      "grad_norm": 0.019205156713724136,
      "learning_rate": 0.29885528512524917,
      "loss": 0.2181,
      "num_input_tokens_seen": 1359232,
      "step": 1575
    },
    {
      "epoch": 0.744931636020745,
      "grad_norm": 0.007677373010665178,
      "learning_rate": 0.29884801026352287,
      "loss": 0.1565,
      "num_input_tokens_seen": 1363408,
      "step": 1580
    },
    {
      "epoch": 0.7472890146157473,
      "grad_norm": 0.04532676562666893,
      "learning_rate": 0.2988407124475629,
      "loss": 0.3095,
      "num_input_tokens_seen": 1368176,
      "step": 1585
    },
    {
      "epoch": 0.7496463932107497,
      "grad_norm": 0.010856139473617077,
      "learning_rate": 0.2988333916784945,
      "loss": 0.2164,
      "num_input_tokens_seen": 1372752,
      "step": 1590
    },
    {
      "epoch": 0.752003771805752,
      "grad_norm": 0.02400929108262062,
      "learning_rate": 0.2988260479574468,
      "loss": 0.3062,
      "num_input_tokens_seen": 1376640,
      "step": 1595
    },
    {
      "epoch": 0.7543611504007544,
      "grad_norm": 0.01615690067410469,
      "learning_rate": 0.2988186812855523,
      "loss": 0.2348,
      "num_input_tokens_seen": 1381792,
      "step": 1600
    },
    {
      "epoch": 0.7543611504007544,
      "eval_loss": 0.3155163824558258,
      "eval_runtime": 21.8857,
      "eval_samples_per_second": 43.087,
      "eval_steps_per_second": 21.567,
      "num_input_tokens_seen": 1381792,
      "step": 1600
    },
    {
      "epoch": 0.7567185289957568,
      "grad_norm": 0.007281403988599777,
      "learning_rate": 0.29881129166394693,
      "loss": 0.2299,
      "num_input_tokens_seen": 1385968,
      "step": 1605
    },
    {
      "epoch": 0.759075907590759,
      "grad_norm": 0.005476027727127075,
      "learning_rate": 0.29880387909377026,
      "loss": 0.2251,
      "num_input_tokens_seen": 1390256,
      "step": 1610
    },
    {
      "epoch": 0.7614332861857614,
      "grad_norm": 0.005885435733944178,
      "learning_rate": 0.2987964435761655,
      "loss": 0.2186,
      "num_input_tokens_seen": 1394272,
      "step": 1615
    },
    {
      "epoch": 0.7637906647807637,
      "grad_norm": 0.005694640800356865,
      "learning_rate": 0.29878898511227925,
      "loss": 0.236,
      "num_input_tokens_seen": 1398672,
      "step": 1620
    },
    {
      "epoch": 0.7661480433757661,
      "grad_norm": 0.008330917917191982,
      "learning_rate": 0.2987815037032617,
      "loss": 0.1737,
      "num_input_tokens_seen": 1403168,
      "step": 1625
    },
    {
      "epoch": 0.7685054219707685,
      "grad_norm": 0.01563160866498947,
      "learning_rate": 0.29877399935026655,
      "loss": 0.2326,
      "num_input_tokens_seen": 1407792,
      "step": 1630
    },
    {
      "epoch": 0.7708628005657708,
      "grad_norm": 0.009742502123117447,
      "learning_rate": 0.2987664720544511,
      "loss": 0.2272,
      "num_input_tokens_seen": 1412080,
      "step": 1635
    },
    {
      "epoch": 0.7732201791607732,
      "grad_norm": 0.006067799869924784,
      "learning_rate": 0.2987589218169761,
      "loss": 0.228,
      "num_input_tokens_seen": 1416752,
      "step": 1640
    },
    {
      "epoch": 0.7755775577557755,
      "grad_norm": 0.0038220356218516827,
      "learning_rate": 0.29875134863900604,
      "loss": 0.2281,
      "num_input_tokens_seen": 1421728,
      "step": 1645
    },
    {
      "epoch": 0.7779349363507779,
      "grad_norm": 0.013462711125612259,
      "learning_rate": 0.29874375252170865,
      "loss": 0.2093,
      "num_input_tokens_seen": 1425744,
      "step": 1650
    },
    {
      "epoch": 0.7802923149457803,
      "grad_norm": 0.01707606017589569,
      "learning_rate": 0.2987361334662553,
      "loss": 0.2286,
      "num_input_tokens_seen": 1430528,
      "step": 1655
    },
    {
      "epoch": 0.7826496935407826,
      "grad_norm": 0.008576737716794014,
      "learning_rate": 0.29872849147382113,
      "loss": 0.271,
      "num_input_tokens_seen": 1435376,
      "step": 1660
    },
    {
      "epoch": 0.785007072135785,
      "grad_norm": 0.01290939748287201,
      "learning_rate": 0.2987208265455845,
      "loss": 0.2282,
      "num_input_tokens_seen": 1440272,
      "step": 1665
    },
    {
      "epoch": 0.7873644507307873,
      "grad_norm": 0.012482976540923119,
      "learning_rate": 0.29871313868272753,
      "loss": 0.2511,
      "num_input_tokens_seen": 1444384,
      "step": 1670
    },
    {
      "epoch": 0.7897218293257897,
      "grad_norm": 0.004670243710279465,
      "learning_rate": 0.29870542788643567,
      "loss": 0.2415,
      "num_input_tokens_seen": 1449056,
      "step": 1675
    },
    {
      "epoch": 0.7920792079207921,
      "grad_norm": 0.00447361683472991,
      "learning_rate": 0.2986976941578981,
      "loss": 0.2365,
      "num_input_tokens_seen": 1453104,
      "step": 1680
    },
    {
      "epoch": 0.7944365865157944,
      "grad_norm": 0.01498006284236908,
      "learning_rate": 0.29868993749830747,
      "loss": 0.2347,
      "num_input_tokens_seen": 1457520,
      "step": 1685
    },
    {
      "epoch": 0.7967939651107968,
      "grad_norm": 0.004394884686917067,
      "learning_rate": 0.2986821579088598,
      "loss": 0.2238,
      "num_input_tokens_seen": 1461984,
      "step": 1690
    },
    {
      "epoch": 0.7991513437057991,
      "grad_norm": 0.009351997636258602,
      "learning_rate": 0.29867435539075504,
      "loss": 0.2366,
      "num_input_tokens_seen": 1465952,
      "step": 1695
    },
    {
      "epoch": 0.8015087223008015,
      "grad_norm": 0.004777699243277311,
      "learning_rate": 0.2986665299451963,
      "loss": 0.2232,
      "num_input_tokens_seen": 1470480,
      "step": 1700
    },
    {
      "epoch": 0.8038661008958039,
      "grad_norm": 0.005686374381184578,
      "learning_rate": 0.29865868157339037,
      "loss": 0.2233,
      "num_input_tokens_seen": 1474736,
      "step": 1705
    },
    {
      "epoch": 0.8062234794908062,
      "grad_norm": 0.006763003300875425,
      "learning_rate": 0.2986508102765476,
      "loss": 0.2117,
      "num_input_tokens_seen": 1478944,
      "step": 1710
    },
    {
      "epoch": 0.8085808580858086,
      "grad_norm": 0.027879104018211365,
      "learning_rate": 0.2986429160558818,
      "loss": 0.2072,
      "num_input_tokens_seen": 1484288,
      "step": 1715
    },
    {
      "epoch": 0.8109382366808109,
      "grad_norm": 0.0064238267950713634,
      "learning_rate": 0.2986349989126104,
      "loss": 0.3049,
      "num_input_tokens_seen": 1489008,
      "step": 1720
    },
    {
      "epoch": 0.8132956152758133,
      "grad_norm": 0.016541101038455963,
      "learning_rate": 0.29862705884795426,
      "loss": 0.2446,
      "num_input_tokens_seen": 1494672,
      "step": 1725
    },
    {
      "epoch": 0.8156529938708157,
      "grad_norm": 0.012036236003041267,
      "learning_rate": 0.2986190958631379,
      "loss": 0.2432,
      "num_input_tokens_seen": 1499072,
      "step": 1730
    },
    {
      "epoch": 0.818010372465818,
      "grad_norm": 0.01372484304010868,
      "learning_rate": 0.29861110995938933,
      "loss": 0.2346,
      "num_input_tokens_seen": 1502816,
      "step": 1735
    },
    {
      "epoch": 0.8203677510608204,
      "grad_norm": 0.0069134049117565155,
      "learning_rate": 0.29860310113794,
      "loss": 0.2642,
      "num_input_tokens_seen": 1506736,
      "step": 1740
    },
    {
      "epoch": 0.8227251296558227,
      "grad_norm": 0.01307134609669447,
      "learning_rate": 0.29859506940002506,
      "loss": 0.2348,
      "num_input_tokens_seen": 1510992,
      "step": 1745
    },
    {
      "epoch": 0.8250825082508251,
      "grad_norm": 0.005410636775195599,
      "learning_rate": 0.298587014746883,
      "loss": 0.2195,
      "num_input_tokens_seen": 1515056,
      "step": 1750
    },
    {
      "epoch": 0.8274398868458275,
      "grad_norm": 0.01701577752828598,
      "learning_rate": 0.298578937179756,
      "loss": 0.2452,
      "num_input_tokens_seen": 1520048,
      "step": 1755
    },
    {
      "epoch": 0.8297972654408298,
      "grad_norm": 0.013892080634832382,
      "learning_rate": 0.29857083669988976,
      "loss": 0.2283,
      "num_input_tokens_seen": 1524176,
      "step": 1760
    },
    {
      "epoch": 0.8321546440358322,
      "grad_norm": 0.015122687444090843,
      "learning_rate": 0.29856271330853346,
      "loss": 0.2267,
      "num_input_tokens_seen": 1528784,
      "step": 1765
    },
    {
      "epoch": 0.8345120226308345,
      "grad_norm": 0.008332401514053345,
      "learning_rate": 0.2985545670069398,
      "loss": 0.2393,
      "num_input_tokens_seen": 1533328,
      "step": 1770
    },
    {
      "epoch": 0.8368694012258369,
      "grad_norm": 0.011506418697535992,
      "learning_rate": 0.29854639779636505,
      "loss": 0.2347,
      "num_input_tokens_seen": 1538064,
      "step": 1775
    },
    {
      "epoch": 0.8392267798208393,
      "grad_norm": 0.004224782343953848,
      "learning_rate": 0.298538205678069,
      "loss": 0.2159,
      "num_input_tokens_seen": 1541952,
      "step": 1780
    },
    {
      "epoch": 0.8415841584158416,
      "grad_norm": 0.005852525122463703,
      "learning_rate": 0.298529990653315,
      "loss": 0.2365,
      "num_input_tokens_seen": 1546032,
      "step": 1785
    },
    {
      "epoch": 0.843941537010844,
      "grad_norm": 0.003291608765721321,
      "learning_rate": 0.29852175272336984,
      "loss": 0.2267,
      "num_input_tokens_seen": 1550480,
      "step": 1790
    },
    {
      "epoch": 0.8462989156058462,
      "grad_norm": 0.01006405521184206,
      "learning_rate": 0.29851349188950405,
      "loss": 0.2215,
      "num_input_tokens_seen": 1554848,
      "step": 1795
    },
    {
      "epoch": 0.8486562942008486,
      "grad_norm": 0.0038961637765169144,
      "learning_rate": 0.2985052081529914,
      "loss": 0.2577,
      "num_input_tokens_seen": 1559456,
      "step": 1800
    },
    {
      "epoch": 0.8486562942008486,
      "eval_loss": 0.22190262377262115,
      "eval_runtime": 21.9289,
      "eval_samples_per_second": 43.003,
      "eval_steps_per_second": 21.524,
      "num_input_tokens_seen": 1559456,
      "step": 1800
    },
    {
      "epoch": 0.851013672795851,
      "grad_norm": 0.008501654490828514,
      "learning_rate": 0.29849690151510944,
      "loss": 0.2235,
      "num_input_tokens_seen": 1563360,
      "step": 1805
    },
    {
      "epoch": 0.8533710513908533,
      "grad_norm": 0.003951655235141516,
      "learning_rate": 0.2984885719771392,
      "loss": 0.192,
      "num_input_tokens_seen": 1567232,
      "step": 1810
    },
    {
      "epoch": 0.8557284299858557,
      "grad_norm": 0.005035221576690674,
      "learning_rate": 0.2984802195403651,
      "loss": 0.2581,
      "num_input_tokens_seen": 1571392,
      "step": 1815
    },
    {
      "epoch": 0.858085808580858,
      "grad_norm": 0.005520319566130638,
      "learning_rate": 0.2984718442060752,
      "loss": 0.2337,
      "num_input_tokens_seen": 1575600,
      "step": 1820
    },
    {
      "epoch": 0.8604431871758604,
      "grad_norm": 0.15033912658691406,
      "learning_rate": 0.2984634459755611,
      "loss": 0.2505,
      "num_input_tokens_seen": 1580640,
      "step": 1825
    },
    {
      "epoch": 0.8628005657708628,
      "grad_norm": 0.012041726149618626,
      "learning_rate": 0.29845502485011793,
      "loss": 0.2359,
      "num_input_tokens_seen": 1584656,
      "step": 1830
    },
    {
      "epoch": 0.8651579443658651,
      "grad_norm": 0.009560885839164257,
      "learning_rate": 0.2984465808310444,
      "loss": 0.2254,
      "num_input_tokens_seen": 1588976,
      "step": 1835
    },
    {
      "epoch": 0.8675153229608675,
      "grad_norm": 0.0041495999321341515,
      "learning_rate": 0.29843811391964253,
      "loss": 0.1744,
      "num_input_tokens_seen": 1592928,
      "step": 1840
    },
    {
      "epoch": 0.8698727015558698,
      "grad_norm": 0.004277557134628296,
      "learning_rate": 0.2984296241172182,
      "loss": 0.2359,
      "num_input_tokens_seen": 1596880,
      "step": 1845
    },
    {
      "epoch": 0.8722300801508722,
      "grad_norm": 0.00727043254300952,
      "learning_rate": 0.29842111142508043,
      "loss": 0.1977,
      "num_input_tokens_seen": 1601440,
      "step": 1850
    },
    {
      "epoch": 0.8745874587458746,
      "grad_norm": 0.005497789941728115,
      "learning_rate": 0.29841257584454217,
      "loss": 0.2538,
      "num_input_tokens_seen": 1606032,
      "step": 1855
    },
    {
      "epoch": 0.8769448373408769,
      "grad_norm": 0.009605045430362225,
      "learning_rate": 0.29840401737691963,
      "loss": 0.2496,
      "num_input_tokens_seen": 1610720,
      "step": 1860
    },
    {
      "epoch": 0.8793022159358793,
      "grad_norm": 0.01595224067568779,
      "learning_rate": 0.29839543602353263,
      "loss": 0.2506,
      "num_input_tokens_seen": 1615520,
      "step": 1865
    },
    {
      "epoch": 0.8816595945308816,
      "grad_norm": 0.005318209063261747,
      "learning_rate": 0.2983868317857046,
      "loss": 0.2333,
      "num_input_tokens_seen": 1620416,
      "step": 1870
    },
    {
      "epoch": 0.884016973125884,
      "grad_norm": 0.006282396614551544,
      "learning_rate": 0.2983782046647623,
      "loss": 0.2395,
      "num_input_tokens_seen": 1624368,
      "step": 1875
    },
    {
      "epoch": 0.8863743517208864,
      "grad_norm": 0.010913478210568428,
      "learning_rate": 0.2983695546620362,
      "loss": 0.2408,
      "num_input_tokens_seen": 1628656,
      "step": 1880
    },
    {
      "epoch": 0.8887317303158887,
      "grad_norm": 0.005084115080535412,
      "learning_rate": 0.2983608817788603,
      "loss": 0.2264,
      "num_input_tokens_seen": 1633072,
      "step": 1885
    },
    {
      "epoch": 0.8910891089108911,
      "grad_norm": 0.011594077572226524,
      "learning_rate": 0.29835218601657193,
      "loss": 0.234,
      "num_input_tokens_seen": 1637328,
      "step": 1890
    },
    {
      "epoch": 0.8934464875058934,
      "grad_norm": 0.007585883606225252,
      "learning_rate": 0.29834346737651224,
      "loss": 0.2136,
      "num_input_tokens_seen": 1642304,
      "step": 1895
    },
    {
      "epoch": 0.8958038661008958,
      "grad_norm": 0.008697092533111572,
      "learning_rate": 0.29833472586002563,
      "loss": 0.2273,
      "num_input_tokens_seen": 1646656,
      "step": 1900
    },
    {
      "epoch": 0.8981612446958982,
      "grad_norm": 0.003734774421900511,
      "learning_rate": 0.29832596146846024,
      "loss": 0.2073,
      "num_input_tokens_seen": 1650720,
      "step": 1905
    },
    {
      "epoch": 0.9005186232909005,
      "grad_norm": 0.003477261168882251,
      "learning_rate": 0.2983171742031676,
      "loss": 0.2737,
      "num_input_tokens_seen": 1655488,
      "step": 1910
    },
    {
      "epoch": 0.9028760018859029,
      "grad_norm": 0.005279069300740957,
      "learning_rate": 0.2983083640655028,
      "loss": 0.2003,
      "num_input_tokens_seen": 1660016,
      "step": 1915
    },
    {
      "epoch": 0.9052333804809052,
      "grad_norm": 0.004082551691681147,
      "learning_rate": 0.29829953105682455,
      "loss": 0.2306,
      "num_input_tokens_seen": 1664448,
      "step": 1920
    },
    {
      "epoch": 0.9075907590759076,
      "grad_norm": 0.016790742054581642,
      "learning_rate": 0.29829067517849495,
      "loss": 0.2208,
      "num_input_tokens_seen": 1669312,
      "step": 1925
    },
    {
      "epoch": 0.90994813767091,
      "grad_norm": 0.0040323324501514435,
      "learning_rate": 0.2982817964318797,
      "loss": 0.2253,
      "num_input_tokens_seen": 1673424,
      "step": 1930
    },
    {
      "epoch": 0.9123055162659123,
      "grad_norm": 0.004748993087559938,
      "learning_rate": 0.298272894818348,
      "loss": 0.2231,
      "num_input_tokens_seen": 1677760,
      "step": 1935
    },
    {
      "epoch": 0.9146628948609147,
      "grad_norm": 0.004319407511502504,
      "learning_rate": 0.2982639703392726,
      "loss": 0.2162,
      "num_input_tokens_seen": 1682064,
      "step": 1940
    },
    {
      "epoch": 0.917020273455917,
      "grad_norm": 0.004806531593203545,
      "learning_rate": 0.29825502299602974,
      "loss": 0.2543,
      "num_input_tokens_seen": 1686752,
      "step": 1945
    },
    {
      "epoch": 0.9193776520509194,
      "grad_norm": 0.0034314647782593966,
      "learning_rate": 0.2982460527899993,
      "loss": 0.1798,
      "num_input_tokens_seen": 1691232,
      "step": 1950
    },
    {
      "epoch": 0.9217350306459218,
      "grad_norm": 0.004376592114567757,
      "learning_rate": 0.29823705972256453,
      "loss": 0.3034,
      "num_input_tokens_seen": 1695440,
      "step": 1955
    },
    {
      "epoch": 0.9240924092409241,
      "grad_norm": 0.0073643955402076244,
      "learning_rate": 0.2982280437951123,
      "loss": 0.2547,
      "num_input_tokens_seen": 1699344,
      "step": 1960
    },
    {
      "epoch": 0.9264497878359265,
      "grad_norm": 0.011101096868515015,
      "learning_rate": 0.298219005009033,
      "loss": 0.2375,
      "num_input_tokens_seen": 1704560,
      "step": 1965
    },
    {
      "epoch": 0.9288071664309288,
      "grad_norm": 0.007933859713375568,
      "learning_rate": 0.29820994336572043,
      "loss": 0.2094,
      "num_input_tokens_seen": 1709312,
      "step": 1970
    },
    {
      "epoch": 0.9311645450259312,
      "grad_norm": 0.007413928396999836,
      "learning_rate": 0.2982008588665721,
      "loss": 0.2066,
      "num_input_tokens_seen": 1714064,
      "step": 1975
    },
    {
      "epoch": 0.9335219236209336,
      "grad_norm": 0.005915017798542976,
      "learning_rate": 0.2981917515129889,
      "loss": 0.2916,
      "num_input_tokens_seen": 1718688,
      "step": 1980
    },
    {
      "epoch": 0.9358793022159358,
      "grad_norm": 0.0036994311958551407,
      "learning_rate": 0.2981826213063753,
      "loss": 0.2458,
      "num_input_tokens_seen": 1723744,
      "step": 1985
    },
    {
      "epoch": 0.9382366808109383,
      "grad_norm": 0.0039099776186048985,
      "learning_rate": 0.2981734682481394,
      "loss": 0.2364,
      "num_input_tokens_seen": 1727824,
      "step": 1990
    },
    {
      "epoch": 0.9405940594059405,
      "grad_norm": 0.0028941142372787,
      "learning_rate": 0.29816429233969255,
      "loss": 0.2007,
      "num_input_tokens_seen": 1731728,
      "step": 1995
    },
    {
      "epoch": 0.9429514380009429,
      "grad_norm": 0.01276901550590992,
      "learning_rate": 0.2981550935824499,
      "loss": 0.3163,
      "num_input_tokens_seen": 1735840,
      "step": 2000
    },
    {
      "epoch": 0.9429514380009429,
      "eval_loss": 0.22208459675312042,
      "eval_runtime": 21.9025,
      "eval_samples_per_second": 43.054,
      "eval_steps_per_second": 21.55,
      "num_input_tokens_seen": 1735840,
      "step": 2000
    },
    {
      "epoch": 0.9453088165959453,
      "grad_norm": 0.002622877014800906,
      "learning_rate": 0.29814587197783,
      "loss": 0.2442,
      "num_input_tokens_seen": 1739856,
      "step": 2005
    },
    {
      "epoch": 0.9476661951909476,
      "grad_norm": 0.005940092261880636,
      "learning_rate": 0.29813662752725495,
      "loss": 0.245,
      "num_input_tokens_seen": 1745168,
      "step": 2010
    },
    {
      "epoch": 0.95002357378595,
      "grad_norm": 0.005352016072720289,
      "learning_rate": 0.29812736023215025,
      "loss": 0.2275,
      "num_input_tokens_seen": 1749936,
      "step": 2015
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.006731404457241297,
      "learning_rate": 0.29811807009394514,
      "loss": 0.2488,
      "num_input_tokens_seen": 1754432,
      "step": 2020
    },
    {
      "epoch": 0.9547383309759547,
      "grad_norm": 0.004008446354418993,
      "learning_rate": 0.2981087571140723,
      "loss": 0.2155,
      "num_input_tokens_seen": 1758960,
      "step": 2025
    },
    {
      "epoch": 0.9570957095709571,
      "grad_norm": 0.0027782730758190155,
      "learning_rate": 0.2980994212939678,
      "loss": 0.1574,
      "num_input_tokens_seen": 1762704,
      "step": 2030
    },
    {
      "epoch": 0.9594530881659594,
      "grad_norm": 0.0029837798792868853,
      "learning_rate": 0.2980900626350715,
      "loss": 0.2808,
      "num_input_tokens_seen": 1767632,
      "step": 2035
    },
    {
      "epoch": 0.9618104667609618,
      "grad_norm": 0.0028322862926870584,
      "learning_rate": 0.29808068113882646,
      "loss": 0.2378,
      "num_input_tokens_seen": 1772032,
      "step": 2040
    },
    {
      "epoch": 0.9641678453559641,
      "grad_norm": 0.004754126071929932,
      "learning_rate": 0.2980712768066795,
      "loss": 0.2333,
      "num_input_tokens_seen": 1776112,
      "step": 2045
    },
    {
      "epoch": 0.9665252239509665,
      "grad_norm": 0.005631719250231981,
      "learning_rate": 0.2980618496400809,
      "loss": 0.234,
      "num_input_tokens_seen": 1780480,
      "step": 2050
    },
    {
      "epoch": 0.9688826025459689,
      "grad_norm": 0.004796988796442747,
      "learning_rate": 0.2980523996404844,
      "loss": 0.2292,
      "num_input_tokens_seen": 1784672,
      "step": 2055
    },
    {
      "epoch": 0.9712399811409712,
      "grad_norm": 0.007834039628505707,
      "learning_rate": 0.2980429268093473,
      "loss": 0.2086,
      "num_input_tokens_seen": 1789056,
      "step": 2060
    },
    {
      "epoch": 0.9735973597359736,
      "grad_norm": 0.010467850603163242,
      "learning_rate": 0.29803343114813047,
      "loss": 0.2401,
      "num_input_tokens_seen": 1793776,
      "step": 2065
    },
    {
      "epoch": 0.9759547383309759,
      "grad_norm": 0.004168745595961809,
      "learning_rate": 0.2980239126582983,
      "loss": 0.2563,
      "num_input_tokens_seen": 1798928,
      "step": 2070
    },
    {
      "epoch": 0.9783121169259783,
      "grad_norm": 0.011788712814450264,
      "learning_rate": 0.2980143713413186,
      "loss": 0.2368,
      "num_input_tokens_seen": 1803120,
      "step": 2075
    },
    {
      "epoch": 0.9806694955209807,
      "grad_norm": 0.005848876666277647,
      "learning_rate": 0.29800480719866274,
      "loss": 0.2363,
      "num_input_tokens_seen": 1807664,
      "step": 2080
    },
    {
      "epoch": 0.983026874115983,
      "grad_norm": 0.007544418331235647,
      "learning_rate": 0.2979952202318057,
      "loss": 0.2454,
      "num_input_tokens_seen": 1811344,
      "step": 2085
    },
    {
      "epoch": 0.9853842527109854,
      "grad_norm": 0.003753883996978402,
      "learning_rate": 0.2979856104422259,
      "loss": 0.2234,
      "num_input_tokens_seen": 1815472,
      "step": 2090
    },
    {
      "epoch": 0.9877416313059877,
      "grad_norm": 0.002804399933665991,
      "learning_rate": 0.2979759778314052,
      "loss": 0.2049,
      "num_input_tokens_seen": 1819744,
      "step": 2095
    },
    {
      "epoch": 0.9900990099009901,
      "grad_norm": 0.006288627162575722,
      "learning_rate": 0.2979663224008292,
      "loss": 0.2779,
      "num_input_tokens_seen": 1823792,
      "step": 2100
    },
    {
      "epoch": 0.9924563884959925,
      "grad_norm": 0.007085269317030907,
      "learning_rate": 0.2979566441519868,
      "loss": 0.2267,
      "num_input_tokens_seen": 1827632,
      "step": 2105
    },
    {
      "epoch": 0.9948137670909948,
      "grad_norm": 0.004938343074172735,
      "learning_rate": 0.29794694308637054,
      "loss": 0.2501,
      "num_input_tokens_seen": 1831376,
      "step": 2110
    },
    {
      "epoch": 0.9971711456859972,
      "grad_norm": 0.002113934373483062,
      "learning_rate": 0.2979372192054764,
      "loss": 0.2306,
      "num_input_tokens_seen": 1835840,
      "step": 2115
    },
    {
      "epoch": 0.9995285242809995,
      "grad_norm": 0.008228533901274204,
      "learning_rate": 0.297927472510804,
      "loss": 0.2152,
      "num_input_tokens_seen": 1840144,
      "step": 2120
    },
    {
      "epoch": 1.0018859028760019,
      "grad_norm": 0.0043676975183188915,
      "learning_rate": 0.29791770300385634,
      "loss": 0.2713,
      "num_input_tokens_seen": 1844720,
      "step": 2125
    },
    {
      "epoch": 1.0042432814710043,
      "grad_norm": 0.007573768496513367,
      "learning_rate": 0.29790791068614003,
      "loss": 0.2381,
      "num_input_tokens_seen": 1848592,
      "step": 2130
    },
    {
      "epoch": 1.0066006600660067,
      "grad_norm": 0.005674374755471945,
      "learning_rate": 0.2978980955591652,
      "loss": 0.2286,
      "num_input_tokens_seen": 1853344,
      "step": 2135
    },
    {
      "epoch": 1.0089580386610089,
      "grad_norm": 0.0025785774923861027,
      "learning_rate": 0.2978882576244454,
      "loss": 0.1867,
      "num_input_tokens_seen": 1858352,
      "step": 2140
    },
    {
      "epoch": 1.0113154172560113,
      "grad_norm": 0.0030189198441803455,
      "learning_rate": 0.2978783968834978,
      "loss": 0.2361,
      "num_input_tokens_seen": 1863120,
      "step": 2145
    },
    {
      "epoch": 1.0136727958510137,
      "grad_norm": 0.004002026282250881,
      "learning_rate": 0.29786851333784303,
      "loss": 0.1929,
      "num_input_tokens_seen": 1868640,
      "step": 2150
    },
    {
      "epoch": 1.016030174446016,
      "grad_norm": 0.004960834514349699,
      "learning_rate": 0.2978586069890053,
      "loss": 0.205,
      "num_input_tokens_seen": 1873088,
      "step": 2155
    },
    {
      "epoch": 1.0183875530410185,
      "grad_norm": 0.0032031885348260403,
      "learning_rate": 0.29784867783851227,
      "loss": 0.2988,
      "num_input_tokens_seen": 1877552,
      "step": 2160
    },
    {
      "epoch": 1.0207449316360206,
      "grad_norm": 0.005285789258778095,
      "learning_rate": 0.2978387258878951,
      "loss": 0.2183,
      "num_input_tokens_seen": 1882240,
      "step": 2165
    },
    {
      "epoch": 1.023102310231023,
      "grad_norm": 0.006060976069420576,
      "learning_rate": 0.29782875113868856,
      "loss": 0.2511,
      "num_input_tokens_seen": 1887120,
      "step": 2170
    },
    {
      "epoch": 1.0254596888260255,
      "grad_norm": 0.005063164047896862,
      "learning_rate": 0.2978187535924309,
      "loss": 0.2577,
      "num_input_tokens_seen": 1891312,
      "step": 2175
    },
    {
      "epoch": 1.0278170674210279,
      "grad_norm": 0.006553592626005411,
      "learning_rate": 0.29780873325066376,
      "loss": 0.2084,
      "num_input_tokens_seen": 1895360,
      "step": 2180
    },
    {
      "epoch": 1.0301744460160303,
      "grad_norm": 0.019467391073703766,
      "learning_rate": 0.2977986901149325,
      "loss": 0.2477,
      "num_input_tokens_seen": 1898704,
      "step": 2185
    },
    {
      "epoch": 1.0325318246110324,
      "grad_norm": 0.0025658246595412493,
      "learning_rate": 0.29778862418678587,
      "loss": 0.2296,
      "num_input_tokens_seen": 1903328,
      "step": 2190
    },
    {
      "epoch": 1.0348892032060348,
      "grad_norm": 0.003836560994386673,
      "learning_rate": 0.29777853546777616,
      "loss": 0.2438,
      "num_input_tokens_seen": 1907024,
      "step": 2195
    },
    {
      "epoch": 1.0372465818010372,
      "grad_norm": 0.002883586334064603,
      "learning_rate": 0.2977684239594592,
      "loss": 0.2406,
      "num_input_tokens_seen": 1910848,
      "step": 2200
    },
    {
      "epoch": 1.0372465818010372,
      "eval_loss": 0.24526606500148773,
      "eval_runtime": 21.9613,
      "eval_samples_per_second": 42.939,
      "eval_steps_per_second": 21.492,
      "num_input_tokens_seen": 1910848,
      "step": 2200
    },
    {
      "epoch": 1.0396039603960396,
      "grad_norm": 0.004581039305776358,
      "learning_rate": 0.29775828966339424,
      "loss": 0.2259,
      "num_input_tokens_seen": 1914672,
      "step": 2205
    },
    {
      "epoch": 1.041961338991042,
      "grad_norm": 0.007325735408812761,
      "learning_rate": 0.29774813258114424,
      "loss": 0.2587,
      "num_input_tokens_seen": 1918880,
      "step": 2210
    },
    {
      "epoch": 1.0443187175860442,
      "grad_norm": 0.004092403221875429,
      "learning_rate": 0.29773795271427544,
      "loss": 0.2302,
      "num_input_tokens_seen": 1923024,
      "step": 2215
    },
    {
      "epoch": 1.0466760961810466,
      "grad_norm": 0.002851843135431409,
      "learning_rate": 0.2977277500643577,
      "loss": 0.2074,
      "num_input_tokens_seen": 1927360,
      "step": 2220
    },
    {
      "epoch": 1.049033474776049,
      "grad_norm": 0.0043287393637001514,
      "learning_rate": 0.29771752463296447,
      "loss": 0.2367,
      "num_input_tokens_seen": 1931776,
      "step": 2225
    },
    {
      "epoch": 1.0513908533710514,
      "grad_norm": 0.003814509604126215,
      "learning_rate": 0.29770727642167266,
      "loss": 0.245,
      "num_input_tokens_seen": 1935376,
      "step": 2230
    },
    {
      "epoch": 1.0537482319660538,
      "grad_norm": 0.0050248634070158005,
      "learning_rate": 0.29769700543206257,
      "loss": 0.2181,
      "num_input_tokens_seen": 1939552,
      "step": 2235
    },
    {
      "epoch": 1.056105610561056,
      "grad_norm": 0.002869442105293274,
      "learning_rate": 0.2976867116657182,
      "loss": 0.2103,
      "num_input_tokens_seen": 1944192,
      "step": 2240
    },
    {
      "epoch": 1.0584629891560584,
      "grad_norm": 0.0022990142460912466,
      "learning_rate": 0.2976763951242269,
      "loss": 0.1913,
      "num_input_tokens_seen": 1948656,
      "step": 2245
    },
    {
      "epoch": 1.0608203677510608,
      "grad_norm": 0.013879192061722279,
      "learning_rate": 0.29766605580917965,
      "loss": 0.195,
      "num_input_tokens_seen": 1953344,
      "step": 2250
    },
    {
      "epoch": 1.0631777463460632,
      "grad_norm": 0.00637099938467145,
      "learning_rate": 0.29765569372217093,
      "loss": 0.2582,
      "num_input_tokens_seen": 1957392,
      "step": 2255
    },
    {
      "epoch": 1.0655351249410656,
      "grad_norm": 0.00394192710518837,
      "learning_rate": 0.2976453088647987,
      "loss": 0.1951,
      "num_input_tokens_seen": 1961440,
      "step": 2260
    },
    {
      "epoch": 1.0678925035360678,
      "grad_norm": 0.003272519912570715,
      "learning_rate": 0.2976349012386644,
      "loss": 0.1994,
      "num_input_tokens_seen": 1965504,
      "step": 2265
    },
    {
      "epoch": 1.0702498821310702,
      "grad_norm": 0.0030303613748401403,
      "learning_rate": 0.29762447084537297,
      "loss": 0.2081,
      "num_input_tokens_seen": 1969152,
      "step": 2270
    },
    {
      "epoch": 1.0726072607260726,
      "grad_norm": 0.0029132701456546783,
      "learning_rate": 0.29761401768653306,
      "loss": 0.2274,
      "num_input_tokens_seen": 1973952,
      "step": 2275
    },
    {
      "epoch": 1.074964639321075,
      "grad_norm": 0.0034993591252714396,
      "learning_rate": 0.29760354176375653,
      "loss": 0.256,
      "num_input_tokens_seen": 1977520,
      "step": 2280
    },
    {
      "epoch": 1.0773220179160774,
      "grad_norm": 0.007542883511632681,
      "learning_rate": 0.29759304307865897,
      "loss": 0.2172,
      "num_input_tokens_seen": 1981664,
      "step": 2285
    },
    {
      "epoch": 1.0796793965110796,
      "grad_norm": 0.004839949309825897,
      "learning_rate": 0.2975825216328594,
      "loss": 0.2284,
      "num_input_tokens_seen": 1986304,
      "step": 2290
    },
    {
      "epoch": 1.082036775106082,
      "grad_norm": 0.003058667993173003,
      "learning_rate": 0.2975719774279804,
      "loss": 0.2233,
      "num_input_tokens_seen": 1990400,
      "step": 2295
    },
    {
      "epoch": 1.0843941537010844,
      "grad_norm": 0.0036426065489649773,
      "learning_rate": 0.29756141046564794,
      "loss": 0.2132,
      "num_input_tokens_seen": 1994368,
      "step": 2300
    },
    {
      "epoch": 1.0867515322960868,
      "grad_norm": 0.0032126728910952806,
      "learning_rate": 0.2975508207474916,
      "loss": 0.245,
      "num_input_tokens_seen": 1999104,
      "step": 2305
    },
    {
      "epoch": 1.0891089108910892,
      "grad_norm": 0.0036312832962721586,
      "learning_rate": 0.2975402082751445,
      "loss": 0.2277,
      "num_input_tokens_seen": 2003248,
      "step": 2310
    },
    {
      "epoch": 1.0914662894860914,
      "grad_norm": 0.00797110516577959,
      "learning_rate": 0.29752957305024313,
      "loss": 0.2149,
      "num_input_tokens_seen": 2007040,
      "step": 2315
    },
    {
      "epoch": 1.0938236680810938,
      "grad_norm": 0.007661684416234493,
      "learning_rate": 0.2975189150744277,
      "loss": 0.2225,
      "num_input_tokens_seen": 2010720,
      "step": 2320
    },
    {
      "epoch": 1.0961810466760962,
      "grad_norm": 0.004179884679615498,
      "learning_rate": 0.29750823434934165,
      "loss": 0.2142,
      "num_input_tokens_seen": 2014752,
      "step": 2325
    },
    {
      "epoch": 1.0985384252710986,
      "grad_norm": 0.004948524758219719,
      "learning_rate": 0.29749753087663217,
      "loss": 0.233,
      "num_input_tokens_seen": 2019632,
      "step": 2330
    },
    {
      "epoch": 1.100895803866101,
      "grad_norm": 0.00805339403450489,
      "learning_rate": 0.29748680465794985,
      "loss": 0.2286,
      "num_input_tokens_seen": 2023712,
      "step": 2335
    },
    {
      "epoch": 1.1032531824611032,
      "grad_norm": 0.003974564839154482,
      "learning_rate": 0.29747605569494884,
      "loss": 0.2345,
      "num_input_tokens_seen": 2027680,
      "step": 2340
    },
    {
      "epoch": 1.1056105610561056,
      "grad_norm": 0.007854514755308628,
      "learning_rate": 0.29746528398928673,
      "loss": 0.241,
      "num_input_tokens_seen": 2032704,
      "step": 2345
    },
    {
      "epoch": 1.107967939651108,
      "grad_norm": 0.007960119284689426,
      "learning_rate": 0.2974544895426247,
      "loss": 0.2517,
      "num_input_tokens_seen": 2036160,
      "step": 2350
    },
    {
      "epoch": 1.1103253182461104,
      "grad_norm": 0.008488879539072514,
      "learning_rate": 0.29744367235662733,
      "loss": 0.239,
      "num_input_tokens_seen": 2040240,
      "step": 2355
    },
    {
      "epoch": 1.1126826968411128,
      "grad_norm": 0.020646855235099792,
      "learning_rate": 0.29743283243296276,
      "loss": 0.2545,
      "num_input_tokens_seen": 2046224,
      "step": 2360
    },
    {
      "epoch": 1.115040075436115,
      "grad_norm": 0.005945291370153427,
      "learning_rate": 0.29742196977330276,
      "loss": 0.226,
      "num_input_tokens_seen": 2050400,
      "step": 2365
    },
    {
      "epoch": 1.1173974540311173,
      "grad_norm": 0.010240446776151657,
      "learning_rate": 0.2974110843793223,
      "loss": 0.2393,
      "num_input_tokens_seen": 2055904,
      "step": 2370
    },
    {
      "epoch": 1.1197548326261197,
      "grad_norm": 0.011086949147284031,
      "learning_rate": 0.2974001762527002,
      "loss": 0.2397,
      "num_input_tokens_seen": 2060208,
      "step": 2375
    },
    {
      "epoch": 1.1221122112211221,
      "grad_norm": 0.011045214720070362,
      "learning_rate": 0.2973892453951186,
      "loss": 0.2349,
      "num_input_tokens_seen": 2064128,
      "step": 2380
    },
    {
      "epoch": 1.1244695898161245,
      "grad_norm": 0.053821492940187454,
      "learning_rate": 0.2973782918082631,
      "loss": 0.2441,
      "num_input_tokens_seen": 2068448,
      "step": 2385
    },
    {
      "epoch": 1.1268269684111267,
      "grad_norm": 0.009985426440834999,
      "learning_rate": 0.29736731549382295,
      "loss": 0.2119,
      "num_input_tokens_seen": 2072784,
      "step": 2390
    },
    {
      "epoch": 1.1291843470061291,
      "grad_norm": 0.01207641139626503,
      "learning_rate": 0.2973563164534908,
      "loss": 0.2299,
      "num_input_tokens_seen": 2076656,
      "step": 2395
    },
    {
      "epoch": 1.1315417256011315,
      "grad_norm": 0.004354174714535475,
      "learning_rate": 0.29734529468896287,
      "loss": 0.1931,
      "num_input_tokens_seen": 2081696,
      "step": 2400
    },
    {
      "epoch": 1.1315417256011315,
      "eval_loss": 0.22883723676204681,
      "eval_runtime": 21.9429,
      "eval_samples_per_second": 42.975,
      "eval_steps_per_second": 21.51,
      "num_input_tokens_seen": 2081696,
      "step": 2400
    },
    {
      "epoch": 1.133899104196134,
      "grad_norm": 0.009082062169909477,
      "learning_rate": 0.2973342502019388,
      "loss": 0.1991,
      "num_input_tokens_seen": 2085824,
      "step": 2405
    },
    {
      "epoch": 1.1362564827911363,
      "grad_norm": 0.006109756883233786,
      "learning_rate": 0.2973231829941219,
      "loss": 0.2215,
      "num_input_tokens_seen": 2090032,
      "step": 2410
    },
    {
      "epoch": 1.1386138613861387,
      "grad_norm": 0.008330211974680424,
      "learning_rate": 0.2973120930672188,
      "loss": 0.2293,
      "num_input_tokens_seen": 2093968,
      "step": 2415
    },
    {
      "epoch": 1.140971239981141,
      "grad_norm": 0.005231280345469713,
      "learning_rate": 0.2973009804229397,
      "loss": 0.2146,
      "num_input_tokens_seen": 2098352,
      "step": 2420
    },
    {
      "epoch": 1.1433286185761433,
      "grad_norm": 0.008973370306193829,
      "learning_rate": 0.29728984506299827,
      "loss": 0.2486,
      "num_input_tokens_seen": 2102624,
      "step": 2425
    },
    {
      "epoch": 1.1456859971711457,
      "grad_norm": 0.007892332971096039,
      "learning_rate": 0.2972786869891118,
      "loss": 0.2478,
      "num_input_tokens_seen": 2106672,
      "step": 2430
    },
    {
      "epoch": 1.1480433757661481,
      "grad_norm": 0.006100038066506386,
      "learning_rate": 0.29726750620300096,
      "loss": 0.2144,
      "num_input_tokens_seen": 2111472,
      "step": 2435
    },
    {
      "epoch": 1.1504007543611503,
      "grad_norm": 0.006571868900209665,
      "learning_rate": 0.29725630270639003,
      "loss": 0.2417,
      "num_input_tokens_seen": 2116624,
      "step": 2440
    },
    {
      "epoch": 1.1527581329561527,
      "grad_norm": 0.006925846450030804,
      "learning_rate": 0.2972450765010067,
      "loss": 0.2399,
      "num_input_tokens_seen": 2121456,
      "step": 2445
    },
    {
      "epoch": 1.155115511551155,
      "grad_norm": 0.004056949634104967,
      "learning_rate": 0.29723382758858213,
      "loss": 0.2124,
      "num_input_tokens_seen": 2125808,
      "step": 2450
    },
    {
      "epoch": 1.1574728901461575,
      "grad_norm": 0.00938568264245987,
      "learning_rate": 0.29722255597085107,
      "loss": 0.2284,
      "num_input_tokens_seen": 2131184,
      "step": 2455
    },
    {
      "epoch": 1.15983026874116,
      "grad_norm": 0.010969400405883789,
      "learning_rate": 0.2972112616495518,
      "loss": 0.2507,
      "num_input_tokens_seen": 2135600,
      "step": 2460
    },
    {
      "epoch": 1.1621876473361623,
      "grad_norm": 0.009343928657472134,
      "learning_rate": 0.297199944626426,
      "loss": 0.2142,
      "num_input_tokens_seen": 2140848,
      "step": 2465
    },
    {
      "epoch": 1.1645450259311645,
      "grad_norm": 0.01069515198469162,
      "learning_rate": 0.2971886049032189,
      "loss": 0.2194,
      "num_input_tokens_seen": 2144624,
      "step": 2470
    },
    {
      "epoch": 1.166902404526167,
      "grad_norm": 0.004001745954155922,
      "learning_rate": 0.29717724248167926,
      "loss": 0.2042,
      "num_input_tokens_seen": 2148752,
      "step": 2475
    },
    {
      "epoch": 1.1692597831211693,
      "grad_norm": 0.006407513283193111,
      "learning_rate": 0.29716585736355927,
      "loss": 0.2258,
      "num_input_tokens_seen": 2152624,
      "step": 2480
    },
    {
      "epoch": 1.1716171617161717,
      "grad_norm": 0.0050386106595396996,
      "learning_rate": 0.2971544495506147,
      "loss": 0.1263,
      "num_input_tokens_seen": 2157200,
      "step": 2485
    },
    {
      "epoch": 1.1739745403111739,
      "grad_norm": 0.013103808276355267,
      "learning_rate": 0.2971430190446048,
      "loss": 0.2598,
      "num_input_tokens_seen": 2161664,
      "step": 2490
    },
    {
      "epoch": 1.1763319189061763,
      "grad_norm": 0.0070473793894052505,
      "learning_rate": 0.2971315658472921,
      "loss": 0.2257,
      "num_input_tokens_seen": 2166464,
      "step": 2495
    },
    {
      "epoch": 1.1786892975011787,
      "grad_norm": 0.0071532223373651505,
      "learning_rate": 0.2971200899604431,
      "loss": 0.2438,
      "num_input_tokens_seen": 2170304,
      "step": 2500
    },
    {
      "epoch": 1.181046676096181,
      "grad_norm": 0.01070361863821745,
      "learning_rate": 0.29710859138582735,
      "loss": 0.2402,
      "num_input_tokens_seen": 2175696,
      "step": 2505
    },
    {
      "epoch": 1.1834040546911835,
      "grad_norm": 0.00840723142027855,
      "learning_rate": 0.29709707012521813,
      "loss": 0.2467,
      "num_input_tokens_seen": 2180240,
      "step": 2510
    },
    {
      "epoch": 1.1857614332861859,
      "grad_norm": 0.02045406401157379,
      "learning_rate": 0.29708552618039213,
      "loss": 0.269,
      "num_input_tokens_seen": 2184736,
      "step": 2515
    },
    {
      "epoch": 1.188118811881188,
      "grad_norm": 0.008667693473398685,
      "learning_rate": 0.2970739595531296,
      "loss": 0.2022,
      "num_input_tokens_seen": 2189072,
      "step": 2520
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 0.004271699581295252,
      "learning_rate": 0.2970623702452143,
      "loss": 0.2602,
      "num_input_tokens_seen": 2193680,
      "step": 2525
    },
    {
      "epoch": 1.1928335690711929,
      "grad_norm": 0.005421479698270559,
      "learning_rate": 0.2970507582584334,
      "loss": 0.2283,
      "num_input_tokens_seen": 2197968,
      "step": 2530
    },
    {
      "epoch": 1.1951909476661953,
      "grad_norm": 0.009646003134548664,
      "learning_rate": 0.2970391235945776,
      "loss": 0.2308,
      "num_input_tokens_seen": 2202096,
      "step": 2535
    },
    {
      "epoch": 1.1975483262611974,
      "grad_norm": 0.004087352659553289,
      "learning_rate": 0.2970274662554412,
      "loss": 0.2276,
      "num_input_tokens_seen": 2205728,
      "step": 2540
    },
    {
      "epoch": 1.1999057048561999,
      "grad_norm": 0.0040877158753573895,
      "learning_rate": 0.2970157862428218,
      "loss": 0.2234,
      "num_input_tokens_seen": 2210128,
      "step": 2545
    },
    {
      "epoch": 1.2022630834512023,
      "grad_norm": 0.0038329861126840115,
      "learning_rate": 0.2970040835585206,
      "loss": 0.2037,
      "num_input_tokens_seen": 2214336,
      "step": 2550
    },
    {
      "epoch": 1.2046204620462047,
      "grad_norm": 0.005805865861475468,
      "learning_rate": 0.2969923582043424,
      "loss": 0.2345,
      "num_input_tokens_seen": 2217968,
      "step": 2555
    },
    {
      "epoch": 1.206977840641207,
      "grad_norm": 0.0071100834757089615,
      "learning_rate": 0.2969806101820953,
      "loss": 0.1823,
      "num_input_tokens_seen": 2221888,
      "step": 2560
    },
    {
      "epoch": 1.2093352192362095,
      "grad_norm": 0.009370304644107819,
      "learning_rate": 0.2969688394935911,
      "loss": 0.2358,
      "num_input_tokens_seen": 2226000,
      "step": 2565
    },
    {
      "epoch": 1.2116925978312116,
      "grad_norm": 0.005700339563190937,
      "learning_rate": 0.2969570461406449,
      "loss": 0.262,
      "num_input_tokens_seen": 2230160,
      "step": 2570
    },
    {
      "epoch": 1.214049976426214,
      "grad_norm": 0.0029907983262091875,
      "learning_rate": 0.29694523012507534,
      "loss": 0.2123,
      "num_input_tokens_seen": 2234192,
      "step": 2575
    },
    {
      "epoch": 1.2164073550212164,
      "grad_norm": 0.007089912425726652,
      "learning_rate": 0.2969333914487048,
      "loss": 0.244,
      "num_input_tokens_seen": 2238400,
      "step": 2580
    },
    {
      "epoch": 1.2187647336162188,
      "grad_norm": 0.0034851315431296825,
      "learning_rate": 0.2969215301133587,
      "loss": 0.2459,
      "num_input_tokens_seen": 2242608,
      "step": 2585
    },
    {
      "epoch": 1.221122112211221,
      "grad_norm": 0.0042421650141477585,
      "learning_rate": 0.29690964612086634,
      "loss": 0.2201,
      "num_input_tokens_seen": 2247104,
      "step": 2590
    },
    {
      "epoch": 1.2234794908062234,
      "grad_norm": 0.003272887784987688,
      "learning_rate": 0.2968977394730604,
      "loss": 0.2121,
      "num_input_tokens_seen": 2251328,
      "step": 2595
    },
    {
      "epoch": 1.2258368694012258,
      "grad_norm": 0.005509421695023775,
      "learning_rate": 0.296885810171777,
      "loss": 0.2177,
      "num_input_tokens_seen": 2255952,
      "step": 2600
    },
    {
      "epoch": 1.2258368694012258,
      "eval_loss": 0.2199508547782898,
      "eval_runtime": 21.883,
      "eval_samples_per_second": 43.093,
      "eval_steps_per_second": 21.569,
      "num_input_tokens_seen": 2255952,
      "step": 2600
    },
    {
      "epoch": 1.2281942479962282,
      "grad_norm": 0.002827633870765567,
      "learning_rate": 0.2968738582188558,
      "loss": 0.1916,
      "num_input_tokens_seen": 2260160,
      "step": 2605
    },
    {
      "epoch": 1.2305516265912306,
      "grad_norm": 0.0024956511333584785,
      "learning_rate": 0.2968618836161399,
      "loss": 0.2045,
      "num_input_tokens_seen": 2264016,
      "step": 2610
    },
    {
      "epoch": 1.232909005186233,
      "grad_norm": 0.005286719184368849,
      "learning_rate": 0.296849886365476,
      "loss": 0.2092,
      "num_input_tokens_seen": 2267856,
      "step": 2615
    },
    {
      "epoch": 1.2352663837812352,
      "grad_norm": 0.0038813233841210604,
      "learning_rate": 0.2968378664687142,
      "loss": 0.2899,
      "num_input_tokens_seen": 2271808,
      "step": 2620
    },
    {
      "epoch": 1.2376237623762376,
      "grad_norm": 0.004714853595942259,
      "learning_rate": 0.296825823927708,
      "loss": 0.2136,
      "num_input_tokens_seen": 2276208,
      "step": 2625
    },
    {
      "epoch": 1.23998114097124,
      "grad_norm": 0.003636020701378584,
      "learning_rate": 0.29681375874431476,
      "loss": 0.2322,
      "num_input_tokens_seen": 2279968,
      "step": 2630
    },
    {
      "epoch": 1.2423385195662424,
      "grad_norm": 0.004139642231166363,
      "learning_rate": 0.29680167092039483,
      "loss": 0.2354,
      "num_input_tokens_seen": 2284208,
      "step": 2635
    },
    {
      "epoch": 1.2446958981612446,
      "grad_norm": 0.007399015594273806,
      "learning_rate": 0.2967895604578125,
      "loss": 0.2104,
      "num_input_tokens_seen": 2288944,
      "step": 2640
    },
    {
      "epoch": 1.247053276756247,
      "grad_norm": 0.002826611278578639,
      "learning_rate": 0.2967774273584352,
      "loss": 0.2437,
      "num_input_tokens_seen": 2293008,
      "step": 2645
    },
    {
      "epoch": 1.2494106553512494,
      "grad_norm": 0.00905651692301035,
      "learning_rate": 0.2967652716241342,
      "loss": 0.2396,
      "num_input_tokens_seen": 2298240,
      "step": 2650
    },
    {
      "epoch": 1.2517680339462518,
      "grad_norm": 0.00439093355089426,
      "learning_rate": 0.29675309325678384,
      "loss": 0.2495,
      "num_input_tokens_seen": 2302832,
      "step": 2655
    },
    {
      "epoch": 1.2541254125412542,
      "grad_norm": 0.00828529056161642,
      "learning_rate": 0.29674089225826233,
      "loss": 0.2587,
      "num_input_tokens_seen": 2306576,
      "step": 2660
    },
    {
      "epoch": 1.2564827911362566,
      "grad_norm": 0.002483899239450693,
      "learning_rate": 0.29672866863045116,
      "loss": 0.2288,
      "num_input_tokens_seen": 2310784,
      "step": 2665
    },
    {
      "epoch": 1.2588401697312588,
      "grad_norm": 0.007964280433952808,
      "learning_rate": 0.2967164223752354,
      "loss": 0.2377,
      "num_input_tokens_seen": 2314912,
      "step": 2670
    },
    {
      "epoch": 1.2611975483262612,
      "grad_norm": 0.013144832104444504,
      "learning_rate": 0.2967041534945035,
      "loss": 0.2334,
      "num_input_tokens_seen": 2319344,
      "step": 2675
    },
    {
      "epoch": 1.2635549269212636,
      "grad_norm": 0.005364317912608385,
      "learning_rate": 0.2966918619901476,
      "loss": 0.2073,
      "num_input_tokens_seen": 2323568,
      "step": 2680
    },
    {
      "epoch": 1.265912305516266,
      "grad_norm": 0.004201872274279594,
      "learning_rate": 0.2966795478640631,
      "loss": 0.1986,
      "num_input_tokens_seen": 2328048,
      "step": 2685
    },
    {
      "epoch": 1.2682696841112682,
      "grad_norm": 0.011573298834264278,
      "learning_rate": 0.29666721111814903,
      "loss": 0.2755,
      "num_input_tokens_seen": 2331712,
      "step": 2690
    },
    {
      "epoch": 1.2706270627062706,
      "grad_norm": 0.011270738206803799,
      "learning_rate": 0.2966548517543079,
      "loss": 0.2599,
      "num_input_tokens_seen": 2335808,
      "step": 2695
    },
    {
      "epoch": 1.272984441301273,
      "grad_norm": 0.01084190420806408,
      "learning_rate": 0.29664246977444564,
      "loss": 0.2541,
      "num_input_tokens_seen": 2340032,
      "step": 2700
    },
    {
      "epoch": 1.2753418198962754,
      "grad_norm": 0.005595136433839798,
      "learning_rate": 0.2966300651804717,
      "loss": 0.2465,
      "num_input_tokens_seen": 2344096,
      "step": 2705
    },
    {
      "epoch": 1.2776991984912778,
      "grad_norm": 0.006088255438953638,
      "learning_rate": 0.296617637974299,
      "loss": 0.2076,
      "num_input_tokens_seen": 2348096,
      "step": 2710
    },
    {
      "epoch": 1.2800565770862802,
      "grad_norm": 0.004901167005300522,
      "learning_rate": 0.2966051881578441,
      "loss": 0.2311,
      "num_input_tokens_seen": 2352304,
      "step": 2715
    },
    {
      "epoch": 1.2824139556812824,
      "grad_norm": 0.007242193445563316,
      "learning_rate": 0.29659271573302676,
      "loss": 0.2566,
      "num_input_tokens_seen": 2356736,
      "step": 2720
    },
    {
      "epoch": 1.2847713342762848,
      "grad_norm": 0.006228514015674591,
      "learning_rate": 0.2965802207017705,
      "loss": 0.2045,
      "num_input_tokens_seen": 2361088,
      "step": 2725
    },
    {
      "epoch": 1.2871287128712872,
      "grad_norm": 0.005607002880424261,
      "learning_rate": 0.2965677030660021,
      "loss": 0.1976,
      "num_input_tokens_seen": 2364976,
      "step": 2730
    },
    {
      "epoch": 1.2894860914662896,
      "grad_norm": 0.005090066231787205,
      "learning_rate": 0.2965551628276521,
      "loss": 0.2165,
      "num_input_tokens_seen": 2369792,
      "step": 2735
    },
    {
      "epoch": 1.2918434700612917,
      "grad_norm": 0.01625673845410347,
      "learning_rate": 0.29654259998865423,
      "loss": 0.2577,
      "num_input_tokens_seen": 2374064,
      "step": 2740
    },
    {
      "epoch": 1.2942008486562941,
      "grad_norm": 0.0029056889470666647,
      "learning_rate": 0.2965300145509458,
      "loss": 0.2539,
      "num_input_tokens_seen": 2378624,
      "step": 2745
    },
    {
      "epoch": 1.2965582272512965,
      "grad_norm": 0.0027273704763501883,
      "learning_rate": 0.2965174065164678,
      "loss": 0.2294,
      "num_input_tokens_seen": 2383088,
      "step": 2750
    },
    {
      "epoch": 1.298915605846299,
      "grad_norm": 0.004698619712144136,
      "learning_rate": 0.2965047758871644,
      "loss": 0.2362,
      "num_input_tokens_seen": 2387952,
      "step": 2755
    },
    {
      "epoch": 1.3012729844413014,
      "grad_norm": 0.0027133452240377665,
      "learning_rate": 0.2964921226649835,
      "loss": 0.2409,
      "num_input_tokens_seen": 2391984,
      "step": 2760
    },
    {
      "epoch": 1.3036303630363038,
      "grad_norm": 0.004819863010197878,
      "learning_rate": 0.2964794468518763,
      "loss": 0.2342,
      "num_input_tokens_seen": 2396944,
      "step": 2765
    },
    {
      "epoch": 1.305987741631306,
      "grad_norm": 0.0041705030016601086,
      "learning_rate": 0.2964667484497977,
      "loss": 0.224,
      "num_input_tokens_seen": 2401200,
      "step": 2770
    },
    {
      "epoch": 1.3083451202263083,
      "grad_norm": 0.0023280326277017593,
      "learning_rate": 0.29645402746070587,
      "loss": 0.2027,
      "num_input_tokens_seen": 2405936,
      "step": 2775
    },
    {
      "epoch": 1.3107024988213107,
      "grad_norm": 0.006026082206517458,
      "learning_rate": 0.2964412838865625,
      "loss": 0.2429,
      "num_input_tokens_seen": 2410256,
      "step": 2780
    },
    {
      "epoch": 1.3130598774163131,
      "grad_norm": 0.008059031330049038,
      "learning_rate": 0.29642851772933293,
      "loss": 0.2647,
      "num_input_tokens_seen": 2415136,
      "step": 2785
    },
    {
      "epoch": 1.3154172560113153,
      "grad_norm": 0.005650942679494619,
      "learning_rate": 0.29641572899098567,
      "loss": 0.225,
      "num_input_tokens_seen": 2418976,
      "step": 2790
    },
    {
      "epoch": 1.3177746346063177,
      "grad_norm": 0.004146013408899307,
      "learning_rate": 0.29640291767349314,
      "loss": 0.2144,
      "num_input_tokens_seen": 2422768,
      "step": 2795
    },
    {
      "epoch": 1.3201320132013201,
      "grad_norm": 0.006985373329371214,
      "learning_rate": 0.2963900837788308,
      "loss": 0.2397,
      "num_input_tokens_seen": 2427152,
      "step": 2800
    },
    {
      "epoch": 1.3201320132013201,
      "eval_loss": 0.21909011900424957,
      "eval_runtime": 21.8711,
      "eval_samples_per_second": 43.116,
      "eval_steps_per_second": 21.581,
      "num_input_tokens_seen": 2427152,
      "step": 2800
    },
    {
      "epoch": 1.3224893917963225,
      "grad_norm": 0.006013814825564623,
      "learning_rate": 0.2963772273089779,
      "loss": 0.2284,
      "num_input_tokens_seen": 2431568,
      "step": 2805
    },
    {
      "epoch": 1.324846770391325,
      "grad_norm": 0.0063807847909629345,
      "learning_rate": 0.2963643482659171,
      "loss": 0.2246,
      "num_input_tokens_seen": 2435680,
      "step": 2810
    },
    {
      "epoch": 1.3272041489863273,
      "grad_norm": 0.0033262642100453377,
      "learning_rate": 0.2963514466516345,
      "loss": 0.241,
      "num_input_tokens_seen": 2439456,
      "step": 2815
    },
    {
      "epoch": 1.3295615275813295,
      "grad_norm": 0.003506594803184271,
      "learning_rate": 0.2963385224681196,
      "loss": 0.2324,
      "num_input_tokens_seen": 2444048,
      "step": 2820
    },
    {
      "epoch": 1.331918906176332,
      "grad_norm": 0.0029124070424586535,
      "learning_rate": 0.29632557571736556,
      "loss": 0.1984,
      "num_input_tokens_seen": 2449792,
      "step": 2825
    },
    {
      "epoch": 1.3342762847713343,
      "grad_norm": 0.007220807950943708,
      "learning_rate": 0.2963126064013689,
      "loss": 0.2138,
      "num_input_tokens_seen": 2454688,
      "step": 2830
    },
    {
      "epoch": 1.3366336633663367,
      "grad_norm": 0.007118983659893274,
      "learning_rate": 0.29629961452212966,
      "loss": 0.2404,
      "num_input_tokens_seen": 2458480,
      "step": 2835
    },
    {
      "epoch": 1.338991041961339,
      "grad_norm": 0.003447006456553936,
      "learning_rate": 0.2962866000816513,
      "loss": 0.2265,
      "num_input_tokens_seen": 2462656,
      "step": 2840
    },
    {
      "epoch": 1.3413484205563413,
      "grad_norm": 0.0035183795262128115,
      "learning_rate": 0.2962735630819409,
      "loss": 0.2203,
      "num_input_tokens_seen": 2466960,
      "step": 2845
    },
    {
      "epoch": 1.3437057991513437,
      "grad_norm": 0.003590841544792056,
      "learning_rate": 0.2962605035250089,
      "loss": 0.2224,
      "num_input_tokens_seen": 2472096,
      "step": 2850
    },
    {
      "epoch": 1.346063177746346,
      "grad_norm": 0.0050714765675365925,
      "learning_rate": 0.29624742141286914,
      "loss": 0.2143,
      "num_input_tokens_seen": 2476624,
      "step": 2855
    },
    {
      "epoch": 1.3484205563413485,
      "grad_norm": 0.008710482157766819,
      "learning_rate": 0.29623431674753925,
      "loss": 0.2095,
      "num_input_tokens_seen": 2480400,
      "step": 2860
    },
    {
      "epoch": 1.350777934936351,
      "grad_norm": 0.00221889466047287,
      "learning_rate": 0.29622118953103993,
      "loss": 0.2403,
      "num_input_tokens_seen": 2484416,
      "step": 2865
    },
    {
      "epoch": 1.353135313531353,
      "grad_norm": 0.0035844026133418083,
      "learning_rate": 0.2962080397653957,
      "loss": 0.2343,
      "num_input_tokens_seen": 2489040,
      "step": 2870
    },
    {
      "epoch": 1.3554926921263555,
      "grad_norm": 0.00659441901370883,
      "learning_rate": 0.29619486745263435,
      "loss": 0.2413,
      "num_input_tokens_seen": 2493328,
      "step": 2875
    },
    {
      "epoch": 1.3578500707213579,
      "grad_norm": 0.005109064280986786,
      "learning_rate": 0.2961816725947873,
      "loss": 0.2496,
      "num_input_tokens_seen": 2496928,
      "step": 2880
    },
    {
      "epoch": 1.3602074493163603,
      "grad_norm": 0.0039333379827439785,
      "learning_rate": 0.29616845519388924,
      "loss": 0.2291,
      "num_input_tokens_seen": 2502416,
      "step": 2885
    },
    {
      "epoch": 1.3625648279113625,
      "grad_norm": 0.008502107113599777,
      "learning_rate": 0.2961552152519785,
      "loss": 0.2331,
      "num_input_tokens_seen": 2507296,
      "step": 2890
    },
    {
      "epoch": 1.3649222065063649,
      "grad_norm": 0.008395630866289139,
      "learning_rate": 0.29614195277109695,
      "loss": 0.2518,
      "num_input_tokens_seen": 2511264,
      "step": 2895
    },
    {
      "epoch": 1.3672795851013673,
      "grad_norm": 0.0033641664776951075,
      "learning_rate": 0.2961286677532897,
      "loss": 0.2394,
      "num_input_tokens_seen": 2515616,
      "step": 2900
    },
    {
      "epoch": 1.3696369636963697,
      "grad_norm": 0.005667990539222956,
      "learning_rate": 0.2961153602006055,
      "loss": 0.2214,
      "num_input_tokens_seen": 2520320,
      "step": 2905
    },
    {
      "epoch": 1.371994342291372,
      "grad_norm": 0.009371232241392136,
      "learning_rate": 0.29610203011509656,
      "loss": 0.2466,
      "num_input_tokens_seen": 2524624,
      "step": 2910
    },
    {
      "epoch": 1.3743517208863745,
      "grad_norm": 0.00920378603041172,
      "learning_rate": 0.29608867749881856,
      "loss": 0.2427,
      "num_input_tokens_seen": 2529632,
      "step": 2915
    },
    {
      "epoch": 1.3767090994813767,
      "grad_norm": 0.006576639134436846,
      "learning_rate": 0.29607530235383067,
      "loss": 0.2258,
      "num_input_tokens_seen": 2533664,
      "step": 2920
    },
    {
      "epoch": 1.379066478076379,
      "grad_norm": 0.005481430795043707,
      "learning_rate": 0.2960619046821954,
      "loss": 0.2244,
      "num_input_tokens_seen": 2537792,
      "step": 2925
    },
    {
      "epoch": 1.3814238566713815,
      "grad_norm": 0.0025994509924203157,
      "learning_rate": 0.2960484844859789,
      "loss": 0.1885,
      "num_input_tokens_seen": 2542368,
      "step": 2930
    },
    {
      "epoch": 1.3837812352663839,
      "grad_norm": 0.005182967986911535,
      "learning_rate": 0.29603504176725076,
      "loss": 0.2249,
      "num_input_tokens_seen": 2546512,
      "step": 2935
    },
    {
      "epoch": 1.386138613861386,
      "grad_norm": 0.007806459907442331,
      "learning_rate": 0.296021576528084,
      "loss": 0.2492,
      "num_input_tokens_seen": 2550400,
      "step": 2940
    },
    {
      "epoch": 1.3884959924563884,
      "grad_norm": 0.005865577608346939,
      "learning_rate": 0.29600808877055507,
      "loss": 0.2041,
      "num_input_tokens_seen": 2554480,
      "step": 2945
    },
    {
      "epoch": 1.3908533710513908,
      "grad_norm": 0.007396891713142395,
      "learning_rate": 0.29599457849674404,
      "loss": 0.2373,
      "num_input_tokens_seen": 2558576,
      "step": 2950
    },
    {
      "epoch": 1.3932107496463932,
      "grad_norm": 0.005331482272595167,
      "learning_rate": 0.2959810457087343,
      "loss": 0.2488,
      "num_input_tokens_seen": 2562768,
      "step": 2955
    },
    {
      "epoch": 1.3955681282413956,
      "grad_norm": 0.0031363656744360924,
      "learning_rate": 0.2959674904086128,
      "loss": 0.2439,
      "num_input_tokens_seen": 2566832,
      "step": 2960
    },
    {
      "epoch": 1.397925506836398,
      "grad_norm": 0.0062361364252865314,
      "learning_rate": 0.2959539125984699,
      "loss": 0.2227,
      "num_input_tokens_seen": 2571824,
      "step": 2965
    },
    {
      "epoch": 1.4002828854314002,
      "grad_norm": 0.008645808324217796,
      "learning_rate": 0.2959403122803996,
      "loss": 0.2126,
      "num_input_tokens_seen": 2576160,
      "step": 2970
    },
    {
      "epoch": 1.4026402640264026,
      "grad_norm": 0.002940503181889653,
      "learning_rate": 0.2959266894564991,
      "loss": 0.1898,
      "num_input_tokens_seen": 2580208,
      "step": 2975
    },
    {
      "epoch": 1.404997642621405,
      "grad_norm": 0.0023231273517012596,
      "learning_rate": 0.2959130441288692,
      "loss": 0.2433,
      "num_input_tokens_seen": 2583952,
      "step": 2980
    },
    {
      "epoch": 1.4073550212164074,
      "grad_norm": 0.0023128725588321686,
      "learning_rate": 0.2958993762996143,
      "loss": 0.1761,
      "num_input_tokens_seen": 2587856,
      "step": 2985
    },
    {
      "epoch": 1.4097123998114096,
      "grad_norm": 0.006500492803752422,
      "learning_rate": 0.2958856859708421,
      "loss": 0.213,
      "num_input_tokens_seen": 2592784,
      "step": 2990
    },
    {
      "epoch": 1.412069778406412,
      "grad_norm": 0.005657103378325701,
      "learning_rate": 0.2958719731446638,
      "loss": 0.1953,
      "num_input_tokens_seen": 2596864,
      "step": 2995
    },
    {
      "epoch": 1.4144271570014144,
      "grad_norm": 0.002139743883162737,
      "learning_rate": 0.29585823782319404,
      "loss": 0.1923,
      "num_input_tokens_seen": 2601296,
      "step": 3000
    },
    {
      "epoch": 1.4144271570014144,
      "eval_loss": 0.22986771166324615,
      "eval_runtime": 21.9265,
      "eval_samples_per_second": 43.007,
      "eval_steps_per_second": 21.527,
      "num_input_tokens_seen": 2601296,
      "step": 3000
    },
    {
      "epoch": 1.4167845355964168,
      "grad_norm": 0.006443341728299856,
      "learning_rate": 0.2958444800085511,
      "loss": 0.2067,
      "num_input_tokens_seen": 2604832,
      "step": 3005
    },
    {
      "epoch": 1.4191419141914192,
      "grad_norm": 0.005475982092320919,
      "learning_rate": 0.2958306997028565,
      "loss": 0.2021,
      "num_input_tokens_seen": 2608896,
      "step": 3010
    },
    {
      "epoch": 1.4214992927864216,
      "grad_norm": 0.003305422607809305,
      "learning_rate": 0.2958168969082354,
      "loss": 0.2454,
      "num_input_tokens_seen": 2612960,
      "step": 3015
    },
    {
      "epoch": 1.4238566713814238,
      "grad_norm": 0.003398105502128601,
      "learning_rate": 0.2958030716268164,
      "loss": 0.2323,
      "num_input_tokens_seen": 2616976,
      "step": 3020
    },
    {
      "epoch": 1.4262140499764262,
      "grad_norm": 0.007261589169502258,
      "learning_rate": 0.2957892238607314,
      "loss": 0.2097,
      "num_input_tokens_seen": 2620928,
      "step": 3025
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.002714838832616806,
      "learning_rate": 0.2957753536121161,
      "loss": 0.2191,
      "num_input_tokens_seen": 2625328,
      "step": 3030
    },
    {
      "epoch": 1.430928807166431,
      "grad_norm": 0.0059647406451404095,
      "learning_rate": 0.29576146088310923,
      "loss": 0.1983,
      "num_input_tokens_seen": 2629824,
      "step": 3035
    },
    {
      "epoch": 1.4332861857614332,
      "grad_norm": 0.0041071041487157345,
      "learning_rate": 0.2957475456758533,
      "loss": 0.2821,
      "num_input_tokens_seen": 2633568,
      "step": 3040
    },
    {
      "epoch": 1.4356435643564356,
      "grad_norm": 0.005326292477548122,
      "learning_rate": 0.2957336079924944,
      "loss": 0.1413,
      "num_input_tokens_seen": 2638400,
      "step": 3045
    },
    {
      "epoch": 1.438000942951438,
      "grad_norm": 0.009144378826022148,
      "learning_rate": 0.2957196478351816,
      "loss": 0.2384,
      "num_input_tokens_seen": 2643200,
      "step": 3050
    },
    {
      "epoch": 1.4403583215464404,
      "grad_norm": 0.0028494142461568117,
      "learning_rate": 0.295705665206068,
      "loss": 0.1904,
      "num_input_tokens_seen": 2647120,
      "step": 3055
    },
    {
      "epoch": 1.4427157001414428,
      "grad_norm": 0.011923309415578842,
      "learning_rate": 0.2956916601073097,
      "loss": 0.2623,
      "num_input_tokens_seen": 2651088,
      "step": 3060
    },
    {
      "epoch": 1.4450730787364452,
      "grad_norm": 0.005525130778551102,
      "learning_rate": 0.29567763254106655,
      "loss": 0.257,
      "num_input_tokens_seen": 2655616,
      "step": 3065
    },
    {
      "epoch": 1.4474304573314474,
      "grad_norm": 0.003473920514807105,
      "learning_rate": 0.29566358250950175,
      "loss": 0.231,
      "num_input_tokens_seen": 2659680,
      "step": 3070
    },
    {
      "epoch": 1.4497878359264498,
      "grad_norm": 0.004484968725591898,
      "learning_rate": 0.295649510014782,
      "loss": 0.2706,
      "num_input_tokens_seen": 2664048,
      "step": 3075
    },
    {
      "epoch": 1.4521452145214522,
      "grad_norm": 0.005521675106137991,
      "learning_rate": 0.2956354150590775,
      "loss": 0.228,
      "num_input_tokens_seen": 2668464,
      "step": 3080
    },
    {
      "epoch": 1.4545025931164546,
      "grad_norm": 0.009474598802626133,
      "learning_rate": 0.2956212976445618,
      "loss": 0.2345,
      "num_input_tokens_seen": 2672528,
      "step": 3085
    },
    {
      "epoch": 1.4568599717114568,
      "grad_norm": 0.005432508420199156,
      "learning_rate": 0.295607157773412,
      "loss": 0.1872,
      "num_input_tokens_seen": 2677152,
      "step": 3090
    },
    {
      "epoch": 1.4592173503064592,
      "grad_norm": 0.0029296104330569506,
      "learning_rate": 0.2955929954478087,
      "loss": 0.1983,
      "num_input_tokens_seen": 2681600,
      "step": 3095
    },
    {
      "epoch": 1.4615747289014616,
      "grad_norm": 0.013688499107956886,
      "learning_rate": 0.29557881066993585,
      "loss": 0.2312,
      "num_input_tokens_seen": 2686096,
      "step": 3100
    },
    {
      "epoch": 1.463932107496464,
      "grad_norm": 0.0052721574902534485,
      "learning_rate": 0.29556460344198093,
      "loss": 0.3167,
      "num_input_tokens_seen": 2690048,
      "step": 3105
    },
    {
      "epoch": 1.4662894860914664,
      "grad_norm": 0.005750676151365042,
      "learning_rate": 0.29555037376613486,
      "loss": 0.2292,
      "num_input_tokens_seen": 2693824,
      "step": 3110
    },
    {
      "epoch": 1.4686468646864688,
      "grad_norm": 0.0030463619623333216,
      "learning_rate": 0.29553612164459203,
      "loss": 0.238,
      "num_input_tokens_seen": 2698448,
      "step": 3115
    },
    {
      "epoch": 1.471004243281471,
      "grad_norm": 0.0037915317807346582,
      "learning_rate": 0.29552184707955037,
      "loss": 0.2368,
      "num_input_tokens_seen": 2702480,
      "step": 3120
    },
    {
      "epoch": 1.4733616218764733,
      "grad_norm": 0.004207046236842871,
      "learning_rate": 0.29550755007321117,
      "loss": 0.2277,
      "num_input_tokens_seen": 2706640,
      "step": 3125
    },
    {
      "epoch": 1.4757190004714758,
      "grad_norm": 0.003830442437902093,
      "learning_rate": 0.29549323062777916,
      "loss": 0.2376,
      "num_input_tokens_seen": 2711056,
      "step": 3130
    },
    {
      "epoch": 1.4780763790664782,
      "grad_norm": 0.003404313465580344,
      "learning_rate": 0.29547888874546263,
      "loss": 0.2209,
      "num_input_tokens_seen": 2715024,
      "step": 3135
    },
    {
      "epoch": 1.4804337576614803,
      "grad_norm": 0.0019108677515760064,
      "learning_rate": 0.2954645244284732,
      "loss": 0.2028,
      "num_input_tokens_seen": 2719120,
      "step": 3140
    },
    {
      "epoch": 1.4827911362564827,
      "grad_norm": 0.010326757095754147,
      "learning_rate": 0.2954501376790261,
      "loss": 0.2393,
      "num_input_tokens_seen": 2722992,
      "step": 3145
    },
    {
      "epoch": 1.4851485148514851,
      "grad_norm": 0.006713184993714094,
      "learning_rate": 0.29543572849933997,
      "loss": 0.22,
      "num_input_tokens_seen": 2727408,
      "step": 3150
    },
    {
      "epoch": 1.4875058934464875,
      "grad_norm": 0.005542915314435959,
      "learning_rate": 0.2954212968916368,
      "loss": 0.1752,
      "num_input_tokens_seen": 2731008,
      "step": 3155
    },
    {
      "epoch": 1.48986327204149,
      "grad_norm": 0.004811374936252832,
      "learning_rate": 0.29540684285814217,
      "loss": 0.2303,
      "num_input_tokens_seen": 2735456,
      "step": 3160
    },
    {
      "epoch": 1.4922206506364923,
      "grad_norm": 0.004465905018150806,
      "learning_rate": 0.2953923664010851,
      "loss": 0.2616,
      "num_input_tokens_seen": 2740544,
      "step": 3165
    },
    {
      "epoch": 1.4945780292314945,
      "grad_norm": 0.009649048559367657,
      "learning_rate": 0.295377867522698,
      "loss": 0.2241,
      "num_input_tokens_seen": 2745424,
      "step": 3170
    },
    {
      "epoch": 1.496935407826497,
      "grad_norm": 0.003139410400763154,
      "learning_rate": 0.2953633462252168,
      "loss": 0.2369,
      "num_input_tokens_seen": 2749536,
      "step": 3175
    },
    {
      "epoch": 1.4992927864214993,
      "grad_norm": 0.006606172304600477,
      "learning_rate": 0.2953488025108809,
      "loss": 0.2072,
      "num_input_tokens_seen": 2755152,
      "step": 3180
    },
    {
      "epoch": 1.5016501650165015,
      "grad_norm": 0.005086277145892382,
      "learning_rate": 0.295334236381933,
      "loss": 0.1791,
      "num_input_tokens_seen": 2760736,
      "step": 3185
    },
    {
      "epoch": 1.504007543611504,
      "grad_norm": 0.014225717633962631,
      "learning_rate": 0.29531964784061954,
      "loss": 0.2315,
      "num_input_tokens_seen": 2765776,
      "step": 3190
    },
    {
      "epoch": 1.5063649222065063,
      "grad_norm": 0.011708388105034828,
      "learning_rate": 0.2953050368891902,
      "loss": 0.303,
      "num_input_tokens_seen": 2770272,
      "step": 3195
    },
    {
      "epoch": 1.5087223008015087,
      "grad_norm": 0.0027984611224383116,
      "learning_rate": 0.29529040352989805,
      "loss": 0.2004,
      "num_input_tokens_seen": 2774672,
      "step": 3200
    },
    {
      "epoch": 1.5087223008015087,
      "eval_loss": 0.21937692165374756,
      "eval_runtime": 21.8923,
      "eval_samples_per_second": 43.075,
      "eval_steps_per_second": 21.56,
      "num_input_tokens_seen": 2774672,
      "step": 3200
    },
    {
      "epoch": 1.511079679396511,
      "grad_norm": 0.0024444940499961376,
      "learning_rate": 0.29527574776499993,
      "loss": 0.2198,
      "num_input_tokens_seen": 2779232,
      "step": 3205
    },
    {
      "epoch": 1.5134370579915135,
      "grad_norm": 0.002786466386169195,
      "learning_rate": 0.2952610695967558,
      "loss": 0.2304,
      "num_input_tokens_seen": 2784320,
      "step": 3210
    },
    {
      "epoch": 1.515794436586516,
      "grad_norm": 0.003062780015170574,
      "learning_rate": 0.29524636902742935,
      "loss": 0.2211,
      "num_input_tokens_seen": 2788400,
      "step": 3215
    },
    {
      "epoch": 1.5181518151815183,
      "grad_norm": 0.002640972612425685,
      "learning_rate": 0.2952316460592875,
      "loss": 0.2342,
      "num_input_tokens_seen": 2792896,
      "step": 3220
    },
    {
      "epoch": 1.5205091937765205,
      "grad_norm": 0.0037808935157954693,
      "learning_rate": 0.29521690069460066,
      "loss": 0.2308,
      "num_input_tokens_seen": 2796640,
      "step": 3225
    },
    {
      "epoch": 1.522866572371523,
      "grad_norm": 0.005789860617369413,
      "learning_rate": 0.29520213293564285,
      "loss": 0.2109,
      "num_input_tokens_seen": 2800752,
      "step": 3230
    },
    {
      "epoch": 1.525223950966525,
      "grad_norm": 0.007786838803440332,
      "learning_rate": 0.29518734278469144,
      "loss": 0.2254,
      "num_input_tokens_seen": 2804464,
      "step": 3235
    },
    {
      "epoch": 1.5275813295615275,
      "grad_norm": 0.003705211915075779,
      "learning_rate": 0.29517253024402723,
      "loss": 0.2017,
      "num_input_tokens_seen": 2809008,
      "step": 3240
    },
    {
      "epoch": 1.5299387081565299,
      "grad_norm": 0.0026761337649077177,
      "learning_rate": 0.2951576953159345,
      "loss": 0.2272,
      "num_input_tokens_seen": 2813344,
      "step": 3245
    },
    {
      "epoch": 1.5322960867515323,
      "grad_norm": 0.0026641436852514744,
      "learning_rate": 0.29514283800270097,
      "loss": 0.2781,
      "num_input_tokens_seen": 2817520,
      "step": 3250
    },
    {
      "epoch": 1.5346534653465347,
      "grad_norm": 0.004338329657912254,
      "learning_rate": 0.2951279583066179,
      "loss": 0.2222,
      "num_input_tokens_seen": 2822160,
      "step": 3255
    },
    {
      "epoch": 1.537010843941537,
      "grad_norm": 0.004540656693279743,
      "learning_rate": 0.2951130562299798,
      "loss": 0.2358,
      "num_input_tokens_seen": 2826624,
      "step": 3260
    },
    {
      "epoch": 1.5393682225365395,
      "grad_norm": 0.0059455507434904575,
      "learning_rate": 0.29509813177508487,
      "loss": 0.2393,
      "num_input_tokens_seen": 2830448,
      "step": 3265
    },
    {
      "epoch": 1.541725601131542,
      "grad_norm": 0.0031284764409065247,
      "learning_rate": 0.2950831849442346,
      "loss": 0.2331,
      "num_input_tokens_seen": 2834192,
      "step": 3270
    },
    {
      "epoch": 1.544082979726544,
      "grad_norm": 0.00307730445638299,
      "learning_rate": 0.2950682157397339,
      "loss": 0.2348,
      "num_input_tokens_seen": 2838736,
      "step": 3275
    },
    {
      "epoch": 1.5464403583215465,
      "grad_norm": 0.0033250777050852776,
      "learning_rate": 0.2950532241638914,
      "loss": 0.2159,
      "num_input_tokens_seen": 2842896,
      "step": 3280
    },
    {
      "epoch": 1.5487977369165487,
      "grad_norm": 0.001959958579391241,
      "learning_rate": 0.2950382102190188,
      "loss": 0.2272,
      "num_input_tokens_seen": 2846496,
      "step": 3285
    },
    {
      "epoch": 1.551155115511551,
      "grad_norm": 0.0016896735178306699,
      "learning_rate": 0.2950231739074316,
      "loss": 0.226,
      "num_input_tokens_seen": 2851568,
      "step": 3290
    },
    {
      "epoch": 1.5535124941065535,
      "grad_norm": 0.003085876815021038,
      "learning_rate": 0.29500811523144843,
      "loss": 0.2137,
      "num_input_tokens_seen": 2855984,
      "step": 3295
    },
    {
      "epoch": 1.5558698727015559,
      "grad_norm": 0.003959935624152422,
      "learning_rate": 0.2949930341933917,
      "loss": 0.2393,
      "num_input_tokens_seen": 2860912,
      "step": 3300
    },
    {
      "epoch": 1.5582272512965583,
      "grad_norm": 0.003104042261838913,
      "learning_rate": 0.29497793079558693,
      "loss": 0.2169,
      "num_input_tokens_seen": 2865312,
      "step": 3305
    },
    {
      "epoch": 1.5605846298915607,
      "grad_norm": 0.0034342617727816105,
      "learning_rate": 0.2949628050403633,
      "loss": 0.231,
      "num_input_tokens_seen": 2869200,
      "step": 3310
    },
    {
      "epoch": 1.562942008486563,
      "grad_norm": 0.005453883204609156,
      "learning_rate": 0.2949476569300535,
      "loss": 0.2159,
      "num_input_tokens_seen": 2872640,
      "step": 3315
    },
    {
      "epoch": 1.5652993870815655,
      "grad_norm": 0.0025117069017142057,
      "learning_rate": 0.29493248646699344,
      "loss": 0.1731,
      "num_input_tokens_seen": 2876848,
      "step": 3320
    },
    {
      "epoch": 1.5676567656765676,
      "grad_norm": 0.0020055363420397043,
      "learning_rate": 0.29491729365352265,
      "loss": 0.2449,
      "num_input_tokens_seen": 2881904,
      "step": 3325
    },
    {
      "epoch": 1.57001414427157,
      "grad_norm": 0.0038363824132829905,
      "learning_rate": 0.29490207849198397,
      "loss": 0.2299,
      "num_input_tokens_seen": 2886192,
      "step": 3330
    },
    {
      "epoch": 1.5723715228665722,
      "grad_norm": 0.006472740322351456,
      "learning_rate": 0.29488684098472384,
      "loss": 0.2158,
      "num_input_tokens_seen": 2890928,
      "step": 3335
    },
    {
      "epoch": 1.5747289014615746,
      "grad_norm": 0.004765896126627922,
      "learning_rate": 0.2948715811340921,
      "loss": 0.182,
      "num_input_tokens_seen": 2895488,
      "step": 3340
    },
    {
      "epoch": 1.577086280056577,
      "grad_norm": 0.004921776708215475,
      "learning_rate": 0.294856298942442,
      "loss": 0.1941,
      "num_input_tokens_seen": 2899712,
      "step": 3345
    },
    {
      "epoch": 1.5794436586515794,
      "grad_norm": 0.00912700779736042,
      "learning_rate": 0.2948409944121302,
      "loss": 0.2422,
      "num_input_tokens_seen": 2903728,
      "step": 3350
    },
    {
      "epoch": 1.5818010372465818,
      "grad_norm": 0.002673919079825282,
      "learning_rate": 0.29482566754551687,
      "loss": 0.249,
      "num_input_tokens_seen": 2907792,
      "step": 3355
    },
    {
      "epoch": 1.5841584158415842,
      "grad_norm": 0.0032459553331136703,
      "learning_rate": 0.2948103183449656,
      "loss": 0.2284,
      "num_input_tokens_seen": 2912352,
      "step": 3360
    },
    {
      "epoch": 1.5865157944365866,
      "grad_norm": 0.006294066086411476,
      "learning_rate": 0.2947949468128435,
      "loss": 0.2191,
      "num_input_tokens_seen": 2916432,
      "step": 3365
    },
    {
      "epoch": 1.588873173031589,
      "grad_norm": 0.002189614111557603,
      "learning_rate": 0.2947795529515209,
      "loss": 0.23,
      "num_input_tokens_seen": 2920416,
      "step": 3370
    },
    {
      "epoch": 1.5912305516265912,
      "grad_norm": 0.007093979511409998,
      "learning_rate": 0.29476413676337193,
      "loss": 0.2505,
      "num_input_tokens_seen": 2924048,
      "step": 3375
    },
    {
      "epoch": 1.5935879302215936,
      "grad_norm": 0.002555058803409338,
      "learning_rate": 0.2947486982507738,
      "loss": 0.2413,
      "num_input_tokens_seen": 2927760,
      "step": 3380
    },
    {
      "epoch": 1.5959453088165958,
      "grad_norm": 0.00733768567442894,
      "learning_rate": 0.29473323741610735,
      "loss": 0.2432,
      "num_input_tokens_seen": 2932448,
      "step": 3385
    },
    {
      "epoch": 1.5983026874115982,
      "grad_norm": 0.005436216481029987,
      "learning_rate": 0.2947177542617569,
      "loss": 0.2511,
      "num_input_tokens_seen": 2936800,
      "step": 3390
    },
    {
      "epoch": 1.6006600660066006,
      "grad_norm": 0.0027357100043445826,
      "learning_rate": 0.2947022487901101,
      "loss": 0.2087,
      "num_input_tokens_seen": 2941136,
      "step": 3395
    },
    {
      "epoch": 1.603017444601603,
      "grad_norm": 0.002693792339414358,
      "learning_rate": 0.2946867210035581,
      "loss": 0.214,
      "num_input_tokens_seen": 2944896,
      "step": 3400
    },
    {
      "epoch": 1.603017444601603,
      "eval_loss": 0.21779467165470123,
      "eval_runtime": 21.9033,
      "eval_samples_per_second": 43.053,
      "eval_steps_per_second": 21.549,
      "num_input_tokens_seen": 2944896,
      "step": 3400
    },
    {
      "epoch": 1.6053748231966054,
      "grad_norm": 0.0027324312832206488,
      "learning_rate": 0.2946711709044954,
      "loss": 0.232,
      "num_input_tokens_seen": 2949328,
      "step": 3405
    },
    {
      "epoch": 1.6077322017916078,
      "grad_norm": 0.0024546149652451277,
      "learning_rate": 0.2946555984953202,
      "loss": 0.201,
      "num_input_tokens_seen": 2953440,
      "step": 3410
    },
    {
      "epoch": 1.6100895803866102,
      "grad_norm": 0.003707753261551261,
      "learning_rate": 0.2946400037784338,
      "loss": 0.2107,
      "num_input_tokens_seen": 2958048,
      "step": 3415
    },
    {
      "epoch": 1.6124469589816126,
      "grad_norm": 0.0018873336957767606,
      "learning_rate": 0.29462438675624114,
      "loss": 0.2459,
      "num_input_tokens_seen": 2962064,
      "step": 3420
    },
    {
      "epoch": 1.6148043375766148,
      "grad_norm": 0.0025487891398370266,
      "learning_rate": 0.2946087474311506,
      "loss": 0.2161,
      "num_input_tokens_seen": 2966208,
      "step": 3425
    },
    {
      "epoch": 1.6171617161716172,
      "grad_norm": 0.0032318630255758762,
      "learning_rate": 0.294593085805574,
      "loss": 0.2324,
      "num_input_tokens_seen": 2970720,
      "step": 3430
    },
    {
      "epoch": 1.6195190947666194,
      "grad_norm": 0.004803461022675037,
      "learning_rate": 0.2945774018819264,
      "loss": 0.2263,
      "num_input_tokens_seen": 2974704,
      "step": 3435
    },
    {
      "epoch": 1.6218764733616218,
      "grad_norm": 0.0028848457150161266,
      "learning_rate": 0.2945616956626266,
      "loss": 0.2225,
      "num_input_tokens_seen": 2978496,
      "step": 3440
    },
    {
      "epoch": 1.6242338519566242,
      "grad_norm": 0.002597562503069639,
      "learning_rate": 0.2945459671500966,
      "loss": 0.2044,
      "num_input_tokens_seen": 2983088,
      "step": 3445
    },
    {
      "epoch": 1.6265912305516266,
      "grad_norm": 0.003441445529460907,
      "learning_rate": 0.2945302163467621,
      "loss": 0.2585,
      "num_input_tokens_seen": 2987632,
      "step": 3450
    },
    {
      "epoch": 1.628948609146629,
      "grad_norm": 0.002385909203439951,
      "learning_rate": 0.2945144432550519,
      "loss": 0.1919,
      "num_input_tokens_seen": 2991520,
      "step": 3455
    },
    {
      "epoch": 1.6313059877416314,
      "grad_norm": 0.009133965708315372,
      "learning_rate": 0.29449864787739843,
      "loss": 0.2424,
      "num_input_tokens_seen": 2995888,
      "step": 3460
    },
    {
      "epoch": 1.6336633663366338,
      "grad_norm": 0.0037928381934762,
      "learning_rate": 0.2944828302162376,
      "loss": 0.2338,
      "num_input_tokens_seen": 2999952,
      "step": 3465
    },
    {
      "epoch": 1.6360207449316362,
      "grad_norm": 0.0028389464132487774,
      "learning_rate": 0.2944669902740087,
      "loss": 0.219,
      "num_input_tokens_seen": 3003440,
      "step": 3470
    },
    {
      "epoch": 1.6383781235266384,
      "grad_norm": 0.006033920217305422,
      "learning_rate": 0.2944511280531544,
      "loss": 0.2345,
      "num_input_tokens_seen": 3008240,
      "step": 3475
    },
    {
      "epoch": 1.6407355021216408,
      "grad_norm": 0.0031067850068211555,
      "learning_rate": 0.29443524355612083,
      "loss": 0.2276,
      "num_input_tokens_seen": 3012672,
      "step": 3480
    },
    {
      "epoch": 1.643092880716643,
      "grad_norm": 0.003312419867143035,
      "learning_rate": 0.29441933678535764,
      "loss": 0.2141,
      "num_input_tokens_seen": 3018048,
      "step": 3485
    },
    {
      "epoch": 1.6454502593116453,
      "grad_norm": 0.004910258110612631,
      "learning_rate": 0.29440340774331786,
      "loss": 0.2392,
      "num_input_tokens_seen": 3021616,
      "step": 3490
    },
    {
      "epoch": 1.6478076379066477,
      "grad_norm": 0.008694070391356945,
      "learning_rate": 0.2943874564324579,
      "loss": 0.2496,
      "num_input_tokens_seen": 3026240,
      "step": 3495
    },
    {
      "epoch": 1.6501650165016502,
      "grad_norm": 0.007466481067240238,
      "learning_rate": 0.2943714828552376,
      "loss": 0.22,
      "num_input_tokens_seen": 3030512,
      "step": 3500
    },
    {
      "epoch": 1.6525223950966526,
      "grad_norm": 0.006169352680444717,
      "learning_rate": 0.29435548701412045,
      "loss": 0.2322,
      "num_input_tokens_seen": 3034656,
      "step": 3505
    },
    {
      "epoch": 1.654879773691655,
      "grad_norm": 0.004392172209918499,
      "learning_rate": 0.2943394689115731,
      "loss": 0.201,
      "num_input_tokens_seen": 3038592,
      "step": 3510
    },
    {
      "epoch": 1.6572371522866574,
      "grad_norm": 0.0062384638004004955,
      "learning_rate": 0.29432342855006577,
      "loss": 0.2792,
      "num_input_tokens_seen": 3042576,
      "step": 3515
    },
    {
      "epoch": 1.6595945308816598,
      "grad_norm": 0.004291083198040724,
      "learning_rate": 0.294307365932072,
      "loss": 0.2427,
      "num_input_tokens_seen": 3046704,
      "step": 3520
    },
    {
      "epoch": 1.661951909476662,
      "grad_norm": 0.0059023890644311905,
      "learning_rate": 0.294291281060069,
      "loss": 0.2121,
      "num_input_tokens_seen": 3051648,
      "step": 3525
    },
    {
      "epoch": 1.6643092880716643,
      "grad_norm": 0.007602191064506769,
      "learning_rate": 0.29427517393653724,
      "loss": 0.2449,
      "num_input_tokens_seen": 3056384,
      "step": 3530
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.0028451127000153065,
      "learning_rate": 0.29425904456396046,
      "loss": 0.2153,
      "num_input_tokens_seen": 3060368,
      "step": 3535
    },
    {
      "epoch": 1.669024045261669,
      "grad_norm": 0.003465003566816449,
      "learning_rate": 0.2942428929448262,
      "loss": 0.2145,
      "num_input_tokens_seen": 3063968,
      "step": 3540
    },
    {
      "epoch": 1.6713814238566713,
      "grad_norm": 0.004281880799680948,
      "learning_rate": 0.2942267190816252,
      "loss": 0.1448,
      "num_input_tokens_seen": 3067984,
      "step": 3545
    },
    {
      "epoch": 1.6737388024516737,
      "grad_norm": 0.0018966725328937173,
      "learning_rate": 0.2942105229768516,
      "loss": 0.1877,
      "num_input_tokens_seen": 3072288,
      "step": 3550
    },
    {
      "epoch": 1.6760961810466761,
      "grad_norm": 0.0031444847118109465,
      "learning_rate": 0.29419430463300306,
      "loss": 0.17,
      "num_input_tokens_seen": 3076448,
      "step": 3555
    },
    {
      "epoch": 1.6784535596416785,
      "grad_norm": 0.0053614722564816475,
      "learning_rate": 0.2941780640525808,
      "loss": 0.2677,
      "num_input_tokens_seen": 3081072,
      "step": 3560
    },
    {
      "epoch": 1.680810938236681,
      "grad_norm": 0.002356649376451969,
      "learning_rate": 0.2941618012380891,
      "loss": 0.2071,
      "num_input_tokens_seen": 3086160,
      "step": 3565
    },
    {
      "epoch": 1.6831683168316833,
      "grad_norm": 0.004262336064130068,
      "learning_rate": 0.29414551619203605,
      "loss": 0.182,
      "num_input_tokens_seen": 3090608,
      "step": 3570
    },
    {
      "epoch": 1.6855256954266855,
      "grad_norm": 0.004276457708328962,
      "learning_rate": 0.29412920891693295,
      "loss": 0.1906,
      "num_input_tokens_seen": 3095504,
      "step": 3575
    },
    {
      "epoch": 1.687883074021688,
      "grad_norm": 0.0033935445826500654,
      "learning_rate": 0.2941128794152946,
      "loss": 0.2702,
      "num_input_tokens_seen": 3100336,
      "step": 3580
    },
    {
      "epoch": 1.69024045261669,
      "grad_norm": 0.002060750499367714,
      "learning_rate": 0.2940965276896392,
      "loss": 0.2611,
      "num_input_tokens_seen": 3104624,
      "step": 3585
    },
    {
      "epoch": 1.6925978312116925,
      "grad_norm": 0.00911793764680624,
      "learning_rate": 0.2940801537424884,
      "loss": 0.2432,
      "num_input_tokens_seen": 3109184,
      "step": 3590
    },
    {
      "epoch": 1.694955209806695,
      "grad_norm": 0.0062773083336651325,
      "learning_rate": 0.2940637575763673,
      "loss": 0.2471,
      "num_input_tokens_seen": 3113248,
      "step": 3595
    },
    {
      "epoch": 1.6973125884016973,
      "grad_norm": 0.0036801807582378387,
      "learning_rate": 0.2940473391938043,
      "loss": 0.2293,
      "num_input_tokens_seen": 3117216,
      "step": 3600
    },
    {
      "epoch": 1.6973125884016973,
      "eval_loss": 0.23388880491256714,
      "eval_runtime": 21.9072,
      "eval_samples_per_second": 43.045,
      "eval_steps_per_second": 21.545,
      "num_input_tokens_seen": 3117216,
      "step": 3600
    },
    {
      "epoch": 1.6996699669966997,
      "grad_norm": 0.003441160311922431,
      "learning_rate": 0.29403089859733145,
      "loss": 0.2278,
      "num_input_tokens_seen": 3122032,
      "step": 3605
    },
    {
      "epoch": 1.702027345591702,
      "grad_norm": 0.0024703308008611202,
      "learning_rate": 0.294014435789484,
      "loss": 0.2191,
      "num_input_tokens_seen": 3126112,
      "step": 3610
    },
    {
      "epoch": 1.7043847241867045,
      "grad_norm": 0.002759689697995782,
      "learning_rate": 0.2939979507728007,
      "loss": 0.1957,
      "num_input_tokens_seen": 3131088,
      "step": 3615
    },
    {
      "epoch": 1.706742102781707,
      "grad_norm": 0.002302541397511959,
      "learning_rate": 0.2939814435498239,
      "loss": 0.1843,
      "num_input_tokens_seen": 3135632,
      "step": 3620
    },
    {
      "epoch": 1.709099481376709,
      "grad_norm": 0.007768738549202681,
      "learning_rate": 0.29396491412309905,
      "loss": 0.3017,
      "num_input_tokens_seen": 3139712,
      "step": 3625
    },
    {
      "epoch": 1.7114568599717115,
      "grad_norm": 0.003679306013509631,
      "learning_rate": 0.2939483624951753,
      "loss": 0.2353,
      "num_input_tokens_seen": 3143680,
      "step": 3630
    },
    {
      "epoch": 1.7138142385667137,
      "grad_norm": 0.008586880750954151,
      "learning_rate": 0.2939317886686051,
      "loss": 0.2396,
      "num_input_tokens_seen": 3148416,
      "step": 3635
    },
    {
      "epoch": 1.716171617161716,
      "grad_norm": 0.00738940667361021,
      "learning_rate": 0.2939151926459443,
      "loss": 0.2404,
      "num_input_tokens_seen": 3152592,
      "step": 3640
    },
    {
      "epoch": 1.7185289957567185,
      "grad_norm": 0.0029339324682950974,
      "learning_rate": 0.2938985744297522,
      "loss": 0.2343,
      "num_input_tokens_seen": 3156816,
      "step": 3645
    },
    {
      "epoch": 1.7208863743517209,
      "grad_norm": 0.0028502715285867453,
      "learning_rate": 0.29388193402259166,
      "loss": 0.2268,
      "num_input_tokens_seen": 3160672,
      "step": 3650
    },
    {
      "epoch": 1.7232437529467233,
      "grad_norm": 0.0030473375227302313,
      "learning_rate": 0.29386527142702873,
      "loss": 0.2258,
      "num_input_tokens_seen": 3164176,
      "step": 3655
    },
    {
      "epoch": 1.7256011315417257,
      "grad_norm": 0.005669806618243456,
      "learning_rate": 0.293848586645633,
      "loss": 0.2657,
      "num_input_tokens_seen": 3169184,
      "step": 3660
    },
    {
      "epoch": 1.727958510136728,
      "grad_norm": 0.0027305183466523886,
      "learning_rate": 0.2938318796809775,
      "loss": 0.2651,
      "num_input_tokens_seen": 3173056,
      "step": 3665
    },
    {
      "epoch": 1.7303158887317305,
      "grad_norm": 0.0023854009341448545,
      "learning_rate": 0.29381515053563867,
      "loss": 0.2436,
      "num_input_tokens_seen": 3177904,
      "step": 3670
    },
    {
      "epoch": 1.7326732673267327,
      "grad_norm": 0.007409377954900265,
      "learning_rate": 0.29379839921219636,
      "loss": 0.2237,
      "num_input_tokens_seen": 3182448,
      "step": 3675
    },
    {
      "epoch": 1.735030645921735,
      "grad_norm": 0.003367022145539522,
      "learning_rate": 0.2937816257132338,
      "loss": 0.2486,
      "num_input_tokens_seen": 3187056,
      "step": 3680
    },
    {
      "epoch": 1.7373880245167372,
      "grad_norm": 0.005471038166433573,
      "learning_rate": 0.2937648300413376,
      "loss": 0.2217,
      "num_input_tokens_seen": 3191120,
      "step": 3685
    },
    {
      "epoch": 1.7397454031117396,
      "grad_norm": 0.002517069224268198,
      "learning_rate": 0.293748012199098,
      "loss": 0.2406,
      "num_input_tokens_seen": 3194992,
      "step": 3690
    },
    {
      "epoch": 1.742102781706742,
      "grad_norm": 0.00541115365922451,
      "learning_rate": 0.29373117218910844,
      "loss": 0.2079,
      "num_input_tokens_seen": 3199040,
      "step": 3695
    },
    {
      "epoch": 1.7444601603017444,
      "grad_norm": 0.005660151597112417,
      "learning_rate": 0.2937143100139659,
      "loss": 0.2484,
      "num_input_tokens_seen": 3203392,
      "step": 3700
    },
    {
      "epoch": 1.7468175388967468,
      "grad_norm": 0.002180088311433792,
      "learning_rate": 0.29369742567627083,
      "loss": 0.2297,
      "num_input_tokens_seen": 3207296,
      "step": 3705
    },
    {
      "epoch": 1.7491749174917492,
      "grad_norm": 0.0022596705239266157,
      "learning_rate": 0.29368051917862675,
      "loss": 0.2124,
      "num_input_tokens_seen": 3211696,
      "step": 3710
    },
    {
      "epoch": 1.7515322960867516,
      "grad_norm": 0.0023608021438121796,
      "learning_rate": 0.2936635905236411,
      "loss": 0.2204,
      "num_input_tokens_seen": 3215920,
      "step": 3715
    },
    {
      "epoch": 1.753889674681754,
      "grad_norm": 0.002080093137919903,
      "learning_rate": 0.2936466397139244,
      "loss": 0.2412,
      "num_input_tokens_seen": 3219680,
      "step": 3720
    },
    {
      "epoch": 1.7562470532767562,
      "grad_norm": 0.00564040569588542,
      "learning_rate": 0.2936296667520907,
      "loss": 0.2454,
      "num_input_tokens_seen": 3223440,
      "step": 3725
    },
    {
      "epoch": 1.7586044318717586,
      "grad_norm": 0.002177602844312787,
      "learning_rate": 0.2936126716407574,
      "loss": 0.2338,
      "num_input_tokens_seen": 3227680,
      "step": 3730
    },
    {
      "epoch": 1.7609618104667608,
      "grad_norm": 0.0021530482918024063,
      "learning_rate": 0.29359565438254537,
      "loss": 0.2395,
      "num_input_tokens_seen": 3231728,
      "step": 3735
    },
    {
      "epoch": 1.7633191890617632,
      "grad_norm": 0.004011159762740135,
      "learning_rate": 0.29357861498007887,
      "loss": 0.2233,
      "num_input_tokens_seen": 3236112,
      "step": 3740
    },
    {
      "epoch": 1.7656765676567656,
      "grad_norm": 0.003476114943623543,
      "learning_rate": 0.29356155343598567,
      "loss": 0.2307,
      "num_input_tokens_seen": 3240704,
      "step": 3745
    },
    {
      "epoch": 1.768033946251768,
      "grad_norm": 0.002762726740911603,
      "learning_rate": 0.2935444697528968,
      "loss": 0.2196,
      "num_input_tokens_seen": 3244608,
      "step": 3750
    },
    {
      "epoch": 1.7703913248467704,
      "grad_norm": 0.003950927406549454,
      "learning_rate": 0.2935273639334468,
      "loss": 0.2121,
      "num_input_tokens_seen": 3249104,
      "step": 3755
    },
    {
      "epoch": 1.7727487034417728,
      "grad_norm": 0.0017755895387381315,
      "learning_rate": 0.29351023598027365,
      "loss": 0.1752,
      "num_input_tokens_seen": 3252656,
      "step": 3760
    },
    {
      "epoch": 1.7751060820367752,
      "grad_norm": 0.003332328051328659,
      "learning_rate": 0.2934930858960186,
      "loss": 0.1603,
      "num_input_tokens_seen": 3256304,
      "step": 3765
    },
    {
      "epoch": 1.7774634606317776,
      "grad_norm": 0.008481809869408607,
      "learning_rate": 0.29347591368332643,
      "loss": 0.2919,
      "num_input_tokens_seen": 3261104,
      "step": 3770
    },
    {
      "epoch": 1.7798208392267798,
      "grad_norm": 0.0029833847656846046,
      "learning_rate": 0.2934587193448454,
      "loss": 0.2307,
      "num_input_tokens_seen": 3264896,
      "step": 3775
    },
    {
      "epoch": 1.7821782178217822,
      "grad_norm": 0.00415279995650053,
      "learning_rate": 0.29344150288322696,
      "loss": 0.2344,
      "num_input_tokens_seen": 3270368,
      "step": 3780
    },
    {
      "epoch": 1.7845355964167844,
      "grad_norm": 0.003057242138311267,
      "learning_rate": 0.2934242643011263,
      "loss": 0.2507,
      "num_input_tokens_seen": 3275200,
      "step": 3785
    },
    {
      "epoch": 1.7868929750117868,
      "grad_norm": 0.0028615714982151985,
      "learning_rate": 0.2934070036012016,
      "loss": 0.2273,
      "num_input_tokens_seen": 3279552,
      "step": 3790
    },
    {
      "epoch": 1.7892503536067892,
      "grad_norm": 0.005318064242601395,
      "learning_rate": 0.29338972078611475,
      "loss": 0.1883,
      "num_input_tokens_seen": 3283712,
      "step": 3795
    },
    {
      "epoch": 1.7916077322017916,
      "grad_norm": 0.0028827835340052843,
      "learning_rate": 0.2933724158585311,
      "loss": 0.2782,
      "num_input_tokens_seen": 3287952,
      "step": 3800
    },
    {
      "epoch": 1.7916077322017916,
      "eval_loss": 0.21904699504375458,
      "eval_runtime": 21.949,
      "eval_samples_per_second": 42.963,
      "eval_steps_per_second": 21.504,
      "num_input_tokens_seen": 3287952,
      "step": 3800
    },
    {
      "epoch": 1.793965110796794,
      "grad_norm": 0.0036394246853888035,
      "learning_rate": 0.29335508882111916,
      "loss": 0.2082,
      "num_input_tokens_seen": 3292448,
      "step": 3805
    },
    {
      "epoch": 1.7963224893917964,
      "grad_norm": 0.0035431573633104563,
      "learning_rate": 0.29333773967655097,
      "loss": 0.203,
      "num_input_tokens_seen": 3297120,
      "step": 3810
    },
    {
      "epoch": 1.7986798679867988,
      "grad_norm": 0.003945441916584969,
      "learning_rate": 0.2933203684275021,
      "loss": 0.2455,
      "num_input_tokens_seen": 3300880,
      "step": 3815
    },
    {
      "epoch": 1.8010372465818012,
      "grad_norm": 0.006601437926292419,
      "learning_rate": 0.2933029750766513,
      "loss": 0.2134,
      "num_input_tokens_seen": 3304864,
      "step": 3820
    },
    {
      "epoch": 1.8033946251768034,
      "grad_norm": 0.007759299594908953,
      "learning_rate": 0.2932855596266809,
      "loss": 0.2645,
      "num_input_tokens_seen": 3309168,
      "step": 3825
    },
    {
      "epoch": 1.8057520037718058,
      "grad_norm": 0.004086011089384556,
      "learning_rate": 0.2932681220802765,
      "loss": 0.2295,
      "num_input_tokens_seen": 3313808,
      "step": 3830
    },
    {
      "epoch": 1.808109382366808,
      "grad_norm": 0.006060452666133642,
      "learning_rate": 0.2932506624401274,
      "loss": 0.2379,
      "num_input_tokens_seen": 3318656,
      "step": 3835
    },
    {
      "epoch": 1.8104667609618104,
      "grad_norm": 0.0024422467686235905,
      "learning_rate": 0.29323318070892584,
      "loss": 0.2442,
      "num_input_tokens_seen": 3322656,
      "step": 3840
    },
    {
      "epoch": 1.8128241395568128,
      "grad_norm": 0.007358995731920004,
      "learning_rate": 0.29321567688936784,
      "loss": 0.2194,
      "num_input_tokens_seen": 3327024,
      "step": 3845
    },
    {
      "epoch": 1.8151815181518152,
      "grad_norm": 0.007445203140377998,
      "learning_rate": 0.29319815098415275,
      "loss": 0.1795,
      "num_input_tokens_seen": 3332512,
      "step": 3850
    },
    {
      "epoch": 1.8175388967468176,
      "grad_norm": 0.0037934784777462482,
      "learning_rate": 0.2931806029959832,
      "loss": 0.1654,
      "num_input_tokens_seen": 3336992,
      "step": 3855
    },
    {
      "epoch": 1.81989627534182,
      "grad_norm": 0.008467917330563068,
      "learning_rate": 0.29316303292756535,
      "loss": 0.2031,
      "num_input_tokens_seen": 3342160,
      "step": 3860
    },
    {
      "epoch": 1.8222536539368224,
      "grad_norm": 0.0024131883401423693,
      "learning_rate": 0.29314544078160876,
      "loss": 0.2826,
      "num_input_tokens_seen": 3346688,
      "step": 3865
    },
    {
      "epoch": 1.8246110325318248,
      "grad_norm": 0.00276275840587914,
      "learning_rate": 0.2931278265608263,
      "loss": 0.2029,
      "num_input_tokens_seen": 3350768,
      "step": 3870
    },
    {
      "epoch": 1.826968411126827,
      "grad_norm": 0.0057418146170675755,
      "learning_rate": 0.29311019026793433,
      "loss": 0.2209,
      "num_input_tokens_seen": 3355072,
      "step": 3875
    },
    {
      "epoch": 1.8293257897218294,
      "grad_norm": 0.0031369219068437815,
      "learning_rate": 0.29309253190565254,
      "loss": 0.2296,
      "num_input_tokens_seen": 3359664,
      "step": 3880
    },
    {
      "epoch": 1.8316831683168315,
      "grad_norm": 0.002749488689005375,
      "learning_rate": 0.2930748514767042,
      "loss": 0.229,
      "num_input_tokens_seen": 3363744,
      "step": 3885
    },
    {
      "epoch": 1.834040546911834,
      "grad_norm": 0.002402936341241002,
      "learning_rate": 0.29305714898381574,
      "loss": 0.2326,
      "num_input_tokens_seen": 3368464,
      "step": 3890
    },
    {
      "epoch": 1.8363979255068363,
      "grad_norm": 0.0023259713780134916,
      "learning_rate": 0.29303942442971714,
      "loss": 0.2365,
      "num_input_tokens_seen": 3373520,
      "step": 3895
    },
    {
      "epoch": 1.8387553041018387,
      "grad_norm": 0.005345901940017939,
      "learning_rate": 0.2930216778171417,
      "loss": 0.2337,
      "num_input_tokens_seen": 3377808,
      "step": 3900
    },
    {
      "epoch": 1.8411126826968411,
      "grad_norm": 0.004513942636549473,
      "learning_rate": 0.2930039091488263,
      "loss": 0.1906,
      "num_input_tokens_seen": 3382080,
      "step": 3905
    },
    {
      "epoch": 1.8434700612918435,
      "grad_norm": 0.004008403513580561,
      "learning_rate": 0.29298611842751093,
      "loss": 0.2193,
      "num_input_tokens_seen": 3386736,
      "step": 3910
    },
    {
      "epoch": 1.845827439886846,
      "grad_norm": 0.009298855438828468,
      "learning_rate": 0.29296830565593923,
      "loss": 0.1996,
      "num_input_tokens_seen": 3391184,
      "step": 3915
    },
    {
      "epoch": 1.8481848184818483,
      "grad_norm": 0.0063628763891756535,
      "learning_rate": 0.2929504708368582,
      "loss": 0.2709,
      "num_input_tokens_seen": 3395328,
      "step": 3920
    },
    {
      "epoch": 1.8505421970768505,
      "grad_norm": 0.008808179758489132,
      "learning_rate": 0.29293261397301806,
      "loss": 0.2453,
      "num_input_tokens_seen": 3400192,
      "step": 3925
    },
    {
      "epoch": 1.852899575671853,
      "grad_norm": 0.009236971847712994,
      "learning_rate": 0.29291473506717275,
      "loss": 0.23,
      "num_input_tokens_seen": 3404544,
      "step": 3930
    },
    {
      "epoch": 1.855256954266855,
      "grad_norm": 0.0059799086302518845,
      "learning_rate": 0.29289683412207923,
      "loss": 0.2382,
      "num_input_tokens_seen": 3408912,
      "step": 3935
    },
    {
      "epoch": 1.8576143328618575,
      "grad_norm": 0.006135224364697933,
      "learning_rate": 0.29287891114049813,
      "loss": 0.2203,
      "num_input_tokens_seen": 3413008,
      "step": 3940
    },
    {
      "epoch": 1.85997171145686,
      "grad_norm": 0.005332720931619406,
      "learning_rate": 0.29286096612519347,
      "loss": 0.2154,
      "num_input_tokens_seen": 3417056,
      "step": 3945
    },
    {
      "epoch": 1.8623290900518623,
      "grad_norm": 0.009877987205982208,
      "learning_rate": 0.2928429990789325,
      "loss": 0.2509,
      "num_input_tokens_seen": 3421744,
      "step": 3950
    },
    {
      "epoch": 1.8646864686468647,
      "grad_norm": 0.0034489184617996216,
      "learning_rate": 0.29282501000448596,
      "loss": 0.1824,
      "num_input_tokens_seen": 3425696,
      "step": 3955
    },
    {
      "epoch": 1.8670438472418671,
      "grad_norm": 0.00753172067925334,
      "learning_rate": 0.2928069989046281,
      "loss": 0.209,
      "num_input_tokens_seen": 3431088,
      "step": 3960
    },
    {
      "epoch": 1.8694012258368695,
      "grad_norm": 0.006011512130498886,
      "learning_rate": 0.2927889657821363,
      "loss": 0.2106,
      "num_input_tokens_seen": 3435872,
      "step": 3965
    },
    {
      "epoch": 1.871758604431872,
      "grad_norm": 0.004222434479743242,
      "learning_rate": 0.2927709106397916,
      "loss": 0.2594,
      "num_input_tokens_seen": 3440016,
      "step": 3970
    },
    {
      "epoch": 1.874115983026874,
      "grad_norm": 0.005142661277204752,
      "learning_rate": 0.29275283348037834,
      "loss": 0.2112,
      "num_input_tokens_seen": 3444096,
      "step": 3975
    },
    {
      "epoch": 1.8764733616218765,
      "grad_norm": 0.0034148667473345995,
      "learning_rate": 0.29273473430668423,
      "loss": 0.2224,
      "num_input_tokens_seen": 3448448,
      "step": 3980
    },
    {
      "epoch": 1.8788307402168787,
      "grad_norm": 0.00545239495113492,
      "learning_rate": 0.2927166131215003,
      "loss": 0.2295,
      "num_input_tokens_seen": 3452320,
      "step": 3985
    },
    {
      "epoch": 1.881188118811881,
      "grad_norm": 0.005119252949953079,
      "learning_rate": 0.2926984699276212,
      "loss": 0.2469,
      "num_input_tokens_seen": 3456576,
      "step": 3990
    },
    {
      "epoch": 1.8835454974068835,
      "grad_norm": 0.010816463269293308,
      "learning_rate": 0.29268030472784473,
      "loss": 0.226,
      "num_input_tokens_seen": 3460288,
      "step": 3995
    },
    {
      "epoch": 1.8859028760018859,
      "grad_norm": 0.003116845153272152,
      "learning_rate": 0.2926621175249723,
      "loss": 0.1944,
      "num_input_tokens_seen": 3464640,
      "step": 4000
    },
    {
      "epoch": 1.8859028760018859,
      "eval_loss": 0.22011378407478333,
      "eval_runtime": 21.8911,
      "eval_samples_per_second": 43.077,
      "eval_steps_per_second": 21.561,
      "num_input_tokens_seen": 3464640,
      "step": 4000
    },
    {
      "epoch": 1.8882602545968883,
      "grad_norm": 0.0037346116732805967,
      "learning_rate": 0.29264390832180853,
      "loss": 0.2565,
      "num_input_tokens_seen": 3468608,
      "step": 4005
    },
    {
      "epoch": 1.8906176331918907,
      "grad_norm": 0.0036061585415154696,
      "learning_rate": 0.29262567712116144,
      "loss": 0.2024,
      "num_input_tokens_seen": 3472704,
      "step": 4010
    },
    {
      "epoch": 1.892975011786893,
      "grad_norm": 0.004128908738493919,
      "learning_rate": 0.29260742392584266,
      "loss": 0.2105,
      "num_input_tokens_seen": 3477024,
      "step": 4015
    },
    {
      "epoch": 1.8953323903818955,
      "grad_norm": 0.003922776784747839,
      "learning_rate": 0.292589148738667,
      "loss": 0.191,
      "num_input_tokens_seen": 3481328,
      "step": 4020
    },
    {
      "epoch": 1.8976897689768977,
      "grad_norm": 0.0026220264844596386,
      "learning_rate": 0.2925708515624527,
      "loss": 0.2685,
      "num_input_tokens_seen": 3485440,
      "step": 4025
    },
    {
      "epoch": 1.9000471475719,
      "grad_norm": 0.010648313909769058,
      "learning_rate": 0.29255253240002144,
      "loss": 0.2179,
      "num_input_tokens_seen": 3490576,
      "step": 4030
    },
    {
      "epoch": 1.9024045261669023,
      "grad_norm": 0.0029913729522377253,
      "learning_rate": 0.2925341912541983,
      "loss": 0.2254,
      "num_input_tokens_seen": 3495568,
      "step": 4035
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 0.003488389542326331,
      "learning_rate": 0.2925158281278116,
      "loss": 0.2206,
      "num_input_tokens_seen": 3499280,
      "step": 4040
    },
    {
      "epoch": 1.907119283356907,
      "grad_norm": 0.0034548789262771606,
      "learning_rate": 0.29249744302369324,
      "loss": 0.2283,
      "num_input_tokens_seen": 3504048,
      "step": 4045
    },
    {
      "epoch": 1.9094766619519095,
      "grad_norm": 0.019813381135463715,
      "learning_rate": 0.29247903594467844,
      "loss": 0.2328,
      "num_input_tokens_seen": 3508000,
      "step": 4050
    },
    {
      "epoch": 1.9118340405469119,
      "grad_norm": 0.010443145409226418,
      "learning_rate": 0.2924606068936058,
      "loss": 0.2234,
      "num_input_tokens_seen": 3511904,
      "step": 4055
    },
    {
      "epoch": 1.9141914191419143,
      "grad_norm": 0.018534457311034203,
      "learning_rate": 0.2924421558733173,
      "loss": 0.2508,
      "num_input_tokens_seen": 3516736,
      "step": 4060
    },
    {
      "epoch": 1.9165487977369167,
      "grad_norm": 0.022313378751277924,
      "learning_rate": 0.2924236828866583,
      "loss": 0.2132,
      "num_input_tokens_seen": 3520800,
      "step": 4065
    },
    {
      "epoch": 1.918906176331919,
      "grad_norm": 0.009341590106487274,
      "learning_rate": 0.29240518793647763,
      "loss": 0.1472,
      "num_input_tokens_seen": 3524784,
      "step": 4070
    },
    {
      "epoch": 1.9212635549269212,
      "grad_norm": 0.01987406425178051,
      "learning_rate": 0.29238667102562743,
      "loss": 0.3822,
      "num_input_tokens_seen": 3528992,
      "step": 4075
    },
    {
      "epoch": 1.9236209335219236,
      "grad_norm": 0.016693130135536194,
      "learning_rate": 0.29236813215696317,
      "loss": 0.2562,
      "num_input_tokens_seen": 3533344,
      "step": 4080
    },
    {
      "epoch": 1.9259783121169258,
      "grad_norm": 0.01725536212325096,
      "learning_rate": 0.2923495713333439,
      "loss": 0.2533,
      "num_input_tokens_seen": 3538112,
      "step": 4085
    },
    {
      "epoch": 1.9283356907119282,
      "grad_norm": 0.009405343793332577,
      "learning_rate": 0.29233098855763173,
      "loss": 0.2534,
      "num_input_tokens_seen": 3542608,
      "step": 4090
    },
    {
      "epoch": 1.9306930693069306,
      "grad_norm": 0.010560731403529644,
      "learning_rate": 0.29231238383269254,
      "loss": 0.2213,
      "num_input_tokens_seen": 3547344,
      "step": 4095
    },
    {
      "epoch": 1.933050447901933,
      "grad_norm": 0.03204645216464996,
      "learning_rate": 0.2922937571613954,
      "loss": 0.3757,
      "num_input_tokens_seen": 3551376,
      "step": 4100
    },
    {
      "epoch": 1.9354078264969354,
      "grad_norm": 0.017623309046030045,
      "learning_rate": 0.29227510854661265,
      "loss": 0.2976,
      "num_input_tokens_seen": 3555744,
      "step": 4105
    },
    {
      "epoch": 1.9377652050919378,
      "grad_norm": 0.008795729838311672,
      "learning_rate": 0.29225643799122025,
      "loss": 0.256,
      "num_input_tokens_seen": 3560256,
      "step": 4110
    },
    {
      "epoch": 1.9401225836869402,
      "grad_norm": 0.012652077712118626,
      "learning_rate": 0.2922377454980974,
      "loss": 0.273,
      "num_input_tokens_seen": 3564096,
      "step": 4115
    },
    {
      "epoch": 1.9424799622819426,
      "grad_norm": 0.01688888669013977,
      "learning_rate": 0.29221903107012676,
      "loss": 0.2421,
      "num_input_tokens_seen": 3567888,
      "step": 4120
    },
    {
      "epoch": 1.9448373408769448,
      "grad_norm": 1.7488819360733032,
      "learning_rate": 0.29220029471019426,
      "loss": 0.3146,
      "num_input_tokens_seen": 3571776,
      "step": 4125
    },
    {
      "epoch": 1.9471947194719472,
      "grad_norm": 0.021230120211839676,
      "learning_rate": 0.2921815364211893,
      "loss": 0.2769,
      "num_input_tokens_seen": 3576528,
      "step": 4130
    },
    {
      "epoch": 1.9495520980669494,
      "grad_norm": 0.32730531692504883,
      "learning_rate": 0.29216275620600474,
      "loss": 0.4779,
      "num_input_tokens_seen": 3580352,
      "step": 4135
    },
    {
      "epoch": 1.9519094766619518,
      "grad_norm": 0.05166056007146835,
      "learning_rate": 0.29214395406753657,
      "loss": 0.5019,
      "num_input_tokens_seen": 3586272,
      "step": 4140
    },
    {
      "epoch": 1.9542668552569542,
      "grad_norm": 0.1959736943244934,
      "learning_rate": 0.2921251300086844,
      "loss": 0.3479,
      "num_input_tokens_seen": 3590144,
      "step": 4145
    },
    {
      "epoch": 1.9566242338519566,
      "grad_norm": 0.03841518610715866,
      "learning_rate": 0.2921062840323511,
      "loss": 0.2291,
      "num_input_tokens_seen": 3594656,
      "step": 4150
    },
    {
      "epoch": 1.958981612446959,
      "grad_norm": 0.013489766977727413,
      "learning_rate": 0.29208741614144307,
      "loss": 0.3711,
      "num_input_tokens_seen": 3599536,
      "step": 4155
    },
    {
      "epoch": 1.9613389910419614,
      "grad_norm": 0.01739419996738434,
      "learning_rate": 0.2920685263388698,
      "loss": 0.3241,
      "num_input_tokens_seen": 3605136,
      "step": 4160
    },
    {
      "epoch": 1.9636963696369638,
      "grad_norm": 0.007645237725228071,
      "learning_rate": 0.2920496146275445,
      "loss": 0.2107,
      "num_input_tokens_seen": 3609728,
      "step": 4165
    },
    {
      "epoch": 1.9660537482319662,
      "grad_norm": 0.015126136131584644,
      "learning_rate": 0.29203068101038343,
      "loss": 0.2291,
      "num_input_tokens_seen": 3614320,
      "step": 4170
    },
    {
      "epoch": 1.9684111268269684,
      "grad_norm": 0.021407537162303925,
      "learning_rate": 0.2920117254903065,
      "loss": 0.3333,
      "num_input_tokens_seen": 3618400,
      "step": 4175
    },
    {
      "epoch": 1.9707685054219708,
      "grad_norm": 0.0060995579697191715,
      "learning_rate": 0.29199274807023695,
      "loss": 0.2546,
      "num_input_tokens_seen": 3622080,
      "step": 4180
    },
    {
      "epoch": 1.973125884016973,
      "grad_norm": 0.025952722877264023,
      "learning_rate": 0.29197374875310117,
      "loss": 0.3482,
      "num_input_tokens_seen": 3626656,
      "step": 4185
    },
    {
      "epoch": 1.9754832626119754,
      "grad_norm": 0.009526766836643219,
      "learning_rate": 0.2919547275418292,
      "loss": 0.2332,
      "num_input_tokens_seen": 3630912,
      "step": 4190
    },
    {
      "epoch": 1.9778406412069778,
      "grad_norm": 0.00838377233594656,
      "learning_rate": 0.29193568443935436,
      "loss": 0.2395,
      "num_input_tokens_seen": 3634992,
      "step": 4195
    },
    {
      "epoch": 1.9801980198019802,
      "grad_norm": 0.004952629096806049,
      "learning_rate": 0.2919166194486133,
      "loss": 0.2094,
      "num_input_tokens_seen": 3638880,
      "step": 4200
    },
    {
      "epoch": 1.9801980198019802,
      "eval_loss": 0.22351698577404022,
      "eval_runtime": 21.8933,
      "eval_samples_per_second": 43.072,
      "eval_steps_per_second": 21.559,
      "num_input_tokens_seen": 3638880,
      "step": 4200
    },
    {
      "epoch": 1.9825553983969826,
      "grad_norm": 0.008834758773446083,
      "learning_rate": 0.2918975325725461,
      "loss": 0.2112,
      "num_input_tokens_seen": 3643232,
      "step": 4205
    },
    {
      "epoch": 1.984912776991985,
      "grad_norm": 0.008356036618351936,
      "learning_rate": 0.29187842381409607,
      "loss": 0.2477,
      "num_input_tokens_seen": 3647728,
      "step": 4210
    },
    {
      "epoch": 1.9872701555869874,
      "grad_norm": 0.0049659935757517815,
      "learning_rate": 0.29185929317621023,
      "loss": 0.2318,
      "num_input_tokens_seen": 3651248,
      "step": 4215
    },
    {
      "epoch": 1.9896275341819898,
      "grad_norm": 0.004459875635802746,
      "learning_rate": 0.29184014066183867,
      "loss": 0.2309,
      "num_input_tokens_seen": 3655712,
      "step": 4220
    },
    {
      "epoch": 1.991984912776992,
      "grad_norm": 0.09021452069282532,
      "learning_rate": 0.2918209662739349,
      "loss": 0.3733,
      "num_input_tokens_seen": 3660336,
      "step": 4225
    },
    {
      "epoch": 1.9943422913719944,
      "grad_norm": 0.019175348803400993,
      "learning_rate": 0.29180177001545593,
      "loss": 0.2613,
      "num_input_tokens_seen": 3664880,
      "step": 4230
    },
    {
      "epoch": 1.9966996699669965,
      "grad_norm": 0.03967911750078201,
      "learning_rate": 0.29178255188936203,
      "loss": 0.2643,
      "num_input_tokens_seen": 3669216,
      "step": 4235
    },
    {
      "epoch": 1.999057048561999,
      "grad_norm": 0.008160686120390892,
      "learning_rate": 0.2917633118986169,
      "loss": 0.199,
      "num_input_tokens_seen": 3673840,
      "step": 4240
    },
    {
      "epoch": 2.0014144271570014,
      "grad_norm": 0.0055540394969284534,
      "learning_rate": 0.2917440500461875,
      "loss": 0.2959,
      "num_input_tokens_seen": 3678416,
      "step": 4245
    },
    {
      "epoch": 2.0037718057520038,
      "grad_norm": 0.005497124511748552,
      "learning_rate": 0.29172476633504435,
      "loss": 0.2011,
      "num_input_tokens_seen": 3682656,
      "step": 4250
    },
    {
      "epoch": 2.006129184347006,
      "grad_norm": 0.004424527287483215,
      "learning_rate": 0.2917054607681612,
      "loss": 0.2217,
      "num_input_tokens_seen": 3687472,
      "step": 4255
    },
    {
      "epoch": 2.0084865629420086,
      "grad_norm": 0.012912609614431858,
      "learning_rate": 0.29168613334851523,
      "loss": 0.2363,
      "num_input_tokens_seen": 3691168,
      "step": 4260
    },
    {
      "epoch": 2.010843941537011,
      "grad_norm": 0.007412537932395935,
      "learning_rate": 0.2916667840790869,
      "loss": 0.218,
      "num_input_tokens_seen": 3695312,
      "step": 4265
    },
    {
      "epoch": 2.0132013201320134,
      "grad_norm": 0.005817023571580648,
      "learning_rate": 0.2916474129628603,
      "loss": 0.163,
      "num_input_tokens_seen": 3699744,
      "step": 4270
    },
    {
      "epoch": 2.0155586987270158,
      "grad_norm": 0.00935419648885727,
      "learning_rate": 0.29162802000282245,
      "loss": 0.3135,
      "num_input_tokens_seen": 3704416,
      "step": 4275
    },
    {
      "epoch": 2.0179160773220177,
      "grad_norm": 0.00395997567102313,
      "learning_rate": 0.2916086052019642,
      "loss": 0.2077,
      "num_input_tokens_seen": 3709344,
      "step": 4280
    },
    {
      "epoch": 2.02027345591702,
      "grad_norm": 0.0050944071263074875,
      "learning_rate": 0.2915891685632794,
      "loss": 0.242,
      "num_input_tokens_seen": 3714368,
      "step": 4285
    },
    {
      "epoch": 2.0226308345120225,
      "grad_norm": 0.00501962611451745,
      "learning_rate": 0.29156971008976545,
      "loss": 0.2324,
      "num_input_tokens_seen": 3718688,
      "step": 4290
    },
    {
      "epoch": 2.024988213107025,
      "grad_norm": 0.004900871776044369,
      "learning_rate": 0.2915502297844232,
      "loss": 0.229,
      "num_input_tokens_seen": 3723120,
      "step": 4295
    },
    {
      "epoch": 2.0273455917020273,
      "grad_norm": 0.0026056463830173016,
      "learning_rate": 0.2915307276502566,
      "loss": 0.22,
      "num_input_tokens_seen": 3727312,
      "step": 4300
    },
    {
      "epoch": 2.0297029702970297,
      "grad_norm": 0.0133593138307333,
      "learning_rate": 0.29151120369027334,
      "loss": 0.2473,
      "num_input_tokens_seen": 3731696,
      "step": 4305
    },
    {
      "epoch": 2.032060348892032,
      "grad_norm": 0.002827686257660389,
      "learning_rate": 0.29149165790748405,
      "loss": 0.236,
      "num_input_tokens_seen": 3735968,
      "step": 4310
    },
    {
      "epoch": 2.0344177274870345,
      "grad_norm": 0.004360114224255085,
      "learning_rate": 0.291472090304903,
      "loss": 0.235,
      "num_input_tokens_seen": 3740256,
      "step": 4315
    },
    {
      "epoch": 2.036775106082037,
      "grad_norm": 0.010880178771913052,
      "learning_rate": 0.2914525008855478,
      "loss": 0.2443,
      "num_input_tokens_seen": 3744752,
      "step": 4320
    },
    {
      "epoch": 2.0391324846770393,
      "grad_norm": 0.004991465248167515,
      "learning_rate": 0.2914328896524394,
      "loss": 0.2475,
      "num_input_tokens_seen": 3749248,
      "step": 4325
    },
    {
      "epoch": 2.0414898632720413,
      "grad_norm": 0.003895543748512864,
      "learning_rate": 0.291413256608602,
      "loss": 0.1126,
      "num_input_tokens_seen": 3753440,
      "step": 4330
    },
    {
      "epoch": 2.0438472418670437,
      "grad_norm": 0.0074448008090257645,
      "learning_rate": 0.29139360175706336,
      "loss": 0.3136,
      "num_input_tokens_seen": 3758272,
      "step": 4335
    },
    {
      "epoch": 2.046204620462046,
      "grad_norm": 0.007494518533349037,
      "learning_rate": 0.2913739251008544,
      "loss": 0.1993,
      "num_input_tokens_seen": 3762128,
      "step": 4340
    },
    {
      "epoch": 2.0485619990570485,
      "grad_norm": 0.00447418075054884,
      "learning_rate": 0.29135422664300964,
      "loss": 0.2308,
      "num_input_tokens_seen": 3766432,
      "step": 4345
    },
    {
      "epoch": 2.050919377652051,
      "grad_norm": 0.006357372272759676,
      "learning_rate": 0.29133450638656677,
      "loss": 0.2466,
      "num_input_tokens_seen": 3770768,
      "step": 4350
    },
    {
      "epoch": 2.0532767562470533,
      "grad_norm": 0.006542697548866272,
      "learning_rate": 0.2913147643345669,
      "loss": 0.2448,
      "num_input_tokens_seen": 3774752,
      "step": 4355
    },
    {
      "epoch": 2.0556341348420557,
      "grad_norm": 0.004987667314708233,
      "learning_rate": 0.29129500049005447,
      "loss": 0.2046,
      "num_input_tokens_seen": 3778576,
      "step": 4360
    },
    {
      "epoch": 2.057991513437058,
      "grad_norm": 0.004037009086459875,
      "learning_rate": 0.2912752148560773,
      "loss": 0.2197,
      "num_input_tokens_seen": 3782384,
      "step": 4365
    },
    {
      "epoch": 2.0603488920320605,
      "grad_norm": 0.0024866056628525257,
      "learning_rate": 0.2912554074356866,
      "loss": 0.2868,
      "num_input_tokens_seen": 3786512,
      "step": 4370
    },
    {
      "epoch": 2.062706270627063,
      "grad_norm": 0.0036324032116681337,
      "learning_rate": 0.2912355782319371,
      "loss": 0.2465,
      "num_input_tokens_seen": 3790416,
      "step": 4375
    },
    {
      "epoch": 2.065063649222065,
      "grad_norm": 0.00657393503934145,
      "learning_rate": 0.2912157272478864,
      "loss": 0.2717,
      "num_input_tokens_seen": 3794992,
      "step": 4380
    },
    {
      "epoch": 2.0674210278170673,
      "grad_norm": 0.003984222188591957,
      "learning_rate": 0.291195854486596,
      "loss": 0.2342,
      "num_input_tokens_seen": 3799376,
      "step": 4385
    },
    {
      "epoch": 2.0697784064120697,
      "grad_norm": 0.002631403971463442,
      "learning_rate": 0.2911759599511305,
      "loss": 0.2178,
      "num_input_tokens_seen": 3803824,
      "step": 4390
    },
    {
      "epoch": 2.072135785007072,
      "grad_norm": 0.00347145670093596,
      "learning_rate": 0.29115604364455777,
      "loss": 0.1021,
      "num_input_tokens_seen": 3808576,
      "step": 4395
    },
    {
      "epoch": 2.0744931636020745,
      "grad_norm": 0.006253664381802082,
      "learning_rate": 0.2911361055699493,
      "loss": 0.3115,
      "num_input_tokens_seen": 3812624,
      "step": 4400
    },
    {
      "epoch": 2.0744931636020745,
      "eval_loss": 0.25132954120635986,
      "eval_runtime": 21.9168,
      "eval_samples_per_second": 43.026,
      "eval_steps_per_second": 21.536,
      "num_input_tokens_seen": 3812624,
      "step": 4400
    },
    {
      "epoch": 2.076850542197077,
      "grad_norm": 0.0021109282970428467,
      "learning_rate": 0.2911161457303797,
      "loss": 0.2398,
      "num_input_tokens_seen": 3816576,
      "step": 4405
    },
    {
      "epoch": 2.0792079207920793,
      "grad_norm": 0.0023716092109680176,
      "learning_rate": 0.291096164128927,
      "loss": 0.2462,
      "num_input_tokens_seen": 3821328,
      "step": 4410
    },
    {
      "epoch": 2.0815652993870817,
      "grad_norm": 0.005214781034737825,
      "learning_rate": 0.2910761607686727,
      "loss": 0.2291,
      "num_input_tokens_seen": 3825488,
      "step": 4415
    },
    {
      "epoch": 2.083922677982084,
      "grad_norm": 0.001994062913581729,
      "learning_rate": 0.2910561356527016,
      "loss": 0.2421,
      "num_input_tokens_seen": 3829712,
      "step": 4420
    },
    {
      "epoch": 2.0862800565770865,
      "grad_norm": 0.008380950428545475,
      "learning_rate": 0.2910360887841017,
      "loss": 0.2618,
      "num_input_tokens_seen": 3833936,
      "step": 4425
    },
    {
      "epoch": 2.0886374351720884,
      "grad_norm": 0.003959075082093477,
      "learning_rate": 0.2910160201659645,
      "loss": 0.2076,
      "num_input_tokens_seen": 3838816,
      "step": 4430
    },
    {
      "epoch": 2.090994813767091,
      "grad_norm": 0.0017179262358695269,
      "learning_rate": 0.29099592980138494,
      "loss": 0.1584,
      "num_input_tokens_seen": 3843136,
      "step": 4435
    },
    {
      "epoch": 2.0933521923620932,
      "grad_norm": 0.0018219833727926016,
      "learning_rate": 0.29097581769346115,
      "loss": 0.2341,
      "num_input_tokens_seen": 3848144,
      "step": 4440
    },
    {
      "epoch": 2.0957095709570956,
      "grad_norm": 0.003199170809239149,
      "learning_rate": 0.29095568384529463,
      "loss": 0.271,
      "num_input_tokens_seen": 3853584,
      "step": 4445
    },
    {
      "epoch": 2.098066949552098,
      "grad_norm": 0.0019119078060612082,
      "learning_rate": 0.2909355282599903,
      "loss": 0.2257,
      "num_input_tokens_seen": 3857808,
      "step": 4450
    },
    {
      "epoch": 2.1004243281471005,
      "grad_norm": 0.0033691106364130974,
      "learning_rate": 0.29091535094065635,
      "loss": 0.2283,
      "num_input_tokens_seen": 3861840,
      "step": 4455
    },
    {
      "epoch": 2.102781706742103,
      "grad_norm": 0.0015494617400690913,
      "learning_rate": 0.2908951518904045,
      "loss": 0.198,
      "num_input_tokens_seen": 3866128,
      "step": 4460
    },
    {
      "epoch": 2.1051390853371053,
      "grad_norm": 0.0022173866163939238,
      "learning_rate": 0.29087493111234963,
      "loss": 0.1705,
      "num_input_tokens_seen": 3870816,
      "step": 4465
    },
    {
      "epoch": 2.1074964639321077,
      "grad_norm": 0.010074043646454811,
      "learning_rate": 0.29085468860961,
      "loss": 0.2587,
      "num_input_tokens_seen": 3875600,
      "step": 4470
    },
    {
      "epoch": 2.10985384252711,
      "grad_norm": 0.0018980744061991572,
      "learning_rate": 0.2908344243853073,
      "loss": 0.2334,
      "num_input_tokens_seen": 3879552,
      "step": 4475
    },
    {
      "epoch": 2.112211221122112,
      "grad_norm": 0.002582897897809744,
      "learning_rate": 0.2908141384425666,
      "loss": 0.1987,
      "num_input_tokens_seen": 3883808,
      "step": 4480
    },
    {
      "epoch": 2.1145685997171144,
      "grad_norm": 0.0039053717628121376,
      "learning_rate": 0.2907938307845161,
      "loss": 0.2096,
      "num_input_tokens_seen": 3888752,
      "step": 4485
    },
    {
      "epoch": 2.116925978312117,
      "grad_norm": 0.003342817071825266,
      "learning_rate": 0.2907735014142876,
      "loss": 0.25,
      "num_input_tokens_seen": 3893376,
      "step": 4490
    },
    {
      "epoch": 2.119283356907119,
      "grad_norm": 0.0026601029094308615,
      "learning_rate": 0.2907531503350161,
      "loss": 0.2259,
      "num_input_tokens_seen": 3896800,
      "step": 4495
    },
    {
      "epoch": 2.1216407355021216,
      "grad_norm": 0.00332530215382576,
      "learning_rate": 0.29073277754983995,
      "loss": 0.2275,
      "num_input_tokens_seen": 3900832,
      "step": 4500
    },
    {
      "epoch": 2.123998114097124,
      "grad_norm": 0.0022627972066402435,
      "learning_rate": 0.290712383061901,
      "loss": 0.238,
      "num_input_tokens_seen": 3904880,
      "step": 4505
    },
    {
      "epoch": 2.1263554926921264,
      "grad_norm": 0.0037707414012402296,
      "learning_rate": 0.2906919668743443,
      "loss": 0.2182,
      "num_input_tokens_seen": 3909488,
      "step": 4510
    },
    {
      "epoch": 2.128712871287129,
      "grad_norm": 0.009261597879230976,
      "learning_rate": 0.29067152899031823,
      "loss": 0.2514,
      "num_input_tokens_seen": 3913824,
      "step": 4515
    },
    {
      "epoch": 2.1310702498821312,
      "grad_norm": 0.008174052461981773,
      "learning_rate": 0.2906510694129746,
      "loss": 0.2616,
      "num_input_tokens_seen": 3917872,
      "step": 4520
    },
    {
      "epoch": 2.1334276284771336,
      "grad_norm": 0.004461281467229128,
      "learning_rate": 0.2906305881454685,
      "loss": 0.2378,
      "num_input_tokens_seen": 3922384,
      "step": 4525
    },
    {
      "epoch": 2.1357850070721356,
      "grad_norm": 0.0032474680338054895,
      "learning_rate": 0.2906100851909585,
      "loss": 0.2401,
      "num_input_tokens_seen": 3927216,
      "step": 4530
    },
    {
      "epoch": 2.138142385667138,
      "grad_norm": 0.0038526973221451044,
      "learning_rate": 0.29058956055260626,
      "loss": 0.2268,
      "num_input_tokens_seen": 3931584,
      "step": 4535
    },
    {
      "epoch": 2.1404997642621404,
      "grad_norm": 0.010077448561787605,
      "learning_rate": 0.2905690142335771,
      "loss": 0.2424,
      "num_input_tokens_seen": 3936224,
      "step": 4540
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.007212420925498009,
      "learning_rate": 0.29054844623703946,
      "loss": 0.2253,
      "num_input_tokens_seen": 3940880,
      "step": 4545
    },
    {
      "epoch": 2.145214521452145,
      "grad_norm": 0.004083470907062292,
      "learning_rate": 0.2905278565661651,
      "loss": 0.2448,
      "num_input_tokens_seen": 3945136,
      "step": 4550
    },
    {
      "epoch": 2.1475719000471476,
      "grad_norm": 0.002446418860927224,
      "learning_rate": 0.2905072452241293,
      "loss": 0.2267,
      "num_input_tokens_seen": 3948992,
      "step": 4555
    },
    {
      "epoch": 2.14992927864215,
      "grad_norm": 0.006347887217998505,
      "learning_rate": 0.2904866122141106,
      "loss": 0.2588,
      "num_input_tokens_seen": 3953200,
      "step": 4560
    },
    {
      "epoch": 2.1522866572371524,
      "grad_norm": 0.00412141066044569,
      "learning_rate": 0.2904659575392908,
      "loss": 0.2444,
      "num_input_tokens_seen": 3957136,
      "step": 4565
    },
    {
      "epoch": 2.154644035832155,
      "grad_norm": 0.004330959636718035,
      "learning_rate": 0.2904452812028551,
      "loss": 0.2367,
      "num_input_tokens_seen": 3961232,
      "step": 4570
    },
    {
      "epoch": 2.157001414427157,
      "grad_norm": 0.00738293444737792,
      "learning_rate": 0.2904245832079922,
      "loss": 0.236,
      "num_input_tokens_seen": 3964784,
      "step": 4575
    },
    {
      "epoch": 2.159358793022159,
      "grad_norm": 0.008906913921236992,
      "learning_rate": 0.29040386355789377,
      "loss": 0.2667,
      "num_input_tokens_seen": 3969648,
      "step": 4580
    },
    {
      "epoch": 2.1617161716171616,
      "grad_norm": 0.00525430915877223,
      "learning_rate": 0.29038312225575524,
      "loss": 0.2167,
      "num_input_tokens_seen": 3973680,
      "step": 4585
    },
    {
      "epoch": 2.164073550212164,
      "grad_norm": 0.004818788729608059,
      "learning_rate": 0.29036235930477505,
      "loss": 0.1991,
      "num_input_tokens_seen": 3977840,
      "step": 4590
    },
    {
      "epoch": 2.1664309288071664,
      "grad_norm": 0.0020086332224309444,
      "learning_rate": 0.29034157470815514,
      "loss": 0.2443,
      "num_input_tokens_seen": 3981920,
      "step": 4595
    },
    {
      "epoch": 2.1687883074021688,
      "grad_norm": 0.004461600910872221,
      "learning_rate": 0.2903207684691008,
      "loss": 0.2463,
      "num_input_tokens_seen": 3986544,
      "step": 4600
    },
    {
      "epoch": 2.1687883074021688,
      "eval_loss": 0.25094202160835266,
      "eval_runtime": 21.9135,
      "eval_samples_per_second": 43.033,
      "eval_steps_per_second": 21.539,
      "num_input_tokens_seen": 3986544,
      "step": 4600
    },
    {
      "epoch": 2.171145685997171,
      "grad_norm": 0.00898516084998846,
      "learning_rate": 0.29029994059082054,
      "loss": 0.2276,
      "num_input_tokens_seen": 3990896,
      "step": 4605
    },
    {
      "epoch": 2.1735030645921736,
      "grad_norm": 0.004172341898083687,
      "learning_rate": 0.2902790910765264,
      "loss": 0.3738,
      "num_input_tokens_seen": 3995296,
      "step": 4610
    },
    {
      "epoch": 2.175860443187176,
      "grad_norm": 0.0023360135965049267,
      "learning_rate": 0.29025821992943346,
      "loss": 0.2422,
      "num_input_tokens_seen": 3999728,
      "step": 4615
    },
    {
      "epoch": 2.1782178217821784,
      "grad_norm": 0.004342793952673674,
      "learning_rate": 0.29023732715276046,
      "loss": 0.2131,
      "num_input_tokens_seen": 4004160,
      "step": 4620
    },
    {
      "epoch": 2.1805752003771808,
      "grad_norm": 0.0015650418354198337,
      "learning_rate": 0.2902164127497293,
      "loss": 0.2393,
      "num_input_tokens_seen": 4008928,
      "step": 4625
    },
    {
      "epoch": 2.1829325789721827,
      "grad_norm": 0.004077928606420755,
      "learning_rate": 0.2901954767235652,
      "loss": 0.2254,
      "num_input_tokens_seen": 4012928,
      "step": 4630
    },
    {
      "epoch": 2.185289957567185,
      "grad_norm": 0.005993286147713661,
      "learning_rate": 0.2901745190774968,
      "loss": 0.2167,
      "num_input_tokens_seen": 4017376,
      "step": 4635
    },
    {
      "epoch": 2.1876473361621875,
      "grad_norm": 0.0023205699399113655,
      "learning_rate": 0.290153539814756,
      "loss": 0.2384,
      "num_input_tokens_seen": 4021904,
      "step": 4640
    },
    {
      "epoch": 2.19000471475719,
      "grad_norm": 0.0025256965309381485,
      "learning_rate": 0.2901325389385781,
      "loss": 0.2021,
      "num_input_tokens_seen": 4026368,
      "step": 4645
    },
    {
      "epoch": 2.1923620933521923,
      "grad_norm": 0.0029006185941398144,
      "learning_rate": 0.2901115164522016,
      "loss": 0.2535,
      "num_input_tokens_seen": 4030464,
      "step": 4650
    },
    {
      "epoch": 2.1947194719471947,
      "grad_norm": 0.0019522409420460463,
      "learning_rate": 0.29009047235886865,
      "loss": 0.2154,
      "num_input_tokens_seen": 4034560,
      "step": 4655
    },
    {
      "epoch": 2.197076850542197,
      "grad_norm": 0.0028362476732581854,
      "learning_rate": 0.2900694066618243,
      "loss": 0.2273,
      "num_input_tokens_seen": 4038432,
      "step": 4660
    },
    {
      "epoch": 2.1994342291371995,
      "grad_norm": 0.0031916063744574785,
      "learning_rate": 0.2900483193643172,
      "loss": 0.2179,
      "num_input_tokens_seen": 4042432,
      "step": 4665
    },
    {
      "epoch": 2.201791607732202,
      "grad_norm": 0.0029954013880342245,
      "learning_rate": 0.29002721046959934,
      "loss": 0.2076,
      "num_input_tokens_seen": 4046416,
      "step": 4670
    },
    {
      "epoch": 2.2041489863272044,
      "grad_norm": 0.010346957482397556,
      "learning_rate": 0.29000607998092587,
      "loss": 0.287,
      "num_input_tokens_seen": 4050208,
      "step": 4675
    },
    {
      "epoch": 2.2065063649222063,
      "grad_norm": 0.0069283414632081985,
      "learning_rate": 0.2899849279015555,
      "loss": 0.2286,
      "num_input_tokens_seen": 4054928,
      "step": 4680
    },
    {
      "epoch": 2.2088637435172087,
      "grad_norm": 0.004666629247367382,
      "learning_rate": 0.28996375423475007,
      "loss": 0.2268,
      "num_input_tokens_seen": 4059360,
      "step": 4685
    },
    {
      "epoch": 2.211221122112211,
      "grad_norm": 0.002812155056744814,
      "learning_rate": 0.28994255898377486,
      "loss": 0.2029,
      "num_input_tokens_seen": 4063424,
      "step": 4690
    },
    {
      "epoch": 2.2135785007072135,
      "grad_norm": 0.0015963328769430518,
      "learning_rate": 0.2899213421518984,
      "loss": 0.225,
      "num_input_tokens_seen": 4067248,
      "step": 4695
    },
    {
      "epoch": 2.215935879302216,
      "grad_norm": 0.003561423858627677,
      "learning_rate": 0.2899001037423926,
      "loss": 0.2338,
      "num_input_tokens_seen": 4070992,
      "step": 4700
    },
    {
      "epoch": 2.2182932578972183,
      "grad_norm": 0.005524960346519947,
      "learning_rate": 0.28987884375853273,
      "loss": 0.2118,
      "num_input_tokens_seen": 4075712,
      "step": 4705
    },
    {
      "epoch": 2.2206506364922207,
      "grad_norm": 0.0031697568483650684,
      "learning_rate": 0.2898575622035974,
      "loss": 0.2407,
      "num_input_tokens_seen": 4079744,
      "step": 4710
    },
    {
      "epoch": 2.223008015087223,
      "grad_norm": 0.002284466754645109,
      "learning_rate": 0.2898362590808683,
      "loss": 0.2387,
      "num_input_tokens_seen": 4083776,
      "step": 4715
    },
    {
      "epoch": 2.2253653936822255,
      "grad_norm": 0.004113307222723961,
      "learning_rate": 0.2898149343936308,
      "loss": 0.2354,
      "num_input_tokens_seen": 4087232,
      "step": 4720
    },
    {
      "epoch": 2.227722772277228,
      "grad_norm": 0.014603461138904095,
      "learning_rate": 0.2897935881451734,
      "loss": 0.2069,
      "num_input_tokens_seen": 4091840,
      "step": 4725
    },
    {
      "epoch": 2.23008015087223,
      "grad_norm": 0.004420225974172354,
      "learning_rate": 0.28977222033878797,
      "loss": 0.2364,
      "num_input_tokens_seen": 4096176,
      "step": 4730
    },
    {
      "epoch": 2.2324375294672323,
      "grad_norm": 0.0038562866393476725,
      "learning_rate": 0.28975083097776966,
      "loss": 0.2378,
      "num_input_tokens_seen": 4100080,
      "step": 4735
    },
    {
      "epoch": 2.2347949080622347,
      "grad_norm": 0.0030011695344001055,
      "learning_rate": 0.28972942006541696,
      "loss": 0.2129,
      "num_input_tokens_seen": 4106304,
      "step": 4740
    },
    {
      "epoch": 2.237152286657237,
      "grad_norm": 0.003008002880960703,
      "learning_rate": 0.2897079876050318,
      "loss": 0.2591,
      "num_input_tokens_seen": 4111664,
      "step": 4745
    },
    {
      "epoch": 2.2395096652522395,
      "grad_norm": 0.002006371971219778,
      "learning_rate": 0.2896865335999192,
      "loss": 0.2336,
      "num_input_tokens_seen": 4115936,
      "step": 4750
    },
    {
      "epoch": 2.241867043847242,
      "grad_norm": 0.00842322688549757,
      "learning_rate": 0.28966505805338777,
      "loss": 0.2465,
      "num_input_tokens_seen": 4120320,
      "step": 4755
    },
    {
      "epoch": 2.2442244224422443,
      "grad_norm": 0.0027929209172725677,
      "learning_rate": 0.2896435609687492,
      "loss": 0.2396,
      "num_input_tokens_seen": 4124928,
      "step": 4760
    },
    {
      "epoch": 2.2465818010372467,
      "grad_norm": 0.0038449065759778023,
      "learning_rate": 0.2896220423493187,
      "loss": 0.2222,
      "num_input_tokens_seen": 4128672,
      "step": 4765
    },
    {
      "epoch": 2.248939179632249,
      "grad_norm": 0.004670136142522097,
      "learning_rate": 0.28960050219841466,
      "loss": 0.2256,
      "num_input_tokens_seen": 4132848,
      "step": 4770
    },
    {
      "epoch": 2.251296558227251,
      "grad_norm": 0.00642580958083272,
      "learning_rate": 0.28957894051935884,
      "loss": 0.2578,
      "num_input_tokens_seen": 4136416,
      "step": 4775
    },
    {
      "epoch": 2.2536539368222535,
      "grad_norm": 0.0026334747672080994,
      "learning_rate": 0.2895573573154764,
      "loss": 0.1895,
      "num_input_tokens_seen": 4141040,
      "step": 4780
    },
    {
      "epoch": 2.256011315417256,
      "grad_norm": 0.00287412921898067,
      "learning_rate": 0.28953575259009556,
      "loss": 0.2193,
      "num_input_tokens_seen": 4145600,
      "step": 4785
    },
    {
      "epoch": 2.2583686940122583,
      "grad_norm": 0.004092961084097624,
      "learning_rate": 0.2895141263465482,
      "loss": 0.1857,
      "num_input_tokens_seen": 4150336,
      "step": 4790
    },
    {
      "epoch": 2.2607260726072607,
      "grad_norm": 0.0011726474622264504,
      "learning_rate": 0.28949247858816934,
      "loss": 0.2378,
      "num_input_tokens_seen": 4154192,
      "step": 4795
    },
    {
      "epoch": 2.263083451202263,
      "grad_norm": 0.004114753101021051,
      "learning_rate": 0.2894708093182973,
      "loss": 0.1836,
      "num_input_tokens_seen": 4158272,
      "step": 4800
    },
    {
      "epoch": 2.263083451202263,
      "eval_loss": 0.22281309962272644,
      "eval_runtime": 21.8831,
      "eval_samples_per_second": 43.093,
      "eval_steps_per_second": 21.569,
      "num_input_tokens_seen": 4158272,
      "step": 4800
    },
    {
      "epoch": 2.2654408297972655,
      "grad_norm": 0.0053550382144749165,
      "learning_rate": 0.2894491185402737,
      "loss": 0.2881,
      "num_input_tokens_seen": 4162816,
      "step": 4805
    },
    {
      "epoch": 2.267798208392268,
      "grad_norm": 0.002786647994071245,
      "learning_rate": 0.2894274062574437,
      "loss": 0.2427,
      "num_input_tokens_seen": 4166432,
      "step": 4810
    },
    {
      "epoch": 2.2701555869872703,
      "grad_norm": 0.0018372753402218223,
      "learning_rate": 0.2894056724731554,
      "loss": 0.2897,
      "num_input_tokens_seen": 4170416,
      "step": 4815
    },
    {
      "epoch": 2.2725129655822727,
      "grad_norm": 0.002965539461001754,
      "learning_rate": 0.28938391719076056,
      "loss": 0.2495,
      "num_input_tokens_seen": 4174528,
      "step": 4820
    },
    {
      "epoch": 2.274870344177275,
      "grad_norm": 0.0033832949120551348,
      "learning_rate": 0.28936214041361413,
      "loss": 0.2152,
      "num_input_tokens_seen": 4178944,
      "step": 4825
    },
    {
      "epoch": 2.2772277227722775,
      "grad_norm": 0.001343653304502368,
      "learning_rate": 0.2893403421450743,
      "loss": 0.165,
      "num_input_tokens_seen": 4183344,
      "step": 4830
    },
    {
      "epoch": 2.2795851013672794,
      "grad_norm": 0.01073263306170702,
      "learning_rate": 0.2893185223885026,
      "loss": 0.2894,
      "num_input_tokens_seen": 4187376,
      "step": 4835
    },
    {
      "epoch": 2.281942479962282,
      "grad_norm": 0.005695690400898457,
      "learning_rate": 0.289296681147264,
      "loss": 0.2753,
      "num_input_tokens_seen": 4191680,
      "step": 4840
    },
    {
      "epoch": 2.2842998585572842,
      "grad_norm": 0.003689198289066553,
      "learning_rate": 0.28927481842472663,
      "loss": 0.2471,
      "num_input_tokens_seen": 4196080,
      "step": 4845
    },
    {
      "epoch": 2.2866572371522866,
      "grad_norm": 0.00356759550049901,
      "learning_rate": 0.28925293422426207,
      "loss": 0.2367,
      "num_input_tokens_seen": 4200720,
      "step": 4850
    },
    {
      "epoch": 2.289014615747289,
      "grad_norm": 0.0024006229359656572,
      "learning_rate": 0.28923102854924504,
      "loss": 0.2357,
      "num_input_tokens_seen": 4205120,
      "step": 4855
    },
    {
      "epoch": 2.2913719943422914,
      "grad_norm": 0.0019002433400601149,
      "learning_rate": 0.2892091014030537,
      "loss": 0.268,
      "num_input_tokens_seen": 4208320,
      "step": 4860
    },
    {
      "epoch": 2.293729372937294,
      "grad_norm": 0.002187137957662344,
      "learning_rate": 0.2891871527890696,
      "loss": 0.2109,
      "num_input_tokens_seen": 4212976,
      "step": 4865
    },
    {
      "epoch": 2.2960867515322962,
      "grad_norm": 0.00423853937536478,
      "learning_rate": 0.2891651827106773,
      "loss": 0.1861,
      "num_input_tokens_seen": 4218096,
      "step": 4870
    },
    {
      "epoch": 2.298444130127298,
      "grad_norm": 0.0038500367663800716,
      "learning_rate": 0.2891431911712651,
      "loss": 0.225,
      "num_input_tokens_seen": 4222176,
      "step": 4875
    },
    {
      "epoch": 2.3008015087223006,
      "grad_norm": 0.008027452975511551,
      "learning_rate": 0.2891211781742241,
      "loss": 0.2495,
      "num_input_tokens_seen": 4226064,
      "step": 4880
    },
    {
      "epoch": 2.303158887317303,
      "grad_norm": 0.002122135367244482,
      "learning_rate": 0.2890991437229492,
      "loss": 0.2412,
      "num_input_tokens_seen": 4229744,
      "step": 4885
    },
    {
      "epoch": 2.3055162659123054,
      "grad_norm": 0.0014693469274789095,
      "learning_rate": 0.2890770878208383,
      "loss": 0.2333,
      "num_input_tokens_seen": 4233744,
      "step": 4890
    },
    {
      "epoch": 2.307873644507308,
      "grad_norm": 0.0016319864662364125,
      "learning_rate": 0.28905501047129273,
      "loss": 0.228,
      "num_input_tokens_seen": 4238336,
      "step": 4895
    },
    {
      "epoch": 2.31023102310231,
      "grad_norm": 0.0023906566202640533,
      "learning_rate": 0.289032911677717,
      "loss": 0.1994,
      "num_input_tokens_seen": 4242176,
      "step": 4900
    },
    {
      "epoch": 2.3125884016973126,
      "grad_norm": 0.0014183217426761985,
      "learning_rate": 0.28901079144351915,
      "loss": 0.2208,
      "num_input_tokens_seen": 4246832,
      "step": 4905
    },
    {
      "epoch": 2.314945780292315,
      "grad_norm": 0.003341552335768938,
      "learning_rate": 0.2889886497721103,
      "loss": 0.2184,
      "num_input_tokens_seen": 4251424,
      "step": 4910
    },
    {
      "epoch": 2.3173031588873174,
      "grad_norm": 0.0015684900572523475,
      "learning_rate": 0.28896648666690505,
      "loss": 0.2582,
      "num_input_tokens_seen": 4256128,
      "step": 4915
    },
    {
      "epoch": 2.31966053748232,
      "grad_norm": 0.003913819789886475,
      "learning_rate": 0.2889443021313212,
      "loss": 0.2243,
      "num_input_tokens_seen": 4260384,
      "step": 4920
    },
    {
      "epoch": 2.322017916077322,
      "grad_norm": 0.0025833460967987776,
      "learning_rate": 0.28892209616877984,
      "loss": 0.2178,
      "num_input_tokens_seen": 4264480,
      "step": 4925
    },
    {
      "epoch": 2.3243752946723246,
      "grad_norm": 0.0018669584533199668,
      "learning_rate": 0.28889986878270546,
      "loss": 0.217,
      "num_input_tokens_seen": 4267776,
      "step": 4930
    },
    {
      "epoch": 2.3267326732673266,
      "grad_norm": 0.0012349698226898909,
      "learning_rate": 0.28887761997652583,
      "loss": 0.2002,
      "num_input_tokens_seen": 4271856,
      "step": 4935
    },
    {
      "epoch": 2.329090051862329,
      "grad_norm": 0.005057424772530794,
      "learning_rate": 0.2888553497536719,
      "loss": 0.2814,
      "num_input_tokens_seen": 4275568,
      "step": 4940
    },
    {
      "epoch": 2.3314474304573314,
      "grad_norm": 0.0019691118504852057,
      "learning_rate": 0.2888330581175781,
      "loss": 0.2148,
      "num_input_tokens_seen": 4279632,
      "step": 4945
    },
    {
      "epoch": 2.333804809052334,
      "grad_norm": 0.0019750020001083612,
      "learning_rate": 0.28881074507168203,
      "loss": 0.2271,
      "num_input_tokens_seen": 4284016,
      "step": 4950
    },
    {
      "epoch": 2.336162187647336,
      "grad_norm": 0.0015469721984118223,
      "learning_rate": 0.2887884106194247,
      "loss": 0.2383,
      "num_input_tokens_seen": 4288032,
      "step": 4955
    },
    {
      "epoch": 2.3385195662423386,
      "grad_norm": 0.0036088062915951014,
      "learning_rate": 0.28876605476425027,
      "loss": 0.2417,
      "num_input_tokens_seen": 4292496,
      "step": 4960
    },
    {
      "epoch": 2.340876944837341,
      "grad_norm": 0.0014500333927571774,
      "learning_rate": 0.2887436775096064,
      "loss": 0.2406,
      "num_input_tokens_seen": 4296752,
      "step": 4965
    },
    {
      "epoch": 2.3432343234323434,
      "grad_norm": 0.0021647480316460133,
      "learning_rate": 0.2887212788589439,
      "loss": 0.2279,
      "num_input_tokens_seen": 4302048,
      "step": 4970
    },
    {
      "epoch": 2.3455917020273453,
      "grad_norm": 0.004856151062995195,
      "learning_rate": 0.2886988588157169,
      "loss": 0.2267,
      "num_input_tokens_seen": 4307088,
      "step": 4975
    },
    {
      "epoch": 2.3479490806223478,
      "grad_norm": 0.0021772270556539297,
      "learning_rate": 0.28867641738338284,
      "loss": 0.2154,
      "num_input_tokens_seen": 4311184,
      "step": 4980
    },
    {
      "epoch": 2.35030645921735,
      "grad_norm": 0.0025714533403515816,
      "learning_rate": 0.2886539545654026,
      "loss": 0.2352,
      "num_input_tokens_seen": 4315104,
      "step": 4985
    },
    {
      "epoch": 2.3526638378123526,
      "grad_norm": 0.002195181557908654,
      "learning_rate": 0.28863147036524006,
      "loss": 0.2169,
      "num_input_tokens_seen": 4319424,
      "step": 4990
    },
    {
      "epoch": 2.355021216407355,
      "grad_norm": 0.0022226301953196526,
      "learning_rate": 0.2886089647863626,
      "loss": 0.2252,
      "num_input_tokens_seen": 4323712,
      "step": 4995
    },
    {
      "epoch": 2.3573785950023574,
      "grad_norm": 0.0020217387937009335,
      "learning_rate": 0.288586437832241,
      "loss": 0.2252,
      "num_input_tokens_seen": 4328240,
      "step": 5000
    },
    {
      "epoch": 2.3573785950023574,
      "eval_loss": 0.22035084664821625,
      "eval_runtime": 21.9674,
      "eval_samples_per_second": 42.927,
      "eval_steps_per_second": 21.486,
      "num_input_tokens_seen": 4328240,
      "step": 5000
    },
    {
      "epoch": 2.3597359735973598,
      "grad_norm": 0.005042655859142542,
      "learning_rate": 0.28856388950634904,
      "loss": 0.2226,
      "num_input_tokens_seen": 4331856,
      "step": 5005
    },
    {
      "epoch": 2.362093352192362,
      "grad_norm": 0.002474566688761115,
      "learning_rate": 0.288541319812164,
      "loss": 0.233,
      "num_input_tokens_seen": 4336288,
      "step": 5010
    },
    {
      "epoch": 2.3644507307873646,
      "grad_norm": 0.0017734490102156997,
      "learning_rate": 0.2885187287531665,
      "loss": 0.218,
      "num_input_tokens_seen": 4340016,
      "step": 5015
    },
    {
      "epoch": 2.366808109382367,
      "grad_norm": 0.00244091241620481,
      "learning_rate": 0.2884961163328402,
      "loss": 0.2516,
      "num_input_tokens_seen": 4345280,
      "step": 5020
    },
    {
      "epoch": 2.3691654879773694,
      "grad_norm": 0.009196504950523376,
      "learning_rate": 0.28847348255467237,
      "loss": 0.2308,
      "num_input_tokens_seen": 4350496,
      "step": 5025
    },
    {
      "epoch": 2.3715228665723718,
      "grad_norm": 0.00429969048127532,
      "learning_rate": 0.28845082742215333,
      "loss": 0.1995,
      "num_input_tokens_seen": 4355024,
      "step": 5030
    },
    {
      "epoch": 2.3738802451673737,
      "grad_norm": 0.003474870231002569,
      "learning_rate": 0.2884281509387769,
      "loss": 0.2311,
      "num_input_tokens_seen": 4359136,
      "step": 5035
    },
    {
      "epoch": 2.376237623762376,
      "grad_norm": 0.0012972319964319468,
      "learning_rate": 0.2884054531080399,
      "loss": 0.2012,
      "num_input_tokens_seen": 4363264,
      "step": 5040
    },
    {
      "epoch": 2.3785950023573785,
      "grad_norm": 0.0036207414232194424,
      "learning_rate": 0.28838273393344277,
      "loss": 0.2261,
      "num_input_tokens_seen": 4367840,
      "step": 5045
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 0.01373889297246933,
      "learning_rate": 0.288359993418489,
      "loss": 0.2305,
      "num_input_tokens_seen": 4372848,
      "step": 5050
    },
    {
      "epoch": 2.3833097595473833,
      "grad_norm": 0.0014804807724431157,
      "learning_rate": 0.28833723156668556,
      "loss": 0.2357,
      "num_input_tokens_seen": 4376624,
      "step": 5055
    },
    {
      "epoch": 2.3856671381423857,
      "grad_norm": 0.0035263265017420053,
      "learning_rate": 0.2883144483815425,
      "loss": 0.2485,
      "num_input_tokens_seen": 4380944,
      "step": 5060
    },
    {
      "epoch": 2.388024516737388,
      "grad_norm": 0.005051702260971069,
      "learning_rate": 0.28829164386657335,
      "loss": 0.2377,
      "num_input_tokens_seen": 4384880,
      "step": 5065
    },
    {
      "epoch": 2.3903818953323905,
      "grad_norm": 0.008408577181398869,
      "learning_rate": 0.28826881802529486,
      "loss": 0.2289,
      "num_input_tokens_seen": 4389184,
      "step": 5070
    },
    {
      "epoch": 2.3927392739273925,
      "grad_norm": 0.0017322440398856997,
      "learning_rate": 0.28824597086122705,
      "loss": 0.2552,
      "num_input_tokens_seen": 4393424,
      "step": 5075
    },
    {
      "epoch": 2.395096652522395,
      "grad_norm": 0.008912693709135056,
      "learning_rate": 0.28822310237789317,
      "loss": 0.246,
      "num_input_tokens_seen": 4397680,
      "step": 5080
    },
    {
      "epoch": 2.3974540311173973,
      "grad_norm": 0.005066584795713425,
      "learning_rate": 0.2882002125788199,
      "loss": 0.2224,
      "num_input_tokens_seen": 4402288,
      "step": 5085
    },
    {
      "epoch": 2.3998114097123997,
      "grad_norm": 0.009886041283607483,
      "learning_rate": 0.2881773014675371,
      "loss": 0.3468,
      "num_input_tokens_seen": 4406528,
      "step": 5090
    },
    {
      "epoch": 2.402168788307402,
      "grad_norm": 0.0060807447880506516,
      "learning_rate": 0.288154369047578,
      "loss": 0.2512,
      "num_input_tokens_seen": 4411392,
      "step": 5095
    },
    {
      "epoch": 2.4045261669024045,
      "grad_norm": 0.006759710144251585,
      "learning_rate": 0.28813141532247905,
      "loss": 0.2544,
      "num_input_tokens_seen": 4415600,
      "step": 5100
    },
    {
      "epoch": 2.406883545497407,
      "grad_norm": 0.0015812201891094446,
      "learning_rate": 0.28810844029578,
      "loss": 0.1868,
      "num_input_tokens_seen": 4420176,
      "step": 5105
    },
    {
      "epoch": 2.4092409240924093,
      "grad_norm": 0.002313104458153248,
      "learning_rate": 0.2880854439710238,
      "loss": 0.2231,
      "num_input_tokens_seen": 4423968,
      "step": 5110
    },
    {
      "epoch": 2.4115983026874117,
      "grad_norm": 0.00309342285618186,
      "learning_rate": 0.28806242635175694,
      "loss": 0.287,
      "num_input_tokens_seen": 4429840,
      "step": 5115
    },
    {
      "epoch": 2.413955681282414,
      "grad_norm": 0.0055124773643910885,
      "learning_rate": 0.2880393874415289,
      "loss": 0.2546,
      "num_input_tokens_seen": 4434592,
      "step": 5120
    },
    {
      "epoch": 2.4163130598774165,
      "grad_norm": 0.008148107677698135,
      "learning_rate": 0.2880163272438926,
      "loss": 0.2653,
      "num_input_tokens_seen": 4439632,
      "step": 5125
    },
    {
      "epoch": 2.418670438472419,
      "grad_norm": 0.004936034791171551,
      "learning_rate": 0.2879932457624042,
      "loss": 0.2439,
      "num_input_tokens_seen": 4444176,
      "step": 5130
    },
    {
      "epoch": 2.421027817067421,
      "grad_norm": 0.0019003687193617225,
      "learning_rate": 0.2879701430006232,
      "loss": 0.2445,
      "num_input_tokens_seen": 4448640,
      "step": 5135
    },
    {
      "epoch": 2.4233851956624233,
      "grad_norm": 0.006011665798723698,
      "learning_rate": 0.28794701896211233,
      "loss": 0.2702,
      "num_input_tokens_seen": 4453456,
      "step": 5140
    },
    {
      "epoch": 2.4257425742574257,
      "grad_norm": 0.0029711551032960415,
      "learning_rate": 0.28792387365043753,
      "loss": 0.2383,
      "num_input_tokens_seen": 4456736,
      "step": 5145
    },
    {
      "epoch": 2.428099952852428,
      "grad_norm": 0.003661549184471369,
      "learning_rate": 0.28790070706916815,
      "loss": 0.2408,
      "num_input_tokens_seen": 4462368,
      "step": 5150
    },
    {
      "epoch": 2.4304573314474305,
      "grad_norm": 0.003339531598612666,
      "learning_rate": 0.2878775192218768,
      "loss": 0.2297,
      "num_input_tokens_seen": 4468560,
      "step": 5155
    },
    {
      "epoch": 2.432814710042433,
      "grad_norm": 0.005055107641965151,
      "learning_rate": 0.2878543101121393,
      "loss": 0.2261,
      "num_input_tokens_seen": 4473184,
      "step": 5160
    },
    {
      "epoch": 2.4351720886374353,
      "grad_norm": 0.00307185179553926,
      "learning_rate": 0.28783107974353483,
      "loss": 0.2413,
      "num_input_tokens_seen": 4477344,
      "step": 5165
    },
    {
      "epoch": 2.4375294672324377,
      "grad_norm": 0.0029498999938368797,
      "learning_rate": 0.2878078281196457,
      "loss": 0.2132,
      "num_input_tokens_seen": 4481904,
      "step": 5170
    },
    {
      "epoch": 2.4398868458274396,
      "grad_norm": 0.001580321229994297,
      "learning_rate": 0.28778455524405777,
      "loss": 0.2763,
      "num_input_tokens_seen": 4485648,
      "step": 5175
    },
    {
      "epoch": 2.442244224422442,
      "grad_norm": 0.0066320146434009075,
      "learning_rate": 0.2877612611203598,
      "loss": 0.2596,
      "num_input_tokens_seen": 4489376,
      "step": 5180
    },
    {
      "epoch": 2.4446016030174444,
      "grad_norm": 0.0028738700784742832,
      "learning_rate": 0.28773794575214423,
      "loss": 0.2316,
      "num_input_tokens_seen": 4493872,
      "step": 5185
    },
    {
      "epoch": 2.446958981612447,
      "grad_norm": 0.0021227935794740915,
      "learning_rate": 0.28771460914300645,
      "loss": 0.2201,
      "num_input_tokens_seen": 4498016,
      "step": 5190
    },
    {
      "epoch": 2.4493163602074493,
      "grad_norm": 0.0033159791491925716,
      "learning_rate": 0.2876912512965454,
      "loss": 0.2363,
      "num_input_tokens_seen": 4503504,
      "step": 5195
    },
    {
      "epoch": 2.4516737388024517,
      "grad_norm": 0.004337094724178314,
      "learning_rate": 0.287667872216363,
      "loss": 0.2139,
      "num_input_tokens_seen": 4507760,
      "step": 5200
    },
    {
      "epoch": 2.4516737388024517,
      "eval_loss": 0.21937020123004913,
      "eval_runtime": 21.9146,
      "eval_samples_per_second": 43.031,
      "eval_steps_per_second": 21.538,
      "num_input_tokens_seen": 4507760,
      "step": 5200
    },
    {
      "epoch": 2.454031117397454,
      "grad_norm": 0.0028338846750557423,
      "learning_rate": 0.2876444719060647,
      "loss": 0.2255,
      "num_input_tokens_seen": 4512336,
      "step": 5205
    },
    {
      "epoch": 2.4563884959924565,
      "grad_norm": 0.0054929559119045734,
      "learning_rate": 0.287621050369259,
      "loss": 0.2433,
      "num_input_tokens_seen": 4515888,
      "step": 5210
    },
    {
      "epoch": 2.458745874587459,
      "grad_norm": 0.002135999035090208,
      "learning_rate": 0.28759760760955794,
      "loss": 0.2303,
      "num_input_tokens_seen": 4520320,
      "step": 5215
    },
    {
      "epoch": 2.4611032531824613,
      "grad_norm": 0.00253095431253314,
      "learning_rate": 0.2875741436305766,
      "loss": 0.2405,
      "num_input_tokens_seen": 4524544,
      "step": 5220
    },
    {
      "epoch": 2.4634606317774637,
      "grad_norm": 0.0063372207805514336,
      "learning_rate": 0.28755065843593347,
      "loss": 0.233,
      "num_input_tokens_seen": 4528464,
      "step": 5225
    },
    {
      "epoch": 2.465818010372466,
      "grad_norm": 0.005203817505389452,
      "learning_rate": 0.2875271520292502,
      "loss": 0.2161,
      "num_input_tokens_seen": 4532512,
      "step": 5230
    },
    {
      "epoch": 2.468175388967468,
      "grad_norm": 0.0031608010176569223,
      "learning_rate": 0.28750362441415184,
      "loss": 0.2236,
      "num_input_tokens_seen": 4536992,
      "step": 5235
    },
    {
      "epoch": 2.4705327675624704,
      "grad_norm": 0.004605791065841913,
      "learning_rate": 0.28748007559426664,
      "loss": 0.2237,
      "num_input_tokens_seen": 4542336,
      "step": 5240
    },
    {
      "epoch": 2.472890146157473,
      "grad_norm": 0.005680263042449951,
      "learning_rate": 0.2874565055732261,
      "loss": 0.2598,
      "num_input_tokens_seen": 4547808,
      "step": 5245
    },
    {
      "epoch": 2.4752475247524752,
      "grad_norm": 0.003358143847435713,
      "learning_rate": 0.28743291435466495,
      "loss": 0.219,
      "num_input_tokens_seen": 4552768,
      "step": 5250
    },
    {
      "epoch": 2.4776049033474776,
      "grad_norm": 0.0024696127511560917,
      "learning_rate": 0.2874093019422214,
      "loss": 0.2323,
      "num_input_tokens_seen": 4557072,
      "step": 5255
    },
    {
      "epoch": 2.47996228194248,
      "grad_norm": 0.004188489634543657,
      "learning_rate": 0.28738566833953666,
      "loss": 0.2473,
      "num_input_tokens_seen": 4561488,
      "step": 5260
    },
    {
      "epoch": 2.4823196605374824,
      "grad_norm": 0.002688789274543524,
      "learning_rate": 0.28736201355025537,
      "loss": 0.2241,
      "num_input_tokens_seen": 4565952,
      "step": 5265
    },
    {
      "epoch": 2.484677039132485,
      "grad_norm": 0.00432366831228137,
      "learning_rate": 0.28733833757802535,
      "loss": 0.2061,
      "num_input_tokens_seen": 4570944,
      "step": 5270
    },
    {
      "epoch": 2.487034417727487,
      "grad_norm": 0.002762431278824806,
      "learning_rate": 0.28731464042649785,
      "loss": 0.2399,
      "num_input_tokens_seen": 4575936,
      "step": 5275
    },
    {
      "epoch": 2.489391796322489,
      "grad_norm": 0.0021695769391953945,
      "learning_rate": 0.2872909220993271,
      "loss": 0.2282,
      "num_input_tokens_seen": 4580560,
      "step": 5280
    },
    {
      "epoch": 2.4917491749174916,
      "grad_norm": 0.002247544936835766,
      "learning_rate": 0.287267182600171,
      "loss": 0.1939,
      "num_input_tokens_seen": 4584800,
      "step": 5285
    },
    {
      "epoch": 2.494106553512494,
      "grad_norm": 0.0033613366540521383,
      "learning_rate": 0.2872434219326902,
      "loss": 0.2252,
      "num_input_tokens_seen": 4590368,
      "step": 5290
    },
    {
      "epoch": 2.4964639321074964,
      "grad_norm": 0.0024813413619995117,
      "learning_rate": 0.28721964010054907,
      "loss": 0.2655,
      "num_input_tokens_seen": 4594480,
      "step": 5295
    },
    {
      "epoch": 2.498821310702499,
      "grad_norm": 0.0018892339430749416,
      "learning_rate": 0.28719583710741503,
      "loss": 0.225,
      "num_input_tokens_seen": 4598880,
      "step": 5300
    },
    {
      "epoch": 2.501178689297501,
      "grad_norm": 0.002346500987187028,
      "learning_rate": 0.28717201295695877,
      "loss": 0.2216,
      "num_input_tokens_seen": 4602960,
      "step": 5305
    },
    {
      "epoch": 2.5035360678925036,
      "grad_norm": 0.001800031866878271,
      "learning_rate": 0.28714816765285434,
      "loss": 0.2214,
      "num_input_tokens_seen": 4607296,
      "step": 5310
    },
    {
      "epoch": 2.505893446487506,
      "grad_norm": 0.0015908876666799188,
      "learning_rate": 0.28712430119877896,
      "loss": 0.2157,
      "num_input_tokens_seen": 4611872,
      "step": 5315
    },
    {
      "epoch": 2.5082508250825084,
      "grad_norm": 0.0014735175063833594,
      "learning_rate": 0.28710041359841304,
      "loss": 0.2142,
      "num_input_tokens_seen": 4615568,
      "step": 5320
    },
    {
      "epoch": 2.510608203677511,
      "grad_norm": 0.0029877829365432262,
      "learning_rate": 0.28707650485544056,
      "loss": 0.1762,
      "num_input_tokens_seen": 4619520,
      "step": 5325
    },
    {
      "epoch": 2.512965582272513,
      "grad_norm": 0.00255657359957695,
      "learning_rate": 0.28705257497354836,
      "loss": 0.2291,
      "num_input_tokens_seen": 4623232,
      "step": 5330
    },
    {
      "epoch": 2.515322960867515,
      "grad_norm": 0.003087384393438697,
      "learning_rate": 0.28702862395642675,
      "loss": 0.2588,
      "num_input_tokens_seen": 4627312,
      "step": 5335
    },
    {
      "epoch": 2.5176803394625176,
      "grad_norm": 0.0013465832453221083,
      "learning_rate": 0.28700465180776935,
      "loss": 0.2035,
      "num_input_tokens_seen": 4631552,
      "step": 5340
    },
    {
      "epoch": 2.52003771805752,
      "grad_norm": 0.0020368569530546665,
      "learning_rate": 0.2869806585312729,
      "loss": 0.2362,
      "num_input_tokens_seen": 4635568,
      "step": 5345
    },
    {
      "epoch": 2.5223950966525224,
      "grad_norm": 0.004941504914313555,
      "learning_rate": 0.28695664413063754,
      "loss": 0.2234,
      "num_input_tokens_seen": 4639968,
      "step": 5350
    },
    {
      "epoch": 2.5247524752475248,
      "grad_norm": 0.004805352073162794,
      "learning_rate": 0.28693260860956654,
      "loss": 0.228,
      "num_input_tokens_seen": 4643536,
      "step": 5355
    },
    {
      "epoch": 2.527109853842527,
      "grad_norm": 0.004428005311638117,
      "learning_rate": 0.2869085519717665,
      "loss": 0.2771,
      "num_input_tokens_seen": 4648256,
      "step": 5360
    },
    {
      "epoch": 2.5294672324375296,
      "grad_norm": 0.0065118190832436085,
      "learning_rate": 0.28688447422094726,
      "loss": 0.2483,
      "num_input_tokens_seen": 4651712,
      "step": 5365
    },
    {
      "epoch": 2.531824611032532,
      "grad_norm": 0.0033065923489630222,
      "learning_rate": 0.2868603753608219,
      "loss": 0.2619,
      "num_input_tokens_seen": 4656128,
      "step": 5370
    },
    {
      "epoch": 2.534181989627534,
      "grad_norm": 0.00488028209656477,
      "learning_rate": 0.28683625539510665,
      "loss": 0.2399,
      "num_input_tokens_seen": 4660080,
      "step": 5375
    },
    {
      "epoch": 2.5365393682225363,
      "grad_norm": 0.002418127143755555,
      "learning_rate": 0.28681211432752135,
      "loss": 0.2249,
      "num_input_tokens_seen": 4664800,
      "step": 5380
    },
    {
      "epoch": 2.5388967468175387,
      "grad_norm": 0.002541935071349144,
      "learning_rate": 0.2867879521617887,
      "loss": 0.2387,
      "num_input_tokens_seen": 4669040,
      "step": 5385
    },
    {
      "epoch": 2.541254125412541,
      "grad_norm": 0.0015006556641310453,
      "learning_rate": 0.28676376890163485,
      "loss": 0.2397,
      "num_input_tokens_seen": 4673712,
      "step": 5390
    },
    {
      "epoch": 2.5436115040075435,
      "grad_norm": 0.0017031584866344929,
      "learning_rate": 0.2867395645507891,
      "loss": 0.2078,
      "num_input_tokens_seen": 4677344,
      "step": 5395
    },
    {
      "epoch": 2.545968882602546,
      "grad_norm": 0.0020535781513899565,
      "learning_rate": 0.2867153391129842,
      "loss": 0.2029,
      "num_input_tokens_seen": 4681664,
      "step": 5400
    },
    {
      "epoch": 2.545968882602546,
      "eval_loss": 0.2191556841135025,
      "eval_runtime": 21.8946,
      "eval_samples_per_second": 43.07,
      "eval_steps_per_second": 21.558,
      "num_input_tokens_seen": 4681664,
      "step": 5400
    },
    {
      "epoch": 2.5483262611975483,
      "grad_norm": 0.0016599649097770452,
      "learning_rate": 0.28669109259195585,
      "loss": 0.2069,
      "num_input_tokens_seen": 4685776,
      "step": 5405
    },
    {
      "epoch": 2.5506836397925507,
      "grad_norm": 0.0031225637067109346,
      "learning_rate": 0.2866668249914433,
      "loss": 0.1904,
      "num_input_tokens_seen": 4690416,
      "step": 5410
    },
    {
      "epoch": 2.553041018387553,
      "grad_norm": 0.0018391337944194674,
      "learning_rate": 0.2866425363151889,
      "loss": 0.1837,
      "num_input_tokens_seen": 4695328,
      "step": 5415
    },
    {
      "epoch": 2.5553983969825556,
      "grad_norm": 0.0030899751000106335,
      "learning_rate": 0.2866182265669382,
      "loss": 0.2126,
      "num_input_tokens_seen": 4698960,
      "step": 5420
    },
    {
      "epoch": 2.557755775577558,
      "grad_norm": 0.0010326363844797015,
      "learning_rate": 0.28659389575044014,
      "loss": 0.1468,
      "num_input_tokens_seen": 4703024,
      "step": 5425
    },
    {
      "epoch": 2.5601131541725604,
      "grad_norm": 0.0011398557107895613,
      "learning_rate": 0.28656954386944683,
      "loss": 0.2017,
      "num_input_tokens_seen": 4707040,
      "step": 5430
    },
    {
      "epoch": 2.5624705327675623,
      "grad_norm": 0.002205483615398407,
      "learning_rate": 0.28654517092771353,
      "loss": 0.2757,
      "num_input_tokens_seen": 4710896,
      "step": 5435
    },
    {
      "epoch": 2.5648279113625647,
      "grad_norm": 0.0039968714118003845,
      "learning_rate": 0.286520776928999,
      "loss": 0.1771,
      "num_input_tokens_seen": 4715456,
      "step": 5440
    },
    {
      "epoch": 2.567185289957567,
      "grad_norm": 0.001598479226231575,
      "learning_rate": 0.286496361877065,
      "loss": 0.238,
      "num_input_tokens_seen": 4720128,
      "step": 5445
    },
    {
      "epoch": 2.5695426685525695,
      "grad_norm": 0.004140135832130909,
      "learning_rate": 0.28647192577567676,
      "loss": 0.1952,
      "num_input_tokens_seen": 4725024,
      "step": 5450
    },
    {
      "epoch": 2.571900047147572,
      "grad_norm": 0.0015684018144384027,
      "learning_rate": 0.28644746862860254,
      "loss": 0.2627,
      "num_input_tokens_seen": 4730592,
      "step": 5455
    },
    {
      "epoch": 2.5742574257425743,
      "grad_norm": 0.001649559591896832,
      "learning_rate": 0.2864229904396139,
      "loss": 0.2376,
      "num_input_tokens_seen": 4734640,
      "step": 5460
    },
    {
      "epoch": 2.5766148043375767,
      "grad_norm": 0.001824932754971087,
      "learning_rate": 0.28639849121248573,
      "loss": 0.2111,
      "num_input_tokens_seen": 4738928,
      "step": 5465
    },
    {
      "epoch": 2.578972182932579,
      "grad_norm": 0.0038809129036962986,
      "learning_rate": 0.28637397095099615,
      "loss": 0.2245,
      "num_input_tokens_seen": 4743520,
      "step": 5470
    },
    {
      "epoch": 2.581329561527581,
      "grad_norm": 0.0017663838807493448,
      "learning_rate": 0.28634942965892646,
      "loss": 0.2502,
      "num_input_tokens_seen": 4746976,
      "step": 5475
    },
    {
      "epoch": 2.5836869401225835,
      "grad_norm": 0.0020731070544570684,
      "learning_rate": 0.28632486734006124,
      "loss": 0.2029,
      "num_input_tokens_seen": 4751152,
      "step": 5480
    },
    {
      "epoch": 2.586044318717586,
      "grad_norm": 0.0020378485787659883,
      "learning_rate": 0.28630028399818835,
      "loss": 0.2128,
      "num_input_tokens_seen": 4756624,
      "step": 5485
    },
    {
      "epoch": 2.5884016973125883,
      "grad_norm": 0.004616464953869581,
      "learning_rate": 0.2862756796370987,
      "loss": 0.2073,
      "num_input_tokens_seen": 4760416,
      "step": 5490
    },
    {
      "epoch": 2.5907590759075907,
      "grad_norm": 0.004779658280313015,
      "learning_rate": 0.2862510542605868,
      "loss": 0.232,
      "num_input_tokens_seen": 4765360,
      "step": 5495
    },
    {
      "epoch": 2.593116454502593,
      "grad_norm": 0.0019398464355617762,
      "learning_rate": 0.2862264078724501,
      "loss": 0.176,
      "num_input_tokens_seen": 4769264,
      "step": 5500
    },
    {
      "epoch": 2.5954738330975955,
      "grad_norm": 0.003660802496597171,
      "learning_rate": 0.28620174047648933,
      "loss": 0.2454,
      "num_input_tokens_seen": 4774064,
      "step": 5505
    },
    {
      "epoch": 2.597831211692598,
      "grad_norm": 0.003288082079961896,
      "learning_rate": 0.2861770520765086,
      "loss": 0.2096,
      "num_input_tokens_seen": 4777968,
      "step": 5510
    },
    {
      "epoch": 2.6001885902876003,
      "grad_norm": 0.002040012739598751,
      "learning_rate": 0.2861523426763151,
      "loss": 0.2291,
      "num_input_tokens_seen": 4781952,
      "step": 5515
    },
    {
      "epoch": 2.6025459688826027,
      "grad_norm": 0.0022936842869967222,
      "learning_rate": 0.2861276122797194,
      "loss": 0.2605,
      "num_input_tokens_seen": 4786448,
      "step": 5520
    },
    {
      "epoch": 2.604903347477605,
      "grad_norm": 0.0027759710792452097,
      "learning_rate": 0.28610286089053516,
      "loss": 0.2248,
      "num_input_tokens_seen": 4790512,
      "step": 5525
    },
    {
      "epoch": 2.6072607260726075,
      "grad_norm": 0.0029014230240136385,
      "learning_rate": 0.28607808851257943,
      "loss": 0.2307,
      "num_input_tokens_seen": 4794832,
      "step": 5530
    },
    {
      "epoch": 2.6096181046676095,
      "grad_norm": 0.0020303779747337103,
      "learning_rate": 0.28605329514967237,
      "loss": 0.2046,
      "num_input_tokens_seen": 4798688,
      "step": 5535
    },
    {
      "epoch": 2.611975483262612,
      "grad_norm": 0.005810845643281937,
      "learning_rate": 0.2860284808056374,
      "loss": 0.2725,
      "num_input_tokens_seen": 4802880,
      "step": 5540
    },
    {
      "epoch": 2.6143328618576143,
      "grad_norm": 0.0018301663221791387,
      "learning_rate": 0.28600364548430135,
      "loss": 0.2574,
      "num_input_tokens_seen": 4807776,
      "step": 5545
    },
    {
      "epoch": 2.6166902404526167,
      "grad_norm": 0.0049169957637786865,
      "learning_rate": 0.28597878918949393,
      "loss": 0.2651,
      "num_input_tokens_seen": 4812320,
      "step": 5550
    },
    {
      "epoch": 2.619047619047619,
      "grad_norm": 0.003385751973837614,
      "learning_rate": 0.2859539119250485,
      "loss": 0.2476,
      "num_input_tokens_seen": 4816464,
      "step": 5555
    },
    {
      "epoch": 2.6214049976426215,
      "grad_norm": 0.002661591162905097,
      "learning_rate": 0.2859290136948013,
      "loss": 0.2377,
      "num_input_tokens_seen": 4821168,
      "step": 5560
    },
    {
      "epoch": 2.623762376237624,
      "grad_norm": 0.006644025910645723,
      "learning_rate": 0.28590409450259197,
      "loss": 0.2105,
      "num_input_tokens_seen": 4826176,
      "step": 5565
    },
    {
      "epoch": 2.6261197548326263,
      "grad_norm": 0.003102001966908574,
      "learning_rate": 0.28587915435226346,
      "loss": 0.211,
      "num_input_tokens_seen": 4829856,
      "step": 5570
    },
    {
      "epoch": 2.6284771334276282,
      "grad_norm": 0.0029168615583330393,
      "learning_rate": 0.2858541932476617,
      "loss": 0.1973,
      "num_input_tokens_seen": 4834288,
      "step": 5575
    },
    {
      "epoch": 2.6308345120226306,
      "grad_norm": 0.003176958067342639,
      "learning_rate": 0.2858292111926361,
      "loss": 0.2102,
      "num_input_tokens_seen": 4838704,
      "step": 5580
    },
    {
      "epoch": 2.633191890617633,
      "grad_norm": 0.0017621120205149055,
      "learning_rate": 0.28580420819103924,
      "loss": 0.2216,
      "num_input_tokens_seen": 4843008,
      "step": 5585
    },
    {
      "epoch": 2.6355492692126354,
      "grad_norm": 0.002810442354530096,
      "learning_rate": 0.2857791842467269,
      "loss": 0.238,
      "num_input_tokens_seen": 4847760,
      "step": 5590
    },
    {
      "epoch": 2.637906647807638,
      "grad_norm": 0.0016119919018819928,
      "learning_rate": 0.2857541393635579,
      "loss": 0.2175,
      "num_input_tokens_seen": 4852352,
      "step": 5595
    },
    {
      "epoch": 2.6402640264026402,
      "grad_norm": 0.001806200947612524,
      "learning_rate": 0.2857290735453948,
      "loss": 0.2325,
      "num_input_tokens_seen": 4856928,
      "step": 5600
    },
    {
      "epoch": 2.6402640264026402,
      "eval_loss": 0.21884675323963165,
      "eval_runtime": 21.9152,
      "eval_samples_per_second": 43.03,
      "eval_steps_per_second": 21.538,
      "num_input_tokens_seen": 4856928,
      "step": 5600
    },
    {
      "epoch": 2.6426214049976426,
      "grad_norm": 0.001278447569347918,
      "learning_rate": 0.28570398679610276,
      "loss": 0.1866,
      "num_input_tokens_seen": 4861408,
      "step": 5605
    },
    {
      "epoch": 2.644978783592645,
      "grad_norm": 0.0029608765617012978,
      "learning_rate": 0.2856788791195506,
      "loss": 0.1827,
      "num_input_tokens_seen": 4865600,
      "step": 5610
    },
    {
      "epoch": 2.6473361621876474,
      "grad_norm": 0.0031368094496428967,
      "learning_rate": 0.28565375051961023,
      "loss": 0.2295,
      "num_input_tokens_seen": 4869744,
      "step": 5615
    },
    {
      "epoch": 2.64969354078265,
      "grad_norm": 0.0014416318153962493,
      "learning_rate": 0.28562860100015686,
      "loss": 0.2418,
      "num_input_tokens_seen": 4874448,
      "step": 5620
    },
    {
      "epoch": 2.6520509193776522,
      "grad_norm": 0.0018157470040023327,
      "learning_rate": 0.2856034305650687,
      "loss": 0.2428,
      "num_input_tokens_seen": 4878128,
      "step": 5625
    },
    {
      "epoch": 2.6544082979726547,
      "grad_norm": 0.0031846179626882076,
      "learning_rate": 0.28557823921822756,
      "loss": 0.2381,
      "num_input_tokens_seen": 4882048,
      "step": 5630
    },
    {
      "epoch": 2.6567656765676566,
      "grad_norm": 0.0014706782530993223,
      "learning_rate": 0.2855530269635181,
      "loss": 0.2457,
      "num_input_tokens_seen": 4886432,
      "step": 5635
    },
    {
      "epoch": 2.659123055162659,
      "grad_norm": 0.0021466852631419897,
      "learning_rate": 0.2855277938048284,
      "loss": 0.2219,
      "num_input_tokens_seen": 4891008,
      "step": 5640
    },
    {
      "epoch": 2.6614804337576614,
      "grad_norm": 0.0025439192540943623,
      "learning_rate": 0.2855025397460498,
      "loss": 0.2348,
      "num_input_tokens_seen": 4894960,
      "step": 5645
    },
    {
      "epoch": 2.663837812352664,
      "grad_norm": 0.0031455811113119125,
      "learning_rate": 0.28547726479107666,
      "loss": 0.2245,
      "num_input_tokens_seen": 4899952,
      "step": 5650
    },
    {
      "epoch": 2.666195190947666,
      "grad_norm": 0.0015372470952570438,
      "learning_rate": 0.2854519689438068,
      "loss": 0.2428,
      "num_input_tokens_seen": 4903712,
      "step": 5655
    },
    {
      "epoch": 2.6685525695426686,
      "grad_norm": 0.0034673293121159077,
      "learning_rate": 0.2854266522081412,
      "loss": 0.2161,
      "num_input_tokens_seen": 4907488,
      "step": 5660
    },
    {
      "epoch": 2.670909948137671,
      "grad_norm": 0.0037774250376969576,
      "learning_rate": 0.28540131458798385,
      "loss": 0.2032,
      "num_input_tokens_seen": 4911648,
      "step": 5665
    },
    {
      "epoch": 2.6732673267326734,
      "grad_norm": 0.0016357001150026917,
      "learning_rate": 0.28537595608724226,
      "loss": 0.2255,
      "num_input_tokens_seen": 4916560,
      "step": 5670
    },
    {
      "epoch": 2.6756247053276754,
      "grad_norm": 0.0021887377370148897,
      "learning_rate": 0.28535057670982705,
      "loss": 0.1899,
      "num_input_tokens_seen": 4920608,
      "step": 5675
    },
    {
      "epoch": 2.677982083922678,
      "grad_norm": 0.0022791773080825806,
      "learning_rate": 0.285325176459652,
      "loss": 0.2664,
      "num_input_tokens_seen": 4924672,
      "step": 5680
    },
    {
      "epoch": 2.68033946251768,
      "grad_norm": 0.0015386005397886038,
      "learning_rate": 0.28529975534063406,
      "loss": 0.2085,
      "num_input_tokens_seen": 4928816,
      "step": 5685
    },
    {
      "epoch": 2.6826968411126826,
      "grad_norm": 0.0017380303470417857,
      "learning_rate": 0.2852743133566936,
      "loss": 0.2277,
      "num_input_tokens_seen": 4933728,
      "step": 5690
    },
    {
      "epoch": 2.685054219707685,
      "grad_norm": 0.0013048667460680008,
      "learning_rate": 0.2852488505117541,
      "loss": 0.2238,
      "num_input_tokens_seen": 4937184,
      "step": 5695
    },
    {
      "epoch": 2.6874115983026874,
      "grad_norm": 0.002688546199351549,
      "learning_rate": 0.28522336680974214,
      "loss": 0.2407,
      "num_input_tokens_seen": 4942016,
      "step": 5700
    },
    {
      "epoch": 2.68976897689769,
      "grad_norm": 0.0024779201485216618,
      "learning_rate": 0.2851978622545877,
      "loss": 0.2237,
      "num_input_tokens_seen": 4945632,
      "step": 5705
    },
    {
      "epoch": 2.692126355492692,
      "grad_norm": 0.003268840489909053,
      "learning_rate": 0.285172336850224,
      "loss": 0.2091,
      "num_input_tokens_seen": 4949408,
      "step": 5710
    },
    {
      "epoch": 2.6944837340876946,
      "grad_norm": 0.0023994813673198223,
      "learning_rate": 0.2851467906005871,
      "loss": 0.2774,
      "num_input_tokens_seen": 4953616,
      "step": 5715
    },
    {
      "epoch": 2.696841112682697,
      "grad_norm": 0.0040252176113426685,
      "learning_rate": 0.28512122350961683,
      "loss": 0.2192,
      "num_input_tokens_seen": 4958304,
      "step": 5720
    },
    {
      "epoch": 2.6991984912776994,
      "grad_norm": 0.0016280869022011757,
      "learning_rate": 0.2850956355812559,
      "loss": 0.237,
      "num_input_tokens_seen": 4962576,
      "step": 5725
    },
    {
      "epoch": 2.701555869872702,
      "grad_norm": 0.0020573402289301157,
      "learning_rate": 0.28507002681945015,
      "loss": 0.2291,
      "num_input_tokens_seen": 4966864,
      "step": 5730
    },
    {
      "epoch": 2.7039132484677038,
      "grad_norm": 0.0045327614061534405,
      "learning_rate": 0.28504439722814895,
      "loss": 0.2345,
      "num_input_tokens_seen": 4971024,
      "step": 5735
    },
    {
      "epoch": 2.706270627062706,
      "grad_norm": 0.001511972164735198,
      "learning_rate": 0.28501874681130457,
      "loss": 0.2274,
      "num_input_tokens_seen": 4975392,
      "step": 5740
    },
    {
      "epoch": 2.7086280056577086,
      "grad_norm": 0.0014178297715261579,
      "learning_rate": 0.2849930755728727,
      "loss": 0.222,
      "num_input_tokens_seen": 4979056,
      "step": 5745
    },
    {
      "epoch": 2.710985384252711,
      "grad_norm": 0.0014265916543081403,
      "learning_rate": 0.28496738351681217,
      "loss": 0.2211,
      "num_input_tokens_seen": 4983296,
      "step": 5750
    },
    {
      "epoch": 2.7133427628477134,
      "grad_norm": 0.0016009480459615588,
      "learning_rate": 0.284941670647085,
      "loss": 0.1962,
      "num_input_tokens_seen": 4987456,
      "step": 5755
    },
    {
      "epoch": 2.7157001414427158,
      "grad_norm": 0.0014200917212292552,
      "learning_rate": 0.2849159369676563,
      "loss": 0.2124,
      "num_input_tokens_seen": 4991200,
      "step": 5760
    },
    {
      "epoch": 2.718057520037718,
      "grad_norm": 0.00255211372859776,
      "learning_rate": 0.2848901824824948,
      "loss": 0.1846,
      "num_input_tokens_seen": 4995008,
      "step": 5765
    },
    {
      "epoch": 2.7204148986327206,
      "grad_norm": 0.002506986493244767,
      "learning_rate": 0.284864407195572,
      "loss": 0.2369,
      "num_input_tokens_seen": 4999056,
      "step": 5770
    },
    {
      "epoch": 2.7227722772277225,
      "grad_norm": 0.0012661832151934505,
      "learning_rate": 0.28483861111086284,
      "loss": 0.2409,
      "num_input_tokens_seen": 5003616,
      "step": 5775
    },
    {
      "epoch": 2.725129655822725,
      "grad_norm": 0.0027129659429192543,
      "learning_rate": 0.2848127942323453,
      "loss": 0.2663,
      "num_input_tokens_seen": 5008816,
      "step": 5780
    },
    {
      "epoch": 2.7274870344177273,
      "grad_norm": 0.004870220087468624,
      "learning_rate": 0.2847869565640007,
      "loss": 0.2071,
      "num_input_tokens_seen": 5013008,
      "step": 5785
    },
    {
      "epoch": 2.7298444130127297,
      "grad_norm": 0.0013695033267140388,
      "learning_rate": 0.2847610981098136,
      "loss": 0.2235,
      "num_input_tokens_seen": 5017072,
      "step": 5790
    },
    {
      "epoch": 2.732201791607732,
      "grad_norm": 0.003665443742647767,
      "learning_rate": 0.2847352188737716,
      "loss": 0.2283,
      "num_input_tokens_seen": 5021024,
      "step": 5795
    },
    {
      "epoch": 2.7345591702027345,
      "grad_norm": 0.0037875790148973465,
      "learning_rate": 0.2847093188598658,
      "loss": 0.2387,
      "num_input_tokens_seen": 5024976,
      "step": 5800
    },
    {
      "epoch": 2.7345591702027345,
      "eval_loss": 0.21710790693759918,
      "eval_runtime": 21.9232,
      "eval_samples_per_second": 43.014,
      "eval_steps_per_second": 21.53,
      "num_input_tokens_seen": 5024976,
      "step": 5800
    },
    {
      "epoch": 2.736916548797737,
      "grad_norm": 0.0016749076312407851,
      "learning_rate": 0.28468339807209003,
      "loss": 0.2337,
      "num_input_tokens_seen": 5030048,
      "step": 5805
    },
    {
      "epoch": 2.7392739273927393,
      "grad_norm": 0.002567852148786187,
      "learning_rate": 0.2846574565144418,
      "loss": 0.2268,
      "num_input_tokens_seen": 5034192,
      "step": 5810
    },
    {
      "epoch": 2.7416313059877417,
      "grad_norm": 0.002483132528141141,
      "learning_rate": 0.28463149419092154,
      "loss": 0.2259,
      "num_input_tokens_seen": 5037888,
      "step": 5815
    },
    {
      "epoch": 2.743988684582744,
      "grad_norm": 0.0014428104041144252,
      "learning_rate": 0.284605511105533,
      "loss": 0.2336,
      "num_input_tokens_seen": 5042496,
      "step": 5820
    },
    {
      "epoch": 2.7463460631777465,
      "grad_norm": 0.0015334381023421884,
      "learning_rate": 0.28457950726228315,
      "loss": 0.1897,
      "num_input_tokens_seen": 5047424,
      "step": 5825
    },
    {
      "epoch": 2.748703441772749,
      "grad_norm": 0.0014513073256239295,
      "learning_rate": 0.28455348266518193,
      "loss": 0.1898,
      "num_input_tokens_seen": 5051536,
      "step": 5830
    },
    {
      "epoch": 2.751060820367751,
      "grad_norm": 0.004799570422619581,
      "learning_rate": 0.28452743731824287,
      "loss": 0.2929,
      "num_input_tokens_seen": 5055648,
      "step": 5835
    },
    {
      "epoch": 2.7534181989627533,
      "grad_norm": 0.0016081944340839982,
      "learning_rate": 0.28450137122548236,
      "loss": 0.1953,
      "num_input_tokens_seen": 5060176,
      "step": 5840
    },
    {
      "epoch": 2.7557755775577557,
      "grad_norm": 0.0018668733537197113,
      "learning_rate": 0.2844752843909201,
      "loss": 0.2237,
      "num_input_tokens_seen": 5065312,
      "step": 5845
    },
    {
      "epoch": 2.758132956152758,
      "grad_norm": 0.0045110383071005344,
      "learning_rate": 0.28444917681857923,
      "loss": 0.2221,
      "num_input_tokens_seen": 5069712,
      "step": 5850
    },
    {
      "epoch": 2.7604903347477605,
      "grad_norm": 0.0019985593389719725,
      "learning_rate": 0.28442304851248557,
      "loss": 0.2423,
      "num_input_tokens_seen": 5073936,
      "step": 5855
    },
    {
      "epoch": 2.762847713342763,
      "grad_norm": 0.004507777281105518,
      "learning_rate": 0.2843968994766686,
      "loss": 0.2105,
      "num_input_tokens_seen": 5078112,
      "step": 5860
    },
    {
      "epoch": 2.7652050919377653,
      "grad_norm": 0.0049678729847073555,
      "learning_rate": 0.28437072971516075,
      "loss": 0.2275,
      "num_input_tokens_seen": 5082000,
      "step": 5865
    },
    {
      "epoch": 2.7675624705327677,
      "grad_norm": 0.003286338411271572,
      "learning_rate": 0.2843445392319979,
      "loss": 0.1998,
      "num_input_tokens_seen": 5086064,
      "step": 5870
    },
    {
      "epoch": 2.7699198491277697,
      "grad_norm": 0.0027955747209489346,
      "learning_rate": 0.28431832803121865,
      "loss": 0.2615,
      "num_input_tokens_seen": 5090256,
      "step": 5875
    },
    {
      "epoch": 2.772277227722772,
      "grad_norm": 0.0017423109384253621,
      "learning_rate": 0.28429209611686534,
      "loss": 0.232,
      "num_input_tokens_seen": 5094880,
      "step": 5880
    },
    {
      "epoch": 2.7746346063177745,
      "grad_norm": 0.0011339137563481927,
      "learning_rate": 0.28426584349298323,
      "loss": 0.2358,
      "num_input_tokens_seen": 5099136,
      "step": 5885
    },
    {
      "epoch": 2.776991984912777,
      "grad_norm": 0.0017370324349030852,
      "learning_rate": 0.2842395701636207,
      "loss": 0.2199,
      "num_input_tokens_seen": 5103152,
      "step": 5890
    },
    {
      "epoch": 2.7793493635077793,
      "grad_norm": 0.00155707448720932,
      "learning_rate": 0.28421327613282954,
      "loss": 0.2313,
      "num_input_tokens_seen": 5107776,
      "step": 5895
    },
    {
      "epoch": 2.7817067421027817,
      "grad_norm": 0.0020146346651017666,
      "learning_rate": 0.28418696140466454,
      "loss": 0.2392,
      "num_input_tokens_seen": 5112480,
      "step": 5900
    },
    {
      "epoch": 2.784064120697784,
      "grad_norm": 0.0038995950017124414,
      "learning_rate": 0.2841606259831838,
      "loss": 0.2261,
      "num_input_tokens_seen": 5116672,
      "step": 5905
    },
    {
      "epoch": 2.7864214992927865,
      "grad_norm": 0.003765585832297802,
      "learning_rate": 0.2841342698724486,
      "loss": 0.225,
      "num_input_tokens_seen": 5121072,
      "step": 5910
    },
    {
      "epoch": 2.788778877887789,
      "grad_norm": 0.004106113687157631,
      "learning_rate": 0.28410789307652334,
      "loss": 0.2476,
      "num_input_tokens_seen": 5125568,
      "step": 5915
    },
    {
      "epoch": 2.7911362564827913,
      "grad_norm": 0.002156180329620838,
      "learning_rate": 0.2840814955994756,
      "loss": 0.2297,
      "num_input_tokens_seen": 5130560,
      "step": 5920
    },
    {
      "epoch": 2.7934936350777937,
      "grad_norm": 0.004338707309216261,
      "learning_rate": 0.2840550774453763,
      "loss": 0.2204,
      "num_input_tokens_seen": 5134976,
      "step": 5925
    },
    {
      "epoch": 2.795851013672796,
      "grad_norm": 0.0024962443858385086,
      "learning_rate": 0.28402863861829947,
      "loss": 0.2124,
      "num_input_tokens_seen": 5139072,
      "step": 5930
    },
    {
      "epoch": 2.798208392267798,
      "grad_norm": 0.0019814539700746536,
      "learning_rate": 0.2840021791223222,
      "loss": 0.2309,
      "num_input_tokens_seen": 5143776,
      "step": 5935
    },
    {
      "epoch": 2.8005657708628005,
      "grad_norm": 0.003615087131038308,
      "learning_rate": 0.2839756989615249,
      "loss": 0.2338,
      "num_input_tokens_seen": 5148080,
      "step": 5940
    },
    {
      "epoch": 2.802923149457803,
      "grad_norm": 0.0017136948881670833,
      "learning_rate": 0.28394919813999125,
      "loss": 0.2026,
      "num_input_tokens_seen": 5153120,
      "step": 5945
    },
    {
      "epoch": 2.8052805280528053,
      "grad_norm": 0.0014885151758790016,
      "learning_rate": 0.28392267666180787,
      "loss": 0.2032,
      "num_input_tokens_seen": 5157648,
      "step": 5950
    },
    {
      "epoch": 2.8076379066478077,
      "grad_norm": 0.0025144980754703283,
      "learning_rate": 0.2838961345310648,
      "loss": 0.2391,
      "num_input_tokens_seen": 5162448,
      "step": 5955
    },
    {
      "epoch": 2.80999528524281,
      "grad_norm": 0.002820675726979971,
      "learning_rate": 0.2838695717518552,
      "loss": 0.1649,
      "num_input_tokens_seen": 5167216,
      "step": 5960
    },
    {
      "epoch": 2.8123526638378125,
      "grad_norm": 0.0013507091207429767,
      "learning_rate": 0.28384298832827526,
      "loss": 0.2027,
      "num_input_tokens_seen": 5171568,
      "step": 5965
    },
    {
      "epoch": 2.814710042432815,
      "grad_norm": 0.0037363776937127113,
      "learning_rate": 0.28381638426442457,
      "loss": 0.2122,
      "num_input_tokens_seen": 5175888,
      "step": 5970
    },
    {
      "epoch": 2.817067421027817,
      "grad_norm": 0.005171388387680054,
      "learning_rate": 0.2837897595644057,
      "loss": 0.2315,
      "num_input_tokens_seen": 5179856,
      "step": 5975
    },
    {
      "epoch": 2.8194247996228192,
      "grad_norm": 0.0018650018610060215,
      "learning_rate": 0.28376311423232475,
      "loss": 0.2499,
      "num_input_tokens_seen": 5184144,
      "step": 5980
    },
    {
      "epoch": 2.8217821782178216,
      "grad_norm": 0.004200541414320469,
      "learning_rate": 0.2837364482722905,
      "loss": 0.2283,
      "num_input_tokens_seen": 5189376,
      "step": 5985
    },
    {
      "epoch": 2.824139556812824,
      "grad_norm": 0.0052624656818807125,
      "learning_rate": 0.28370976168841533,
      "loss": 0.2458,
      "num_input_tokens_seen": 5193568,
      "step": 5990
    },
    {
      "epoch": 2.8264969354078264,
      "grad_norm": 0.004971596412360668,
      "learning_rate": 0.2836830544848146,
      "loss": 0.2204,
      "num_input_tokens_seen": 5198368,
      "step": 5995
    },
    {
      "epoch": 2.828854314002829,
      "grad_norm": 0.0026749190874397755,
      "learning_rate": 0.2836563266656069,
      "loss": 0.3,
      "num_input_tokens_seen": 5202368,
      "step": 6000
    },
    {
      "epoch": 2.828854314002829,
      "eval_loss": 0.2189222127199173,
      "eval_runtime": 21.9478,
      "eval_samples_per_second": 42.966,
      "eval_steps_per_second": 21.506,
      "num_input_tokens_seen": 5202368,
      "step": 6000
    },
    {
      "epoch": 2.8312116925978312,
      "grad_norm": 0.003617185400798917,
      "learning_rate": 0.283629578234914,
      "loss": 0.1939,
      "num_input_tokens_seen": 5206368,
      "step": 6005
    },
    {
      "epoch": 2.8335690711928336,
      "grad_norm": 0.0013957360060885549,
      "learning_rate": 0.2836028091968608,
      "loss": 0.1953,
      "num_input_tokens_seen": 5210160,
      "step": 6010
    },
    {
      "epoch": 2.835926449787836,
      "grad_norm": 0.001142903114669025,
      "learning_rate": 0.28357601955557554,
      "loss": 0.1664,
      "num_input_tokens_seen": 5214704,
      "step": 6015
    },
    {
      "epoch": 2.8382838283828384,
      "grad_norm": 0.00453013414517045,
      "learning_rate": 0.2835492093151894,
      "loss": 0.287,
      "num_input_tokens_seen": 5218896,
      "step": 6020
    },
    {
      "epoch": 2.840641206977841,
      "grad_norm": 0.001208429574035108,
      "learning_rate": 0.2835223784798369,
      "loss": 0.1667,
      "num_input_tokens_seen": 5223232,
      "step": 6025
    },
    {
      "epoch": 2.8429985855728432,
      "grad_norm": 0.0019413250265643,
      "learning_rate": 0.2834955270536557,
      "loss": 0.253,
      "num_input_tokens_seen": 5227536,
      "step": 6030
    },
    {
      "epoch": 2.845355964167845,
      "grad_norm": 0.001728573814034462,
      "learning_rate": 0.2834686550407866,
      "loss": 0.2015,
      "num_input_tokens_seen": 5232512,
      "step": 6035
    },
    {
      "epoch": 2.8477133427628476,
      "grad_norm": 0.0021283936221152544,
      "learning_rate": 0.28344176244537367,
      "loss": 0.199,
      "num_input_tokens_seen": 5237472,
      "step": 6040
    },
    {
      "epoch": 2.85007072135785,
      "grad_norm": 0.0024944115430116653,
      "learning_rate": 0.28341484927156396,
      "loss": 0.2717,
      "num_input_tokens_seen": 5241904,
      "step": 6045
    },
    {
      "epoch": 2.8524280999528524,
      "grad_norm": 0.001851448556408286,
      "learning_rate": 0.28338791552350795,
      "loss": 0.2384,
      "num_input_tokens_seen": 5245712,
      "step": 6050
    },
    {
      "epoch": 2.854785478547855,
      "grad_norm": 0.0036855118814855814,
      "learning_rate": 0.28336096120535914,
      "loss": 0.2067,
      "num_input_tokens_seen": 5250704,
      "step": 6055
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.001916488166898489,
      "learning_rate": 0.2833339863212741,
      "loss": 0.2289,
      "num_input_tokens_seen": 5255056,
      "step": 6060
    },
    {
      "epoch": 2.8595002357378596,
      "grad_norm": 0.0038230468053370714,
      "learning_rate": 0.28330699087541283,
      "loss": 0.2099,
      "num_input_tokens_seen": 5259792,
      "step": 6065
    },
    {
      "epoch": 2.861857614332862,
      "grad_norm": 0.0024399685207754374,
      "learning_rate": 0.2832799748719384,
      "loss": 0.2364,
      "num_input_tokens_seen": 5263968,
      "step": 6070
    },
    {
      "epoch": 2.864214992927864,
      "grad_norm": 0.0018647192046046257,
      "learning_rate": 0.28325293831501686,
      "loss": 0.2164,
      "num_input_tokens_seen": 5268112,
      "step": 6075
    },
    {
      "epoch": 2.8665723715228664,
      "grad_norm": 0.002013539196923375,
      "learning_rate": 0.2832258812088177,
      "loss": 0.2359,
      "num_input_tokens_seen": 5272576,
      "step": 6080
    },
    {
      "epoch": 2.8689297501178688,
      "grad_norm": 0.0029203263111412525,
      "learning_rate": 0.2831988035575134,
      "loss": 0.234,
      "num_input_tokens_seen": 5276608,
      "step": 6085
    },
    {
      "epoch": 2.871287128712871,
      "grad_norm": 0.0020659794099628925,
      "learning_rate": 0.28317170536527975,
      "loss": 0.2156,
      "num_input_tokens_seen": 5281344,
      "step": 6090
    },
    {
      "epoch": 2.8736445073078736,
      "grad_norm": 0.0035701135639101267,
      "learning_rate": 0.2831445866362956,
      "loss": 0.2057,
      "num_input_tokens_seen": 5285088,
      "step": 6095
    },
    {
      "epoch": 2.876001885902876,
      "grad_norm": 0.0022322030272334814,
      "learning_rate": 0.2831174473747429,
      "loss": 0.1958,
      "num_input_tokens_seen": 5289648,
      "step": 6100
    },
    {
      "epoch": 2.8783592644978784,
      "grad_norm": 0.005428030155599117,
      "learning_rate": 0.2830902875848071,
      "loss": 0.2557,
      "num_input_tokens_seen": 5294688,
      "step": 6105
    },
    {
      "epoch": 2.880716643092881,
      "grad_norm": 0.0017749059479683638,
      "learning_rate": 0.28306310727067635,
      "loss": 0.2181,
      "num_input_tokens_seen": 5299264,
      "step": 6110
    },
    {
      "epoch": 2.883074021687883,
      "grad_norm": 0.0055986144579946995,
      "learning_rate": 0.2830359064365423,
      "loss": 0.2958,
      "num_input_tokens_seen": 5303664,
      "step": 6115
    },
    {
      "epoch": 2.8854314002828856,
      "grad_norm": 0.002622120315209031,
      "learning_rate": 0.28300868508659965,
      "loss": 0.2345,
      "num_input_tokens_seen": 5308448,
      "step": 6120
    },
    {
      "epoch": 2.887788778877888,
      "grad_norm": 0.0019304427551105618,
      "learning_rate": 0.28298144322504626,
      "loss": 0.2325,
      "num_input_tokens_seen": 5312080,
      "step": 6125
    },
    {
      "epoch": 2.8901461574728904,
      "grad_norm": 0.0015272133750841022,
      "learning_rate": 0.2829541808560832,
      "loss": 0.2294,
      "num_input_tokens_seen": 5315936,
      "step": 6130
    },
    {
      "epoch": 2.8925035360678923,
      "grad_norm": 0.0020525394938886166,
      "learning_rate": 0.2829268979839146,
      "loss": 0.2345,
      "num_input_tokens_seen": 5320640,
      "step": 6135
    },
    {
      "epoch": 2.8948609146628947,
      "grad_norm": 0.001981817651540041,
      "learning_rate": 0.2828995946127479,
      "loss": 0.2238,
      "num_input_tokens_seen": 5324816,
      "step": 6140
    },
    {
      "epoch": 2.897218293257897,
      "grad_norm": 0.002629582304507494,
      "learning_rate": 0.2828722707467936,
      "loss": 0.2301,
      "num_input_tokens_seen": 5328512,
      "step": 6145
    },
    {
      "epoch": 2.8995756718528995,
      "grad_norm": 0.0041525582782924175,
      "learning_rate": 0.2828449263902653,
      "loss": 0.2219,
      "num_input_tokens_seen": 5332848,
      "step": 6150
    },
    {
      "epoch": 2.901933050447902,
      "grad_norm": 0.0021330085583031178,
      "learning_rate": 0.28281756154738,
      "loss": 0.2268,
      "num_input_tokens_seen": 5337920,
      "step": 6155
    },
    {
      "epoch": 2.9042904290429044,
      "grad_norm": 0.0022140624932944775,
      "learning_rate": 0.28279017622235764,
      "loss": 0.205,
      "num_input_tokens_seen": 5342432,
      "step": 6160
    },
    {
      "epoch": 2.9066478076379068,
      "grad_norm": 0.002364426851272583,
      "learning_rate": 0.28276277041942127,
      "loss": 0.2196,
      "num_input_tokens_seen": 5346464,
      "step": 6165
    },
    {
      "epoch": 2.909005186232909,
      "grad_norm": 0.0017965259030461311,
      "learning_rate": 0.2827353441427974,
      "loss": 0.256,
      "num_input_tokens_seen": 5350944,
      "step": 6170
    },
    {
      "epoch": 2.911362564827911,
      "grad_norm": 0.002053775591775775,
      "learning_rate": 0.2827078973967153,
      "loss": 0.1755,
      "num_input_tokens_seen": 5354912,
      "step": 6175
    },
    {
      "epoch": 2.9137199434229135,
      "grad_norm": 0.0017593017546460032,
      "learning_rate": 0.2826804301854078,
      "loss": 0.2561,
      "num_input_tokens_seen": 5359456,
      "step": 6180
    },
    {
      "epoch": 2.916077322017916,
      "grad_norm": 0.0038356217555701733,
      "learning_rate": 0.2826529425131105,
      "loss": 0.2129,
      "num_input_tokens_seen": 5364336,
      "step": 6185
    },
    {
      "epoch": 2.9184347006129183,
      "grad_norm": 0.003234922420233488,
      "learning_rate": 0.2826254343840625,
      "loss": 0.2141,
      "num_input_tokens_seen": 5368416,
      "step": 6190
    },
    {
      "epoch": 2.9207920792079207,
      "grad_norm": 0.0017645374173298478,
      "learning_rate": 0.2825979058025059,
      "loss": 0.2454,
      "num_input_tokens_seen": 5372848,
      "step": 6195
    },
    {
      "epoch": 2.923149457802923,
      "grad_norm": 0.0019703989382833242,
      "learning_rate": 0.2825703567726858,
      "loss": 0.2308,
      "num_input_tokens_seen": 5377360,
      "step": 6200
    },
    {
      "epoch": 2.923149457802923,
      "eval_loss": 0.24366508424282074,
      "eval_runtime": 21.9044,
      "eval_samples_per_second": 43.051,
      "eval_steps_per_second": 21.548,
      "num_input_tokens_seen": 5377360,
      "step": 6200
    },
    {
      "epoch": 2.9255068363979255,
      "grad_norm": 0.002375092590227723,
      "learning_rate": 0.2825427872988508,
      "loss": 0.248,
      "num_input_tokens_seen": 5381136,
      "step": 6205
    },
    {
      "epoch": 2.927864214992928,
      "grad_norm": 0.003724662121385336,
      "learning_rate": 0.28251519738525227,
      "loss": 0.2553,
      "num_input_tokens_seen": 5385296,
      "step": 6210
    },
    {
      "epoch": 2.9302215935879303,
      "grad_norm": 0.0024688835255801678,
      "learning_rate": 0.28248758703614507,
      "loss": 0.2244,
      "num_input_tokens_seen": 5389728,
      "step": 6215
    },
    {
      "epoch": 2.9325789721829327,
      "grad_norm": 0.002140147378668189,
      "learning_rate": 0.28245995625578696,
      "loss": 0.2352,
      "num_input_tokens_seen": 5393984,
      "step": 6220
    },
    {
      "epoch": 2.934936350777935,
      "grad_norm": 0.002051682909950614,
      "learning_rate": 0.282432305048439,
      "loss": 0.222,
      "num_input_tokens_seen": 5398048,
      "step": 6225
    },
    {
      "epoch": 2.9372937293729375,
      "grad_norm": 0.00481885951012373,
      "learning_rate": 0.28240463341836536,
      "loss": 0.2513,
      "num_input_tokens_seen": 5402384,
      "step": 6230
    },
    {
      "epoch": 2.9396511079679395,
      "grad_norm": 0.0016708329785615206,
      "learning_rate": 0.2823769413698334,
      "loss": 0.2155,
      "num_input_tokens_seen": 5406752,
      "step": 6235
    },
    {
      "epoch": 2.942008486562942,
      "grad_norm": 0.002223607385531068,
      "learning_rate": 0.2823492289071135,
      "loss": 0.2241,
      "num_input_tokens_seen": 5411120,
      "step": 6240
    },
    {
      "epoch": 2.9443658651579443,
      "grad_norm": 0.005155738443136215,
      "learning_rate": 0.2823214960344793,
      "loss": 0.2766,
      "num_input_tokens_seen": 5414992,
      "step": 6245
    },
    {
      "epoch": 2.9467232437529467,
      "grad_norm": 0.00252379453741014,
      "learning_rate": 0.28229374275620756,
      "loss": 0.2301,
      "num_input_tokens_seen": 5419808,
      "step": 6250
    },
    {
      "epoch": 2.949080622347949,
      "grad_norm": 0.0022023646160960197,
      "learning_rate": 0.28226596907657814,
      "loss": 0.1918,
      "num_input_tokens_seen": 5423840,
      "step": 6255
    },
    {
      "epoch": 2.9514380009429515,
      "grad_norm": 0.003191194264218211,
      "learning_rate": 0.28223817499987414,
      "loss": 0.2423,
      "num_input_tokens_seen": 5428208,
      "step": 6260
    },
    {
      "epoch": 2.953795379537954,
      "grad_norm": 0.005027868784964085,
      "learning_rate": 0.2822103605303818,
      "loss": 0.2772,
      "num_input_tokens_seen": 5433744,
      "step": 6265
    },
    {
      "epoch": 2.9561527581329563,
      "grad_norm": 0.0029444764368236065,
      "learning_rate": 0.2821825256723903,
      "loss": 0.2364,
      "num_input_tokens_seen": 5437568,
      "step": 6270
    },
    {
      "epoch": 2.9585101367279583,
      "grad_norm": 0.004891206976026297,
      "learning_rate": 0.2821546704301923,
      "loss": 0.2418,
      "num_input_tokens_seen": 5442016,
      "step": 6275
    },
    {
      "epoch": 2.9608675153229607,
      "grad_norm": 0.0017353928415104747,
      "learning_rate": 0.2821267948080834,
      "loss": 0.2242,
      "num_input_tokens_seen": 5445984,
      "step": 6280
    },
    {
      "epoch": 2.963224893917963,
      "grad_norm": 0.002085199346765876,
      "learning_rate": 0.28209889881036226,
      "loss": 0.2513,
      "num_input_tokens_seen": 5450192,
      "step": 6285
    },
    {
      "epoch": 2.9655822725129655,
      "grad_norm": 0.005125412251800299,
      "learning_rate": 0.28207098244133094,
      "loss": 0.2382,
      "num_input_tokens_seen": 5455088,
      "step": 6290
    },
    {
      "epoch": 2.967939651107968,
      "grad_norm": 0.005192582029849291,
      "learning_rate": 0.2820430457052943,
      "loss": 0.252,
      "num_input_tokens_seen": 5459152,
      "step": 6295
    },
    {
      "epoch": 2.9702970297029703,
      "grad_norm": 0.0017111661145463586,
      "learning_rate": 0.28201508860656077,
      "loss": 0.2398,
      "num_input_tokens_seen": 5462448,
      "step": 6300
    },
    {
      "epoch": 2.9726544082979727,
      "grad_norm": 0.004531386308372021,
      "learning_rate": 0.2819871111494415,
      "loss": 0.2472,
      "num_input_tokens_seen": 5466768,
      "step": 6305
    },
    {
      "epoch": 2.975011786892975,
      "grad_norm": 0.001943724462762475,
      "learning_rate": 0.28195911333825113,
      "loss": 0.2381,
      "num_input_tokens_seen": 5470448,
      "step": 6310
    },
    {
      "epoch": 2.9773691654879775,
      "grad_norm": 0.0014728866517543793,
      "learning_rate": 0.28193109517730713,
      "loss": 0.2441,
      "num_input_tokens_seen": 5475104,
      "step": 6315
    },
    {
      "epoch": 2.97972654408298,
      "grad_norm": 0.0012446832843124866,
      "learning_rate": 0.2819030566709303,
      "loss": 0.2283,
      "num_input_tokens_seen": 5479632,
      "step": 6320
    },
    {
      "epoch": 2.9820839226779823,
      "grad_norm": 0.0017083230195567012,
      "learning_rate": 0.2818749978234445,
      "loss": 0.1867,
      "num_input_tokens_seen": 5483776,
      "step": 6325
    },
    {
      "epoch": 2.9844413012729847,
      "grad_norm": 0.0021480051800608635,
      "learning_rate": 0.2818469186391768,
      "loss": 0.26,
      "num_input_tokens_seen": 5487952,
      "step": 6330
    },
    {
      "epoch": 2.9867986798679866,
      "grad_norm": 0.0018338409718126059,
      "learning_rate": 0.28181881912245743,
      "loss": 0.2199,
      "num_input_tokens_seen": 5491840,
      "step": 6335
    },
    {
      "epoch": 2.989156058462989,
      "grad_norm": 0.001991529716178775,
      "learning_rate": 0.2817906992776195,
      "loss": 0.236,
      "num_input_tokens_seen": 5496976,
      "step": 6340
    },
    {
      "epoch": 2.9915134370579914,
      "grad_norm": 0.002213717671111226,
      "learning_rate": 0.28176255910899967,
      "loss": 0.226,
      "num_input_tokens_seen": 5500800,
      "step": 6345
    },
    {
      "epoch": 2.993870815652994,
      "grad_norm": 0.0012342786649242043,
      "learning_rate": 0.2817343986209373,
      "loss": 0.2369,
      "num_input_tokens_seen": 5504688,
      "step": 6350
    },
    {
      "epoch": 2.9962281942479962,
      "grad_norm": 0.002489026403054595,
      "learning_rate": 0.2817062178177753,
      "loss": 0.2348,
      "num_input_tokens_seen": 5508864,
      "step": 6355
    },
    {
      "epoch": 2.9985855728429986,
      "grad_norm": 0.002911111107096076,
      "learning_rate": 0.2816780167038593,
      "loss": 0.2398,
      "num_input_tokens_seen": 5512784,
      "step": 6360
    },
    {
      "epoch": 3.000942951438001,
      "grad_norm": 0.002814113162457943,
      "learning_rate": 0.28164979528353834,
      "loss": 0.2374,
      "num_input_tokens_seen": 5516928,
      "step": 6365
    },
    {
      "epoch": 3.0033003300330035,
      "grad_norm": 0.0022455279249697924,
      "learning_rate": 0.28162155356116453,
      "loss": 0.2349,
      "num_input_tokens_seen": 5521024,
      "step": 6370
    },
    {
      "epoch": 3.005657708628006,
      "grad_norm": 0.006765482947230339,
      "learning_rate": 0.28159329154109314,
      "loss": 0.2362,
      "num_input_tokens_seen": 5526416,
      "step": 6375
    },
    {
      "epoch": 3.008015087223008,
      "grad_norm": 0.0020693691913038492,
      "learning_rate": 0.28156500922768246,
      "loss": 0.2298,
      "num_input_tokens_seen": 5531440,
      "step": 6380
    },
    {
      "epoch": 3.01037246581801,
      "grad_norm": 0.0027854018844664097,
      "learning_rate": 0.28153670662529406,
      "loss": 0.2256,
      "num_input_tokens_seen": 5535920,
      "step": 6385
    },
    {
      "epoch": 3.0127298444130126,
      "grad_norm": 0.003534893272444606,
      "learning_rate": 0.28150838373829246,
      "loss": 0.2172,
      "num_input_tokens_seen": 5540432,
      "step": 6390
    },
    {
      "epoch": 3.015087223008015,
      "grad_norm": 0.0021791390608996153,
      "learning_rate": 0.2814800405710455,
      "loss": 0.2345,
      "num_input_tokens_seen": 5546240,
      "step": 6395
    },
    {
      "epoch": 3.0174446016030174,
      "grad_norm": 0.0028235982172191143,
      "learning_rate": 0.2814516771279239,
      "loss": 0.2332,
      "num_input_tokens_seen": 5550480,
      "step": 6400
    },
    {
      "epoch": 3.0174446016030174,
      "eval_loss": 0.21977819502353668,
      "eval_runtime": 21.8991,
      "eval_samples_per_second": 43.061,
      "eval_steps_per_second": 21.553,
      "num_input_tokens_seen": 5550480,
      "step": 6400
    },
    {
      "epoch": 3.01980198019802,
      "grad_norm": 0.004244680982083082,
      "learning_rate": 0.28142329341330186,
      "loss": 0.2365,
      "num_input_tokens_seen": 5554144,
      "step": 6405
    },
    {
      "epoch": 3.022159358793022,
      "grad_norm": 0.0022190907038748264,
      "learning_rate": 0.2813948894315564,
      "loss": 0.2199,
      "num_input_tokens_seen": 5558128,
      "step": 6410
    },
    {
      "epoch": 3.0245167373880246,
      "grad_norm": 0.0020653260871767998,
      "learning_rate": 0.2813664651870677,
      "loss": 0.2466,
      "num_input_tokens_seen": 5562896,
      "step": 6415
    },
    {
      "epoch": 3.026874115983027,
      "grad_norm": 0.0012289370642974973,
      "learning_rate": 0.28133802068421926,
      "loss": 0.2336,
      "num_input_tokens_seen": 5567184,
      "step": 6420
    },
    {
      "epoch": 3.0292314945780294,
      "grad_norm": 0.0032292315736413,
      "learning_rate": 0.28130955592739754,
      "loss": 0.2127,
      "num_input_tokens_seen": 5571984,
      "step": 6425
    },
    {
      "epoch": 3.0315888731730314,
      "grad_norm": 0.0021824019495397806,
      "learning_rate": 0.2812810709209922,
      "loss": 0.2302,
      "num_input_tokens_seen": 5576528,
      "step": 6430
    },
    {
      "epoch": 3.033946251768034,
      "grad_norm": 0.0033686519600450993,
      "learning_rate": 0.2812525656693959,
      "loss": 0.2384,
      "num_input_tokens_seen": 5581632,
      "step": 6435
    },
    {
      "epoch": 3.036303630363036,
      "grad_norm": 0.002757749520242214,
      "learning_rate": 0.28122404017700453,
      "loss": 0.2092,
      "num_input_tokens_seen": 5586720,
      "step": 6440
    },
    {
      "epoch": 3.0386610089580386,
      "grad_norm": 0.0032411690335720778,
      "learning_rate": 0.2811954944482171,
      "loss": 0.1811,
      "num_input_tokens_seen": 5590256,
      "step": 6445
    },
    {
      "epoch": 3.041018387553041,
      "grad_norm": 0.0028251393232494593,
      "learning_rate": 0.2811669284874358,
      "loss": 0.1974,
      "num_input_tokens_seen": 5594912,
      "step": 6450
    },
    {
      "epoch": 3.0433757661480434,
      "grad_norm": 0.00428414810448885,
      "learning_rate": 0.2811383422990657,
      "loss": 0.2046,
      "num_input_tokens_seen": 5599248,
      "step": 6455
    },
    {
      "epoch": 3.045733144743046,
      "grad_norm": 0.00322744925506413,
      "learning_rate": 0.2811097358875152,
      "loss": 0.2642,
      "num_input_tokens_seen": 5603840,
      "step": 6460
    },
    {
      "epoch": 3.048090523338048,
      "grad_norm": 0.007903686724603176,
      "learning_rate": 0.2810811092571959,
      "loss": 0.2476,
      "num_input_tokens_seen": 5608688,
      "step": 6465
    },
    {
      "epoch": 3.0504479019330506,
      "grad_norm": 0.003340390743687749,
      "learning_rate": 0.28105246241252224,
      "loss": 0.2277,
      "num_input_tokens_seen": 5612896,
      "step": 6470
    },
    {
      "epoch": 3.052805280528053,
      "grad_norm": 0.004942734260112047,
      "learning_rate": 0.28102379535791194,
      "loss": 0.2363,
      "num_input_tokens_seen": 5617232,
      "step": 6475
    },
    {
      "epoch": 3.055162659123055,
      "grad_norm": 0.0014105342561379075,
      "learning_rate": 0.2809951080977859,
      "loss": 0.2238,
      "num_input_tokens_seen": 5621216,
      "step": 6480
    },
    {
      "epoch": 3.0575200377180574,
      "grad_norm": 0.0012574895517900586,
      "learning_rate": 0.28096640063656797,
      "loss": 0.2254,
      "num_input_tokens_seen": 5625664,
      "step": 6485
    },
    {
      "epoch": 3.0598774163130598,
      "grad_norm": 0.0010611230973154306,
      "learning_rate": 0.2809376729786852,
      "loss": 0.2511,
      "num_input_tokens_seen": 5629472,
      "step": 6490
    },
    {
      "epoch": 3.062234794908062,
      "grad_norm": 0.0030765223782509565,
      "learning_rate": 0.28090892512856785,
      "loss": 0.2279,
      "num_input_tokens_seen": 5633936,
      "step": 6495
    },
    {
      "epoch": 3.0645921735030646,
      "grad_norm": 0.0016410150565207005,
      "learning_rate": 0.2808801570906491,
      "loss": 0.2347,
      "num_input_tokens_seen": 5638112,
      "step": 6500
    },
    {
      "epoch": 3.066949552098067,
      "grad_norm": 0.00168778991792351,
      "learning_rate": 0.2808513688693654,
      "loss": 0.2093,
      "num_input_tokens_seen": 5642112,
      "step": 6505
    },
    {
      "epoch": 3.0693069306930694,
      "grad_norm": 0.00450535025447607,
      "learning_rate": 0.28082256046915627,
      "loss": 0.1896,
      "num_input_tokens_seen": 5646384,
      "step": 6510
    },
    {
      "epoch": 3.0716643092880718,
      "grad_norm": 0.0023907064460217953,
      "learning_rate": 0.28079373189446427,
      "loss": 0.2515,
      "num_input_tokens_seen": 5650912,
      "step": 6515
    },
    {
      "epoch": 3.074021687883074,
      "grad_norm": 0.0010663475841283798,
      "learning_rate": 0.28076488314973513,
      "loss": 0.1886,
      "num_input_tokens_seen": 5655472,
      "step": 6520
    },
    {
      "epoch": 3.0763790664780766,
      "grad_norm": 0.00347104761749506,
      "learning_rate": 0.28073601423941774,
      "loss": 0.2565,
      "num_input_tokens_seen": 5659296,
      "step": 6525
    },
    {
      "epoch": 3.0787364450730785,
      "grad_norm": 0.0025123795494437218,
      "learning_rate": 0.28070712516796403,
      "loss": 0.2324,
      "num_input_tokens_seen": 5663920,
      "step": 6530
    },
    {
      "epoch": 3.081093823668081,
      "grad_norm": 0.002451820531859994,
      "learning_rate": 0.28067821593982906,
      "loss": 0.2412,
      "num_input_tokens_seen": 5667840,
      "step": 6535
    },
    {
      "epoch": 3.0834512022630833,
      "grad_norm": 0.004395333118736744,
      "learning_rate": 0.28064928655947097,
      "loss": 0.2486,
      "num_input_tokens_seen": 5672656,
      "step": 6540
    },
    {
      "epoch": 3.0858085808580857,
      "grad_norm": 0.0016445236979052424,
      "learning_rate": 0.28062033703135103,
      "loss": 0.2204,
      "num_input_tokens_seen": 5677040,
      "step": 6545
    },
    {
      "epoch": 3.088165959453088,
      "grad_norm": 0.0038351030088961124,
      "learning_rate": 0.2805913673599337,
      "loss": 0.2683,
      "num_input_tokens_seen": 5680528,
      "step": 6550
    },
    {
      "epoch": 3.0905233380480905,
      "grad_norm": 0.002283252077177167,
      "learning_rate": 0.2805623775496864,
      "loss": 0.2099,
      "num_input_tokens_seen": 5685248,
      "step": 6555
    },
    {
      "epoch": 3.092880716643093,
      "grad_norm": 0.001359613612294197,
      "learning_rate": 0.2805333676050797,
      "loss": 0.2573,
      "num_input_tokens_seen": 5688976,
      "step": 6560
    },
    {
      "epoch": 3.0952380952380953,
      "grad_norm": 0.0030670214910060167,
      "learning_rate": 0.2805043375305873,
      "loss": 0.2179,
      "num_input_tokens_seen": 5692960,
      "step": 6565
    },
    {
      "epoch": 3.0975954738330977,
      "grad_norm": 0.0019961423240602016,
      "learning_rate": 0.2804752873306861,
      "loss": 0.2336,
      "num_input_tokens_seen": 5697248,
      "step": 6570
    },
    {
      "epoch": 3.0999528524281,
      "grad_norm": 0.004063866566866636,
      "learning_rate": 0.2804462170098559,
      "loss": 0.2256,
      "num_input_tokens_seen": 5701328,
      "step": 6575
    },
    {
      "epoch": 3.102310231023102,
      "grad_norm": 0.002048760186880827,
      "learning_rate": 0.2804171265725797,
      "loss": 0.2063,
      "num_input_tokens_seen": 5706528,
      "step": 6580
    },
    {
      "epoch": 3.1046676096181045,
      "grad_norm": 0.00381287420168519,
      "learning_rate": 0.28038801602334373,
      "loss": 0.2236,
      "num_input_tokens_seen": 5710752,
      "step": 6585
    },
    {
      "epoch": 3.107024988213107,
      "grad_norm": 0.0024239225313067436,
      "learning_rate": 0.28035888536663717,
      "loss": 0.2235,
      "num_input_tokens_seen": 5715520,
      "step": 6590
    },
    {
      "epoch": 3.1093823668081093,
      "grad_norm": 0.0027272114530205727,
      "learning_rate": 0.2803297346069522,
      "loss": 0.2289,
      "num_input_tokens_seen": 5719760,
      "step": 6595
    },
    {
      "epoch": 3.1117397454031117,
      "grad_norm": 0.003628681180998683,
      "learning_rate": 0.28030056374878437,
      "loss": 0.2422,
      "num_input_tokens_seen": 5724080,
      "step": 6600
    },
    {
      "epoch": 3.1117397454031117,
      "eval_loss": 0.22279927134513855,
      "eval_runtime": 21.9125,
      "eval_samples_per_second": 43.035,
      "eval_steps_per_second": 21.54,
      "num_input_tokens_seen": 5724080,
      "step": 6600
    },
    {
      "epoch": 3.114097123998114,
      "grad_norm": 0.001797323115170002,
      "learning_rate": 0.2802713727966321,
      "loss": 0.2068,
      "num_input_tokens_seen": 5728128,
      "step": 6605
    },
    {
      "epoch": 3.1164545025931165,
      "grad_norm": 0.0011620894074440002,
      "learning_rate": 0.28024216175499717,
      "loss": 0.2115,
      "num_input_tokens_seen": 5732576,
      "step": 6610
    },
    {
      "epoch": 3.118811881188119,
      "grad_norm": 0.003845749655738473,
      "learning_rate": 0.2802129306283841,
      "loss": 0.2201,
      "num_input_tokens_seen": 5735888,
      "step": 6615
    },
    {
      "epoch": 3.1211692597831213,
      "grad_norm": 0.0012406290043145418,
      "learning_rate": 0.28018367942130074,
      "loss": 0.2567,
      "num_input_tokens_seen": 5739968,
      "step": 6620
    },
    {
      "epoch": 3.1235266383781237,
      "grad_norm": 0.0015564817003905773,
      "learning_rate": 0.28015440813825804,
      "loss": 0.2291,
      "num_input_tokens_seen": 5743824,
      "step": 6625
    },
    {
      "epoch": 3.1258840169731257,
      "grad_norm": 0.003645627060905099,
      "learning_rate": 0.28012511678377006,
      "loss": 0.2465,
      "num_input_tokens_seen": 5748000,
      "step": 6630
    },
    {
      "epoch": 3.128241395568128,
      "grad_norm": 0.0034321791026741266,
      "learning_rate": 0.28009580536235373,
      "loss": 0.228,
      "num_input_tokens_seen": 5752384,
      "step": 6635
    },
    {
      "epoch": 3.1305987741631305,
      "grad_norm": 0.00308528495952487,
      "learning_rate": 0.28006647387852934,
      "loss": 0.2198,
      "num_input_tokens_seen": 5757216,
      "step": 6640
    },
    {
      "epoch": 3.132956152758133,
      "grad_norm": 0.0016433403361588717,
      "learning_rate": 0.28003712233682015,
      "loss": 0.226,
      "num_input_tokens_seen": 5762368,
      "step": 6645
    },
    {
      "epoch": 3.1353135313531353,
      "grad_norm": 0.0017365470994263887,
      "learning_rate": 0.2800077507417526,
      "loss": 0.222,
      "num_input_tokens_seen": 5766592,
      "step": 6650
    },
    {
      "epoch": 3.1376709099481377,
      "grad_norm": 0.001847127452492714,
      "learning_rate": 0.2799783590978561,
      "loss": 0.2358,
      "num_input_tokens_seen": 5770624,
      "step": 6655
    },
    {
      "epoch": 3.14002828854314,
      "grad_norm": 0.0014438219368457794,
      "learning_rate": 0.2799489474096632,
      "loss": 0.2483,
      "num_input_tokens_seen": 5775296,
      "step": 6660
    },
    {
      "epoch": 3.1423856671381425,
      "grad_norm": 0.0018508339999243617,
      "learning_rate": 0.27991951568170953,
      "loss": 0.23,
      "num_input_tokens_seen": 5779280,
      "step": 6665
    },
    {
      "epoch": 3.144743045733145,
      "grad_norm": 0.003142034634947777,
      "learning_rate": 0.2798900639185339,
      "loss": 0.2174,
      "num_input_tokens_seen": 5783856,
      "step": 6670
    },
    {
      "epoch": 3.1471004243281473,
      "grad_norm": 0.001180998864583671,
      "learning_rate": 0.2798605921246781,
      "loss": 0.1774,
      "num_input_tokens_seen": 5788336,
      "step": 6675
    },
    {
      "epoch": 3.1494578029231493,
      "grad_norm": 0.0009475816041231155,
      "learning_rate": 0.2798311003046871,
      "loss": 0.2162,
      "num_input_tokens_seen": 5793040,
      "step": 6680
    },
    {
      "epoch": 3.1518151815181517,
      "grad_norm": 0.0027856584638357162,
      "learning_rate": 0.2798015884631089,
      "loss": 0.2485,
      "num_input_tokens_seen": 5798576,
      "step": 6685
    },
    {
      "epoch": 3.154172560113154,
      "grad_norm": 0.0022963955998420715,
      "learning_rate": 0.27977205660449445,
      "loss": 0.2207,
      "num_input_tokens_seen": 5803760,
      "step": 6690
    },
    {
      "epoch": 3.1565299387081565,
      "grad_norm": 0.002540385816246271,
      "learning_rate": 0.2797425047333981,
      "loss": 0.2336,
      "num_input_tokens_seen": 5808192,
      "step": 6695
    },
    {
      "epoch": 3.158887317303159,
      "grad_norm": 0.0018690063152462244,
      "learning_rate": 0.27971293285437715,
      "loss": 0.2156,
      "num_input_tokens_seen": 5812400,
      "step": 6700
    },
    {
      "epoch": 3.1612446958981613,
      "grad_norm": 0.0016668778844177723,
      "learning_rate": 0.2796833409719918,
      "loss": 0.2239,
      "num_input_tokens_seen": 5817216,
      "step": 6705
    },
    {
      "epoch": 3.1636020744931637,
      "grad_norm": 0.0016467575915157795,
      "learning_rate": 0.27965372909080566,
      "loss": 0.2063,
      "num_input_tokens_seen": 5820944,
      "step": 6710
    },
    {
      "epoch": 3.165959453088166,
      "grad_norm": 0.0010513500310480595,
      "learning_rate": 0.27962409721538506,
      "loss": 0.2492,
      "num_input_tokens_seen": 5824912,
      "step": 6715
    },
    {
      "epoch": 3.1683168316831685,
      "grad_norm": 0.003182976972311735,
      "learning_rate": 0.27959444535029976,
      "loss": 0.2705,
      "num_input_tokens_seen": 5828944,
      "step": 6720
    },
    {
      "epoch": 3.170674210278171,
      "grad_norm": 0.0019462519558146596,
      "learning_rate": 0.27956477350012243,
      "loss": 0.2414,
      "num_input_tokens_seen": 5832608,
      "step": 6725
    },
    {
      "epoch": 3.173031588873173,
      "grad_norm": 0.0011494166683405638,
      "learning_rate": 0.27953508166942875,
      "loss": 0.248,
      "num_input_tokens_seen": 5836608,
      "step": 6730
    },
    {
      "epoch": 3.1753889674681752,
      "grad_norm": 0.0018198189791291952,
      "learning_rate": 0.27950536986279767,
      "loss": 0.2303,
      "num_input_tokens_seen": 5840416,
      "step": 6735
    },
    {
      "epoch": 3.1777463460631776,
      "grad_norm": 0.0017197655979543924,
      "learning_rate": 0.2794756380848111,
      "loss": 0.2412,
      "num_input_tokens_seen": 5844032,
      "step": 6740
    },
    {
      "epoch": 3.18010372465818,
      "grad_norm": 0.0013186497380957007,
      "learning_rate": 0.279445886340054,
      "loss": 0.1994,
      "num_input_tokens_seen": 5848336,
      "step": 6745
    },
    {
      "epoch": 3.1824611032531824,
      "grad_norm": 0.0009659620700404048,
      "learning_rate": 0.27941611463311455,
      "loss": 0.2229,
      "num_input_tokens_seen": 5852976,
      "step": 6750
    },
    {
      "epoch": 3.184818481848185,
      "grad_norm": 0.0025989271234720945,
      "learning_rate": 0.2793863229685839,
      "loss": 0.2748,
      "num_input_tokens_seen": 5857552,
      "step": 6755
    },
    {
      "epoch": 3.1871758604431872,
      "grad_norm": 0.001701919361948967,
      "learning_rate": 0.27935651135105627,
      "loss": 0.2513,
      "num_input_tokens_seen": 5862336,
      "step": 6760
    },
    {
      "epoch": 3.1895332390381896,
      "grad_norm": 0.003486581612378359,
      "learning_rate": 0.279326679785129,
      "loss": 0.2313,
      "num_input_tokens_seen": 5866368,
      "step": 6765
    },
    {
      "epoch": 3.191890617633192,
      "grad_norm": 0.0012486559571698308,
      "learning_rate": 0.2792968282754024,
      "loss": 0.2137,
      "num_input_tokens_seen": 5871600,
      "step": 6770
    },
    {
      "epoch": 3.1942479962281944,
      "grad_norm": 0.002585432957857847,
      "learning_rate": 0.2792669568264801,
      "loss": 0.2432,
      "num_input_tokens_seen": 5876560,
      "step": 6775
    },
    {
      "epoch": 3.1966053748231964,
      "grad_norm": 0.0017338250763714314,
      "learning_rate": 0.27923706544296856,
      "loss": 0.2075,
      "num_input_tokens_seen": 5880256,
      "step": 6780
    },
    {
      "epoch": 3.198962753418199,
      "grad_norm": 0.001195418299175799,
      "learning_rate": 0.2792071541294775,
      "loss": 0.2042,
      "num_input_tokens_seen": 5884096,
      "step": 6785
    },
    {
      "epoch": 3.201320132013201,
      "grad_norm": 0.0014828711282461882,
      "learning_rate": 0.27917722289061947,
      "loss": 0.2303,
      "num_input_tokens_seen": 5888256,
      "step": 6790
    },
    {
      "epoch": 3.2036775106082036,
      "grad_norm": 0.0013843594351783395,
      "learning_rate": 0.27914727173101034,
      "loss": 0.221,
      "num_input_tokens_seen": 5892768,
      "step": 6795
    },
    {
      "epoch": 3.206034889203206,
      "grad_norm": 0.0013452813727781177,
      "learning_rate": 0.279117300655269,
      "loss": 0.2444,
      "num_input_tokens_seen": 5896688,
      "step": 6800
    },
    {
      "epoch": 3.206034889203206,
      "eval_loss": 0.22003114223480225,
      "eval_runtime": 21.9705,
      "eval_samples_per_second": 42.921,
      "eval_steps_per_second": 21.483,
      "num_input_tokens_seen": 5896688,
      "step": 6800
    },
    {
      "epoch": 3.2083922677982084,
      "grad_norm": 0.0028393915854394436,
      "learning_rate": 0.2790873096680173,
      "loss": 0.2168,
      "num_input_tokens_seen": 5901520,
      "step": 6805
    },
    {
      "epoch": 3.210749646393211,
      "grad_norm": 0.0011864913394674659,
      "learning_rate": 0.2790572987738802,
      "loss": 0.2215,
      "num_input_tokens_seen": 5905920,
      "step": 6810
    },
    {
      "epoch": 3.213107024988213,
      "grad_norm": 0.002324012340977788,
      "learning_rate": 0.27902726797748584,
      "loss": 0.1941,
      "num_input_tokens_seen": 5909840,
      "step": 6815
    },
    {
      "epoch": 3.2154644035832156,
      "grad_norm": 0.0019916014280170202,
      "learning_rate": 0.2789972172834652,
      "loss": 0.1844,
      "num_input_tokens_seen": 5914752,
      "step": 6820
    },
    {
      "epoch": 3.217821782178218,
      "grad_norm": 0.002011052565649152,
      "learning_rate": 0.2789671466964527,
      "loss": 0.2439,
      "num_input_tokens_seen": 5919424,
      "step": 6825
    },
    {
      "epoch": 3.22017916077322,
      "grad_norm": 0.0011960356496274471,
      "learning_rate": 0.2789370562210854,
      "loss": 0.2498,
      "num_input_tokens_seen": 5924432,
      "step": 6830
    },
    {
      "epoch": 3.2225365393682224,
      "grad_norm": 0.0015195002779364586,
      "learning_rate": 0.27890694586200376,
      "loss": 0.2208,
      "num_input_tokens_seen": 5928848,
      "step": 6835
    },
    {
      "epoch": 3.2248939179632248,
      "grad_norm": 0.0013107417616993189,
      "learning_rate": 0.2788768156238511,
      "loss": 0.2411,
      "num_input_tokens_seen": 5933152,
      "step": 6840
    },
    {
      "epoch": 3.227251296558227,
      "grad_norm": 0.0017446527490392327,
      "learning_rate": 0.27884666551127385,
      "loss": 0.2343,
      "num_input_tokens_seen": 5937120,
      "step": 6845
    },
    {
      "epoch": 3.2296086751532296,
      "grad_norm": 0.002317015780135989,
      "learning_rate": 0.2788164955289217,
      "loss": 0.2509,
      "num_input_tokens_seen": 5941248,
      "step": 6850
    },
    {
      "epoch": 3.231966053748232,
      "grad_norm": 0.0017115857917815447,
      "learning_rate": 0.27878630568144697,
      "loss": 0.2325,
      "num_input_tokens_seen": 5945472,
      "step": 6855
    },
    {
      "epoch": 3.2343234323432344,
      "grad_norm": 0.0012043580645695329,
      "learning_rate": 0.2787560959735056,
      "loss": 0.2135,
      "num_input_tokens_seen": 5949888,
      "step": 6860
    },
    {
      "epoch": 3.236680810938237,
      "grad_norm": 0.0025454151909798384,
      "learning_rate": 0.27872586640975616,
      "loss": 0.249,
      "num_input_tokens_seen": 5953632,
      "step": 6865
    },
    {
      "epoch": 3.239038189533239,
      "grad_norm": 0.001849035150371492,
      "learning_rate": 0.27869561699486045,
      "loss": 0.2182,
      "num_input_tokens_seen": 5957808,
      "step": 6870
    },
    {
      "epoch": 3.2413955681282416,
      "grad_norm": 0.0020729422103613615,
      "learning_rate": 0.2786653477334833,
      "loss": 0.2328,
      "num_input_tokens_seen": 5961360,
      "step": 6875
    },
    {
      "epoch": 3.2437529467232435,
      "grad_norm": 0.002039064420387149,
      "learning_rate": 0.2786350586302926,
      "loss": 0.2327,
      "num_input_tokens_seen": 5966528,
      "step": 6880
    },
    {
      "epoch": 3.246110325318246,
      "grad_norm": 0.0016803611069917679,
      "learning_rate": 0.27860474968995935,
      "loss": 0.2383,
      "num_input_tokens_seen": 5970400,
      "step": 6885
    },
    {
      "epoch": 3.2484677039132484,
      "grad_norm": 0.0013120239600539207,
      "learning_rate": 0.27857442091715756,
      "loss": 0.2004,
      "num_input_tokens_seen": 5975616,
      "step": 6890
    },
    {
      "epoch": 3.2508250825082508,
      "grad_norm": 0.0033225726801902056,
      "learning_rate": 0.27854407231656425,
      "loss": 0.2458,
      "num_input_tokens_seen": 5979856,
      "step": 6895
    },
    {
      "epoch": 3.253182461103253,
      "grad_norm": 0.04170410707592964,
      "learning_rate": 0.2785137038928596,
      "loss": 0.292,
      "num_input_tokens_seen": 5985008,
      "step": 6900
    },
    {
      "epoch": 3.2555398396982556,
      "grad_norm": 0.001562706078402698,
      "learning_rate": 0.27848331565072687,
      "loss": 0.2531,
      "num_input_tokens_seen": 5990464,
      "step": 6905
    },
    {
      "epoch": 3.257897218293258,
      "grad_norm": 0.001959222834557295,
      "learning_rate": 0.27845290759485225,
      "loss": 0.2153,
      "num_input_tokens_seen": 5995024,
      "step": 6910
    },
    {
      "epoch": 3.2602545968882604,
      "grad_norm": 0.0030264370143413544,
      "learning_rate": 0.278422479729925,
      "loss": 0.2367,
      "num_input_tokens_seen": 5999488,
      "step": 6915
    },
    {
      "epoch": 3.2626119754832628,
      "grad_norm": 0.00212351162917912,
      "learning_rate": 0.2783920320606375,
      "loss": 0.2244,
      "num_input_tokens_seen": 6003536,
      "step": 6920
    },
    {
      "epoch": 3.264969354078265,
      "grad_norm": 0.0015657602343708277,
      "learning_rate": 0.2783615645916852,
      "loss": 0.2367,
      "num_input_tokens_seen": 6008432,
      "step": 6925
    },
    {
      "epoch": 3.2673267326732676,
      "grad_norm": 0.0020443841349333525,
      "learning_rate": 0.2783310773277666,
      "loss": 0.2359,
      "num_input_tokens_seen": 6012016,
      "step": 6930
    },
    {
      "epoch": 3.2696841112682695,
      "grad_norm": 0.0026490509044378996,
      "learning_rate": 0.2783005702735831,
      "loss": 0.2279,
      "num_input_tokens_seen": 6017088,
      "step": 6935
    },
    {
      "epoch": 3.272041489863272,
      "grad_norm": 0.003506440669298172,
      "learning_rate": 0.2782700434338394,
      "loss": 0.2326,
      "num_input_tokens_seen": 6021360,
      "step": 6940
    },
    {
      "epoch": 3.2743988684582743,
      "grad_norm": 0.002338959136977792,
      "learning_rate": 0.278239496813243,
      "loss": 0.2397,
      "num_input_tokens_seen": 6025504,
      "step": 6945
    },
    {
      "epoch": 3.2767562470532767,
      "grad_norm": 0.002607919042930007,
      "learning_rate": 0.27820893041650463,
      "loss": 0.2349,
      "num_input_tokens_seen": 6029824,
      "step": 6950
    },
    {
      "epoch": 3.279113625648279,
      "grad_norm": 0.0027701915241777897,
      "learning_rate": 0.27817834424833804,
      "loss": 0.2246,
      "num_input_tokens_seen": 6034432,
      "step": 6955
    },
    {
      "epoch": 3.2814710042432815,
      "grad_norm": 0.0016291236970573664,
      "learning_rate": 0.27814773831345996,
      "loss": 0.2174,
      "num_input_tokens_seen": 6038448,
      "step": 6960
    },
    {
      "epoch": 3.283828382838284,
      "grad_norm": 0.0015494412509724498,
      "learning_rate": 0.2781171126165902,
      "loss": 0.218,
      "num_input_tokens_seen": 6042256,
      "step": 6965
    },
    {
      "epoch": 3.2861857614332863,
      "grad_norm": 0.0030371190514415503,
      "learning_rate": 0.2780864671624517,
      "loss": 0.2768,
      "num_input_tokens_seen": 6046064,
      "step": 6970
    },
    {
      "epoch": 3.2885431400282887,
      "grad_norm": 0.003019363386556506,
      "learning_rate": 0.27805580195577034,
      "loss": 0.2107,
      "num_input_tokens_seen": 6050064,
      "step": 6975
    },
    {
      "epoch": 3.2909005186232907,
      "grad_norm": 0.00227131019346416,
      "learning_rate": 0.2780251170012751,
      "loss": 0.231,
      "num_input_tokens_seen": 6054176,
      "step": 6980
    },
    {
      "epoch": 3.293257897218293,
      "grad_norm": 0.0017293553100898862,
      "learning_rate": 0.27799441230369787,
      "loss": 0.2244,
      "num_input_tokens_seen": 6058368,
      "step": 6985
    },
    {
      "epoch": 3.2956152758132955,
      "grad_norm": 0.0032269610092043877,
      "learning_rate": 0.27796368786777387,
      "loss": 0.2286,
      "num_input_tokens_seen": 6062272,
      "step": 6990
    },
    {
      "epoch": 3.297972654408298,
      "grad_norm": 0.003913054242730141,
      "learning_rate": 0.277932943698241,
      "loss": 0.2252,
      "num_input_tokens_seen": 6065888,
      "step": 6995
    },
    {
      "epoch": 3.3003300330033003,
      "grad_norm": 0.005522516090422869,
      "learning_rate": 0.2779021797998406,
      "loss": 0.2385,
      "num_input_tokens_seen": 6070544,
      "step": 7000
    },
    {
      "epoch": 3.3003300330033003,
      "eval_loss": 0.21738483011722565,
      "eval_runtime": 21.8656,
      "eval_samples_per_second": 43.127,
      "eval_steps_per_second": 21.586,
      "num_input_tokens_seen": 6070544,
      "step": 7000
    },
    {
      "epoch": 3.3026874115983027,
      "grad_norm": 0.003493984928354621,
      "learning_rate": 0.2778713961773167,
      "loss": 0.2342,
      "num_input_tokens_seen": 6075024,
      "step": 7005
    },
    {
      "epoch": 3.305044790193305,
      "grad_norm": 0.0025782797019928694,
      "learning_rate": 0.2778405928354166,
      "loss": 0.1719,
      "num_input_tokens_seen": 6079120,
      "step": 7010
    },
    {
      "epoch": 3.3074021687883075,
      "grad_norm": 0.0037321806885302067,
      "learning_rate": 0.27780976977889055,
      "loss": 0.2646,
      "num_input_tokens_seen": 6082512,
      "step": 7015
    },
    {
      "epoch": 3.30975954738331,
      "grad_norm": 0.003182155778631568,
      "learning_rate": 0.27777892701249185,
      "loss": 0.2501,
      "num_input_tokens_seen": 6086992,
      "step": 7020
    },
    {
      "epoch": 3.3121169259783123,
      "grad_norm": 0.0037068689707666636,
      "learning_rate": 0.2777480645409768,
      "loss": 0.2501,
      "num_input_tokens_seen": 6091984,
      "step": 7025
    },
    {
      "epoch": 3.3144743045733147,
      "grad_norm": 0.0015096596907824278,
      "learning_rate": 0.27771718236910486,
      "loss": 0.2286,
      "num_input_tokens_seen": 6095632,
      "step": 7030
    },
    {
      "epoch": 3.3168316831683167,
      "grad_norm": 0.0020208507776260376,
      "learning_rate": 0.27768628050163835,
      "loss": 0.2363,
      "num_input_tokens_seen": 6100656,
      "step": 7035
    },
    {
      "epoch": 3.319189061763319,
      "grad_norm": 0.0013854358112439513,
      "learning_rate": 0.2776553589433428,
      "loss": 0.2264,
      "num_input_tokens_seen": 6105248,
      "step": 7040
    },
    {
      "epoch": 3.3215464403583215,
      "grad_norm": 0.0016285768942907453,
      "learning_rate": 0.27762441769898666,
      "loss": 0.2162,
      "num_input_tokens_seen": 6109184,
      "step": 7045
    },
    {
      "epoch": 3.323903818953324,
      "grad_norm": 0.0013114437460899353,
      "learning_rate": 0.2775934567733415,
      "loss": 0.2452,
      "num_input_tokens_seen": 6113696,
      "step": 7050
    },
    {
      "epoch": 3.3262611975483263,
      "grad_norm": 0.0021063259337097406,
      "learning_rate": 0.2775624761711819,
      "loss": 0.2284,
      "num_input_tokens_seen": 6118080,
      "step": 7055
    },
    {
      "epoch": 3.3286185761433287,
      "grad_norm": 0.001402810332365334,
      "learning_rate": 0.2775314758972854,
      "loss": 0.2188,
      "num_input_tokens_seen": 6122576,
      "step": 7060
    },
    {
      "epoch": 3.330975954738331,
      "grad_norm": 0.002076199045404792,
      "learning_rate": 0.2775004559564327,
      "loss": 0.2239,
      "num_input_tokens_seen": 6126816,
      "step": 7065
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.0018991362303495407,
      "learning_rate": 0.2774694163534073,
      "loss": 0.2513,
      "num_input_tokens_seen": 6131184,
      "step": 7070
    },
    {
      "epoch": 3.335690711928336,
      "grad_norm": 0.0017437102505937219,
      "learning_rate": 0.27743835709299614,
      "loss": 0.2232,
      "num_input_tokens_seen": 6135424,
      "step": 7075
    },
    {
      "epoch": 3.338048090523338,
      "grad_norm": 0.0017402650555595756,
      "learning_rate": 0.2774072781799888,
      "loss": 0.2185,
      "num_input_tokens_seen": 6139744,
      "step": 7080
    },
    {
      "epoch": 3.3404054691183402,
      "grad_norm": 0.001483156462199986,
      "learning_rate": 0.27737617961917804,
      "loss": 0.2011,
      "num_input_tokens_seen": 6143584,
      "step": 7085
    },
    {
      "epoch": 3.3427628477133426,
      "grad_norm": 0.0016855536960065365,
      "learning_rate": 0.27734506141535964,
      "loss": 0.2503,
      "num_input_tokens_seen": 6148512,
      "step": 7090
    },
    {
      "epoch": 3.345120226308345,
      "grad_norm": 0.0013885904336348176,
      "learning_rate": 0.2773139235733325,
      "loss": 0.2341,
      "num_input_tokens_seen": 6152928,
      "step": 7095
    },
    {
      "epoch": 3.3474776049033474,
      "grad_norm": 0.0015265572583302855,
      "learning_rate": 0.2772827660978984,
      "loss": 0.2251,
      "num_input_tokens_seen": 6156992,
      "step": 7100
    },
    {
      "epoch": 3.34983498349835,
      "grad_norm": 0.0019305471796542406,
      "learning_rate": 0.27725158899386226,
      "loss": 0.2316,
      "num_input_tokens_seen": 6160880,
      "step": 7105
    },
    {
      "epoch": 3.3521923620933523,
      "grad_norm": 0.0010126088745892048,
      "learning_rate": 0.27722039226603196,
      "loss": 0.2223,
      "num_input_tokens_seen": 6165120,
      "step": 7110
    },
    {
      "epoch": 3.3545497406883547,
      "grad_norm": 0.0018479155842214823,
      "learning_rate": 0.2771891759192184,
      "loss": 0.2329,
      "num_input_tokens_seen": 6169648,
      "step": 7115
    },
    {
      "epoch": 3.356907119283357,
      "grad_norm": 0.002539560664445162,
      "learning_rate": 0.2771579399582355,
      "loss": 0.2213,
      "num_input_tokens_seen": 6173552,
      "step": 7120
    },
    {
      "epoch": 3.3592644978783595,
      "grad_norm": 0.004805716220289469,
      "learning_rate": 0.2771266843879004,
      "loss": 0.2238,
      "num_input_tokens_seen": 6178144,
      "step": 7125
    },
    {
      "epoch": 3.361621876473362,
      "grad_norm": 0.001320306328125298,
      "learning_rate": 0.2770954092130329,
      "loss": 0.2167,
      "num_input_tokens_seen": 6182448,
      "step": 7130
    },
    {
      "epoch": 3.363979255068364,
      "grad_norm": 0.0015767494915053248,
      "learning_rate": 0.27706411443845613,
      "loss": 0.2556,
      "num_input_tokens_seen": 6186768,
      "step": 7135
    },
    {
      "epoch": 3.366336633663366,
      "grad_norm": 0.0031018180307000875,
      "learning_rate": 0.27703280006899617,
      "loss": 0.2155,
      "num_input_tokens_seen": 6191120,
      "step": 7140
    },
    {
      "epoch": 3.3686940122583686,
      "grad_norm": 0.0015641606878489256,
      "learning_rate": 0.277001466109482,
      "loss": 0.2382,
      "num_input_tokens_seen": 6194688,
      "step": 7145
    },
    {
      "epoch": 3.371051390853371,
      "grad_norm": 0.0015605957014486194,
      "learning_rate": 0.2769701125647458,
      "loss": 0.2206,
      "num_input_tokens_seen": 6199520,
      "step": 7150
    },
    {
      "epoch": 3.3734087694483734,
      "grad_norm": 0.0027610515244305134,
      "learning_rate": 0.27693873943962266,
      "loss": 0.263,
      "num_input_tokens_seen": 6203456,
      "step": 7155
    },
    {
      "epoch": 3.375766148043376,
      "grad_norm": 0.0019244597060605884,
      "learning_rate": 0.2769073467389506,
      "loss": 0.2171,
      "num_input_tokens_seen": 6208640,
      "step": 7160
    },
    {
      "epoch": 3.3781235266383782,
      "grad_norm": 0.0013964230893179774,
      "learning_rate": 0.2768759344675709,
      "loss": 0.2093,
      "num_input_tokens_seen": 6213392,
      "step": 7165
    },
    {
      "epoch": 3.3804809052333806,
      "grad_norm": 0.0013954725582152605,
      "learning_rate": 0.27684450263032767,
      "loss": 0.2274,
      "num_input_tokens_seen": 6217952,
      "step": 7170
    },
    {
      "epoch": 3.382838283828383,
      "grad_norm": 0.0012220799690112472,
      "learning_rate": 0.2768130512320682,
      "loss": 0.2229,
      "num_input_tokens_seen": 6222256,
      "step": 7175
    },
    {
      "epoch": 3.385195662423385,
      "grad_norm": 0.005545579828321934,
      "learning_rate": 0.27678158027764244,
      "loss": 0.2269,
      "num_input_tokens_seen": 6226496,
      "step": 7180
    },
    {
      "epoch": 3.3875530410183874,
      "grad_norm": 0.001451921765692532,
      "learning_rate": 0.27675008977190385,
      "loss": 0.1661,
      "num_input_tokens_seen": 6231312,
      "step": 7185
    },
    {
      "epoch": 3.38991041961339,
      "grad_norm": 0.0010781936580315232,
      "learning_rate": 0.2767185797197086,
      "loss": 0.2262,
      "num_input_tokens_seen": 6236048,
      "step": 7190
    },
    {
      "epoch": 3.392267798208392,
      "grad_norm": 0.002580445259809494,
      "learning_rate": 0.2766870501259159,
      "loss": 0.2799,
      "num_input_tokens_seen": 6239840,
      "step": 7195
    },
    {
      "epoch": 3.3946251768033946,
      "grad_norm": 0.0026025348342955112,
      "learning_rate": 0.276655500995388,
      "loss": 0.2365,
      "num_input_tokens_seen": 6244624,
      "step": 7200
    },
    {
      "epoch": 3.3946251768033946,
      "eval_loss": 0.2422030121088028,
      "eval_runtime": 21.8902,
      "eval_samples_per_second": 43.079,
      "eval_steps_per_second": 21.562,
      "num_input_tokens_seen": 6244624,
      "step": 7200
    },
    {
      "epoch": 3.396982555398397,
      "grad_norm": 0.002089216373860836,
      "learning_rate": 0.27662393233299015,
      "loss": 0.2468,
      "num_input_tokens_seen": 6248816,
      "step": 7205
    },
    {
      "epoch": 3.3993399339933994,
      "grad_norm": 0.001023127930238843,
      "learning_rate": 0.27659234414359074,
      "loss": 0.2213,
      "num_input_tokens_seen": 6252736,
      "step": 7210
    },
    {
      "epoch": 3.401697312588402,
      "grad_norm": 0.0016543582314625382,
      "learning_rate": 0.27656073643206097,
      "loss": 0.2212,
      "num_input_tokens_seen": 6256704,
      "step": 7215
    },
    {
      "epoch": 3.404054691183404,
      "grad_norm": 0.0019142511300742626,
      "learning_rate": 0.27652910920327517,
      "loss": 0.2214,
      "num_input_tokens_seen": 6260096,
      "step": 7220
    },
    {
      "epoch": 3.4064120697784066,
      "grad_norm": 0.001541539211757481,
      "learning_rate": 0.2764974624621107,
      "loss": 0.2095,
      "num_input_tokens_seen": 6264032,
      "step": 7225
    },
    {
      "epoch": 3.408769448373409,
      "grad_norm": 0.0027256172616034746,
      "learning_rate": 0.2764657962134479,
      "loss": 0.2085,
      "num_input_tokens_seen": 6268096,
      "step": 7230
    },
    {
      "epoch": 3.411126826968411,
      "grad_norm": 0.002020112005993724,
      "learning_rate": 0.27643411046217,
      "loss": 0.198,
      "num_input_tokens_seen": 6272848,
      "step": 7235
    },
    {
      "epoch": 3.4134842055634134,
      "grad_norm": 0.0038403095677495003,
      "learning_rate": 0.27640240521316334,
      "loss": 0.2779,
      "num_input_tokens_seen": 6277616,
      "step": 7240
    },
    {
      "epoch": 3.4158415841584158,
      "grad_norm": 0.0017760127084329724,
      "learning_rate": 0.2763706804713174,
      "loss": 0.2113,
      "num_input_tokens_seen": 6282768,
      "step": 7245
    },
    {
      "epoch": 3.418198962753418,
      "grad_norm": 0.0017612036317586899,
      "learning_rate": 0.2763389362415245,
      "loss": 0.2208,
      "num_input_tokens_seen": 6287680,
      "step": 7250
    },
    {
      "epoch": 3.4205563413484206,
      "grad_norm": 0.0023202865850180387,
      "learning_rate": 0.27630717252867987,
      "loss": 0.1709,
      "num_input_tokens_seen": 6292032,
      "step": 7255
    },
    {
      "epoch": 3.422913719943423,
      "grad_norm": 0.003048066282644868,
      "learning_rate": 0.276275389337682,
      "loss": 0.1757,
      "num_input_tokens_seen": 6296176,
      "step": 7260
    },
    {
      "epoch": 3.4252710985384254,
      "grad_norm": 0.0016517804469913244,
      "learning_rate": 0.2762435866734322,
      "loss": 0.2036,
      "num_input_tokens_seen": 6300896,
      "step": 7265
    },
    {
      "epoch": 3.4276284771334278,
      "grad_norm": 0.002743727993220091,
      "learning_rate": 0.27621176454083485,
      "loss": 0.2664,
      "num_input_tokens_seen": 6304912,
      "step": 7270
    },
    {
      "epoch": 3.42998585572843,
      "grad_norm": 0.0014227003557607532,
      "learning_rate": 0.2761799229447973,
      "loss": 0.2139,
      "num_input_tokens_seen": 6309328,
      "step": 7275
    },
    {
      "epoch": 3.432343234323432,
      "grad_norm": 0.0024173902347683907,
      "learning_rate": 0.27614806189023006,
      "loss": 0.2174,
      "num_input_tokens_seen": 6314192,
      "step": 7280
    },
    {
      "epoch": 3.4347006129184345,
      "grad_norm": 0.002194293076172471,
      "learning_rate": 0.27611618138204636,
      "loss": 0.2102,
      "num_input_tokens_seen": 6318464,
      "step": 7285
    },
    {
      "epoch": 3.437057991513437,
      "grad_norm": 0.003310837782919407,
      "learning_rate": 0.2760842814251626,
      "loss": 0.1846,
      "num_input_tokens_seen": 6322560,
      "step": 7290
    },
    {
      "epoch": 3.4394153701084393,
      "grad_norm": 0.0026300654280930758,
      "learning_rate": 0.2760523620244982,
      "loss": 0.2021,
      "num_input_tokens_seen": 6326320,
      "step": 7295
    },
    {
      "epoch": 3.4417727487034417,
      "grad_norm": 0.0044151959009468555,
      "learning_rate": 0.27602042318497544,
      "loss": 0.233,
      "num_input_tokens_seen": 6330192,
      "step": 7300
    },
    {
      "epoch": 3.444130127298444,
      "grad_norm": 0.0018041933653876185,
      "learning_rate": 0.2759884649115198,
      "loss": 0.2056,
      "num_input_tokens_seen": 6334288,
      "step": 7305
    },
    {
      "epoch": 3.4464875058934465,
      "grad_norm": 0.0023777398746460676,
      "learning_rate": 0.2759564872090596,
      "loss": 0.2035,
      "num_input_tokens_seen": 6338400,
      "step": 7310
    },
    {
      "epoch": 3.448844884488449,
      "grad_norm": 0.002719591837376356,
      "learning_rate": 0.2759244900825262,
      "loss": 0.2184,
      "num_input_tokens_seen": 6342976,
      "step": 7315
    },
    {
      "epoch": 3.4512022630834513,
      "grad_norm": 0.00211810739710927,
      "learning_rate": 0.2758924735368539,
      "loss": 0.234,
      "num_input_tokens_seen": 6346944,
      "step": 7320
    },
    {
      "epoch": 3.4535596416784538,
      "grad_norm": 0.0014107570750638843,
      "learning_rate": 0.27586043757698014,
      "loss": 0.2192,
      "num_input_tokens_seen": 6351424,
      "step": 7325
    },
    {
      "epoch": 3.455917020273456,
      "grad_norm": 0.0032008313573896885,
      "learning_rate": 0.27582838220784534,
      "loss": 0.1935,
      "num_input_tokens_seen": 6355392,
      "step": 7330
    },
    {
      "epoch": 3.458274398868458,
      "grad_norm": 0.0025628795847296715,
      "learning_rate": 0.27579630743439265,
      "loss": 0.1731,
      "num_input_tokens_seen": 6359040,
      "step": 7335
    },
    {
      "epoch": 3.4606317774634605,
      "grad_norm": 0.0017497598892077804,
      "learning_rate": 0.2757642132615686,
      "loss": 0.2006,
      "num_input_tokens_seen": 6363280,
      "step": 7340
    },
    {
      "epoch": 3.462989156058463,
      "grad_norm": 0.002716874470934272,
      "learning_rate": 0.2757320996943223,
      "loss": 0.2698,
      "num_input_tokens_seen": 6368224,
      "step": 7345
    },
    {
      "epoch": 3.4653465346534653,
      "grad_norm": 0.002742862794548273,
      "learning_rate": 0.2756999667376062,
      "loss": 0.2664,
      "num_input_tokens_seen": 6372240,
      "step": 7350
    },
    {
      "epoch": 3.4677039132484677,
      "grad_norm": 0.003939539194107056,
      "learning_rate": 0.2756678143963756,
      "loss": 0.2445,
      "num_input_tokens_seen": 6376352,
      "step": 7355
    },
    {
      "epoch": 3.47006129184347,
      "grad_norm": 0.004275953397154808,
      "learning_rate": 0.2756356426755888,
      "loss": 0.242,
      "num_input_tokens_seen": 6380864,
      "step": 7360
    },
    {
      "epoch": 3.4724186704384725,
      "grad_norm": 0.0020021451637148857,
      "learning_rate": 0.27560345158020705,
      "loss": 0.2079,
      "num_input_tokens_seen": 6385984,
      "step": 7365
    },
    {
      "epoch": 3.474776049033475,
      "grad_norm": 0.0025579510256648064,
      "learning_rate": 0.27557124111519465,
      "loss": 0.2258,
      "num_input_tokens_seen": 6390864,
      "step": 7370
    },
    {
      "epoch": 3.4771334276284773,
      "grad_norm": 0.001775909448042512,
      "learning_rate": 0.27553901128551883,
      "loss": 0.2774,
      "num_input_tokens_seen": 6394768,
      "step": 7375
    },
    {
      "epoch": 3.4794908062234793,
      "grad_norm": 0.00175549800042063,
      "learning_rate": 0.2755067620961498,
      "loss": 0.2605,
      "num_input_tokens_seen": 6399344,
      "step": 7380
    },
    {
      "epoch": 3.4818481848184817,
      "grad_norm": 0.00382578675635159,
      "learning_rate": 0.27547449355206094,
      "loss": 0.2383,
      "num_input_tokens_seen": 6403904,
      "step": 7385
    },
    {
      "epoch": 3.484205563413484,
      "grad_norm": 0.0027317763306200504,
      "learning_rate": 0.2754422056582283,
      "loss": 0.2525,
      "num_input_tokens_seen": 6408240,
      "step": 7390
    },
    {
      "epoch": 3.4865629420084865,
      "grad_norm": 0.0036534599494189024,
      "learning_rate": 0.27540989841963115,
      "loss": 0.2305,
      "num_input_tokens_seen": 6412352,
      "step": 7395
    },
    {
      "epoch": 3.488920320603489,
      "grad_norm": 0.0017173553351312876,
      "learning_rate": 0.27537757184125167,
      "loss": 0.2499,
      "num_input_tokens_seen": 6416176,
      "step": 7400
    },
    {
      "epoch": 3.488920320603489,
      "eval_loss": 0.22008851170539856,
      "eval_runtime": 21.9493,
      "eval_samples_per_second": 42.963,
      "eval_steps_per_second": 21.504,
      "num_input_tokens_seen": 6416176,
      "step": 7400
    },
    {
      "epoch": 3.4912776991984913,
      "grad_norm": 0.0015983363846316934,
      "learning_rate": 0.275345225928075,
      "loss": 0.2229,
      "num_input_tokens_seen": 6420416,
      "step": 7405
    },
    {
      "epoch": 3.4936350777934937,
      "grad_norm": 0.0015859893755987287,
      "learning_rate": 0.2753128606850893,
      "loss": 0.2134,
      "num_input_tokens_seen": 6424592,
      "step": 7410
    },
    {
      "epoch": 3.495992456388496,
      "grad_norm": 0.0019025172805413604,
      "learning_rate": 0.2752804761172858,
      "loss": 0.2288,
      "num_input_tokens_seen": 6429328,
      "step": 7415
    },
    {
      "epoch": 3.4983498349834985,
      "grad_norm": 0.0035407382529228926,
      "learning_rate": 0.27524807222965836,
      "loss": 0.249,
      "num_input_tokens_seen": 6433776,
      "step": 7420
    },
    {
      "epoch": 3.500707213578501,
      "grad_norm": 0.002059070859104395,
      "learning_rate": 0.27521564902720436,
      "loss": 0.2117,
      "num_input_tokens_seen": 6437840,
      "step": 7425
    },
    {
      "epoch": 3.5030645921735033,
      "grad_norm": 0.0020002226810902357,
      "learning_rate": 0.2751832065149236,
      "loss": 0.2144,
      "num_input_tokens_seen": 6441744,
      "step": 7430
    },
    {
      "epoch": 3.5054219707685053,
      "grad_norm": 0.002228444442152977,
      "learning_rate": 0.2751507446978193,
      "loss": 0.1877,
      "num_input_tokens_seen": 6445440,
      "step": 7435
    },
    {
      "epoch": 3.5077793493635077,
      "grad_norm": 0.0012377174571156502,
      "learning_rate": 0.2751182635808974,
      "loss": 0.1826,
      "num_input_tokens_seen": 6449888,
      "step": 7440
    },
    {
      "epoch": 3.51013672795851,
      "grad_norm": 0.005509630311280489,
      "learning_rate": 0.27508576316916694,
      "loss": 0.2039,
      "num_input_tokens_seen": 6453808,
      "step": 7445
    },
    {
      "epoch": 3.5124941065535125,
      "grad_norm": 0.0023684920743107796,
      "learning_rate": 0.2750532434676399,
      "loss": 0.2596,
      "num_input_tokens_seen": 6457696,
      "step": 7450
    },
    {
      "epoch": 3.514851485148515,
      "grad_norm": 0.0028551174327731133,
      "learning_rate": 0.27502070448133115,
      "loss": 0.1979,
      "num_input_tokens_seen": 6462960,
      "step": 7455
    },
    {
      "epoch": 3.5172088637435173,
      "grad_norm": 0.001305055688135326,
      "learning_rate": 0.2749881462152587,
      "loss": 0.2025,
      "num_input_tokens_seen": 6467536,
      "step": 7460
    },
    {
      "epoch": 3.5195662423385197,
      "grad_norm": 0.004031731281429529,
      "learning_rate": 0.2749555686744434,
      "loss": 0.2452,
      "num_input_tokens_seen": 6471616,
      "step": 7465
    },
    {
      "epoch": 3.521923620933522,
      "grad_norm": 0.001082683796994388,
      "learning_rate": 0.2749229718639091,
      "loss": 0.2318,
      "num_input_tokens_seen": 6475824,
      "step": 7470
    },
    {
      "epoch": 3.524280999528524,
      "grad_norm": 0.0013505717506632209,
      "learning_rate": 0.27489035578868265,
      "loss": 0.2329,
      "num_input_tokens_seen": 6479808,
      "step": 7475
    },
    {
      "epoch": 3.5266383781235264,
      "grad_norm": 0.0033814788330346346,
      "learning_rate": 0.2748577204537939,
      "loss": 0.2364,
      "num_input_tokens_seen": 6484048,
      "step": 7480
    },
    {
      "epoch": 3.528995756718529,
      "grad_norm": 0.0017407538834959269,
      "learning_rate": 0.2748250658642756,
      "loss": 0.2321,
      "num_input_tokens_seen": 6488784,
      "step": 7485
    },
    {
      "epoch": 3.5313531353135312,
      "grad_norm": 0.0013526425464078784,
      "learning_rate": 0.2747923920251634,
      "loss": 0.2098,
      "num_input_tokens_seen": 6493264,
      "step": 7490
    },
    {
      "epoch": 3.5337105139085336,
      "grad_norm": 0.0012853119987994432,
      "learning_rate": 0.27475969894149627,
      "loss": 0.2469,
      "num_input_tokens_seen": 6498368,
      "step": 7495
    },
    {
      "epoch": 3.536067892503536,
      "grad_norm": 0.0012513651745393872,
      "learning_rate": 0.2747269866183156,
      "loss": 0.2355,
      "num_input_tokens_seen": 6502416,
      "step": 7500
    },
    {
      "epoch": 3.5384252710985384,
      "grad_norm": 0.0013078675838187337,
      "learning_rate": 0.27469425506066625,
      "loss": 0.2035,
      "num_input_tokens_seen": 6507264,
      "step": 7505
    },
    {
      "epoch": 3.540782649693541,
      "grad_norm": 0.0013946976978331804,
      "learning_rate": 0.27466150427359576,
      "loss": 0.2354,
      "num_input_tokens_seen": 6511280,
      "step": 7510
    },
    {
      "epoch": 3.5431400282885432,
      "grad_norm": 0.001269750064238906,
      "learning_rate": 0.2746287342621547,
      "loss": 0.2088,
      "num_input_tokens_seen": 6515312,
      "step": 7515
    },
    {
      "epoch": 3.5454974068835456,
      "grad_norm": 0.0023586887400597334,
      "learning_rate": 0.2745959450313966,
      "loss": 0.1954,
      "num_input_tokens_seen": 6519312,
      "step": 7520
    },
    {
      "epoch": 3.547854785478548,
      "grad_norm": 0.0049217464402318,
      "learning_rate": 0.27456313658637804,
      "loss": 0.2771,
      "num_input_tokens_seen": 6524416,
      "step": 7525
    },
    {
      "epoch": 3.5502121640735504,
      "grad_norm": 0.001406246330589056,
      "learning_rate": 0.27453030893215846,
      "loss": 0.2125,
      "num_input_tokens_seen": 6528720,
      "step": 7530
    },
    {
      "epoch": 3.5525695426685524,
      "grad_norm": 0.001640518195927143,
      "learning_rate": 0.2744974620738003,
      "loss": 0.2326,
      "num_input_tokens_seen": 6533296,
      "step": 7535
    },
    {
      "epoch": 3.554926921263555,
      "grad_norm": 0.0025143090169876814,
      "learning_rate": 0.27446459601636897,
      "loss": 0.2319,
      "num_input_tokens_seen": 6537792,
      "step": 7540
    },
    {
      "epoch": 3.557284299858557,
      "grad_norm": 0.0016905986703932285,
      "learning_rate": 0.2744317107649328,
      "loss": 0.2246,
      "num_input_tokens_seen": 6542208,
      "step": 7545
    },
    {
      "epoch": 3.5596416784535596,
      "grad_norm": 0.0020396022591739893,
      "learning_rate": 0.2743988063245631,
      "loss": 0.258,
      "num_input_tokens_seen": 6546496,
      "step": 7550
    },
    {
      "epoch": 3.561999057048562,
      "grad_norm": 0.002980960300192237,
      "learning_rate": 0.2743658827003342,
      "loss": 0.2256,
      "num_input_tokens_seen": 6550288,
      "step": 7555
    },
    {
      "epoch": 3.5643564356435644,
      "grad_norm": 0.002304153284057975,
      "learning_rate": 0.27433293989732327,
      "loss": 0.203,
      "num_input_tokens_seen": 6554416,
      "step": 7560
    },
    {
      "epoch": 3.566713814238567,
      "grad_norm": 0.001734415884129703,
      "learning_rate": 0.27429997792061056,
      "loss": 0.2125,
      "num_input_tokens_seen": 6559232,
      "step": 7565
    },
    {
      "epoch": 3.569071192833569,
      "grad_norm": 0.0008995212847366929,
      "learning_rate": 0.27426699677527927,
      "loss": 0.2433,
      "num_input_tokens_seen": 6563008,
      "step": 7570
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.001496580196544528,
      "learning_rate": 0.2742339964664154,
      "loss": 0.2617,
      "num_input_tokens_seen": 6566848,
      "step": 7575
    },
    {
      "epoch": 3.5737859500235736,
      "grad_norm": 0.0024066439364105463,
      "learning_rate": 0.274200976999108,
      "loss": 0.2322,
      "num_input_tokens_seen": 6571664,
      "step": 7580
    },
    {
      "epoch": 3.576143328618576,
      "grad_norm": 0.0023740872275084257,
      "learning_rate": 0.27416793837844916,
      "loss": 0.2047,
      "num_input_tokens_seen": 6575616,
      "step": 7585
    },
    {
      "epoch": 3.5785007072135784,
      "grad_norm": 0.0024361128453165293,
      "learning_rate": 0.27413488060953384,
      "loss": 0.2437,
      "num_input_tokens_seen": 6579280,
      "step": 7590
    },
    {
      "epoch": 3.580858085808581,
      "grad_norm": 0.0012804957805201411,
      "learning_rate": 0.27410180369745996,
      "loss": 0.2294,
      "num_input_tokens_seen": 6583472,
      "step": 7595
    },
    {
      "epoch": 3.583215464403583,
      "grad_norm": 0.0010128882713615894,
      "learning_rate": 0.27406870764732844,
      "loss": 0.2372,
      "num_input_tokens_seen": 6587616,
      "step": 7600
    },
    {
      "epoch": 3.583215464403583,
      "eval_loss": 0.22464415431022644,
      "eval_runtime": 21.9075,
      "eval_samples_per_second": 43.045,
      "eval_steps_per_second": 21.545,
      "num_input_tokens_seen": 6587616,
      "step": 7600
    },
    {
      "epoch": 3.5855728429985856,
      "grad_norm": 0.0010941236978396773,
      "learning_rate": 0.27403559246424297,
      "loss": 0.2348,
      "num_input_tokens_seen": 6591680,
      "step": 7605
    },
    {
      "epoch": 3.587930221593588,
      "grad_norm": 0.0013178616063669324,
      "learning_rate": 0.2740024581533105,
      "loss": 0.2158,
      "num_input_tokens_seen": 6595344,
      "step": 7610
    },
    {
      "epoch": 3.5902876001885904,
      "grad_norm": 0.0009237665217369795,
      "learning_rate": 0.2739693047196406,
      "loss": 0.1956,
      "num_input_tokens_seen": 6598832,
      "step": 7615
    },
    {
      "epoch": 3.592644978783593,
      "grad_norm": 0.0018076588166877627,
      "learning_rate": 0.27393613216834606,
      "loss": 0.2635,
      "num_input_tokens_seen": 6603056,
      "step": 7620
    },
    {
      "epoch": 3.595002357378595,
      "grad_norm": 0.002524383831769228,
      "learning_rate": 0.2739029405045424,
      "loss": 0.2527,
      "num_input_tokens_seen": 6607040,
      "step": 7625
    },
    {
      "epoch": 3.5973597359735976,
      "grad_norm": 0.002406864194199443,
      "learning_rate": 0.2738697297333483,
      "loss": 0.1826,
      "num_input_tokens_seen": 6611840,
      "step": 7630
    },
    {
      "epoch": 3.5997171145685996,
      "grad_norm": 0.0014414064353331923,
      "learning_rate": 0.2738364998598852,
      "loss": 0.2329,
      "num_input_tokens_seen": 6616608,
      "step": 7635
    },
    {
      "epoch": 3.602074493163602,
      "grad_norm": 0.0011891728499904275,
      "learning_rate": 0.27380325088927765,
      "loss": 0.2146,
      "num_input_tokens_seen": 6620624,
      "step": 7640
    },
    {
      "epoch": 3.6044318717586044,
      "grad_norm": 0.001107415766455233,
      "learning_rate": 0.27376998282665294,
      "loss": 0.2238,
      "num_input_tokens_seen": 6624656,
      "step": 7645
    },
    {
      "epoch": 3.6067892503536068,
      "grad_norm": 0.0012750504538416862,
      "learning_rate": 0.27373669567714154,
      "loss": 0.2206,
      "num_input_tokens_seen": 6628976,
      "step": 7650
    },
    {
      "epoch": 3.609146628948609,
      "grad_norm": 0.0017991869244724512,
      "learning_rate": 0.27370338944587663,
      "loss": 0.2313,
      "num_input_tokens_seen": 6633744,
      "step": 7655
    },
    {
      "epoch": 3.6115040075436116,
      "grad_norm": 0.0022403558250516653,
      "learning_rate": 0.27367006413799455,
      "loss": 0.2156,
      "num_input_tokens_seen": 6638256,
      "step": 7660
    },
    {
      "epoch": 3.613861386138614,
      "grad_norm": 0.0018583911005407572,
      "learning_rate": 0.2736367197586345,
      "loss": 0.2335,
      "num_input_tokens_seen": 6643408,
      "step": 7665
    },
    {
      "epoch": 3.6162187647336164,
      "grad_norm": 0.0026185179594904184,
      "learning_rate": 0.2736033563129385,
      "loss": 0.2347,
      "num_input_tokens_seen": 6647744,
      "step": 7670
    },
    {
      "epoch": 3.6185761433286183,
      "grad_norm": 0.00175188307184726,
      "learning_rate": 0.27356997380605164,
      "loss": 0.2146,
      "num_input_tokens_seen": 6652448,
      "step": 7675
    },
    {
      "epoch": 3.6209335219236207,
      "grad_norm": 0.0016861824551597238,
      "learning_rate": 0.27353657224312194,
      "loss": 0.198,
      "num_input_tokens_seen": 6656576,
      "step": 7680
    },
    {
      "epoch": 3.623290900518623,
      "grad_norm": 0.002153982874006033,
      "learning_rate": 0.2735031516293004,
      "loss": 0.21,
      "num_input_tokens_seen": 6661088,
      "step": 7685
    },
    {
      "epoch": 3.6256482791136255,
      "grad_norm": 0.0012487162603065372,
      "learning_rate": 0.2734697119697408,
      "loss": 0.2118,
      "num_input_tokens_seen": 6665008,
      "step": 7690
    },
    {
      "epoch": 3.628005657708628,
      "grad_norm": 0.0018250724533572793,
      "learning_rate": 0.27343625326959997,
      "loss": 0.1919,
      "num_input_tokens_seen": 6668880,
      "step": 7695
    },
    {
      "epoch": 3.6303630363036303,
      "grad_norm": 0.0011959741823375225,
      "learning_rate": 0.27340277553403775,
      "loss": 0.238,
      "num_input_tokens_seen": 6673984,
      "step": 7700
    },
    {
      "epoch": 3.6327204148986327,
      "grad_norm": 0.0015136012807488441,
      "learning_rate": 0.2733692787682167,
      "loss": 0.1635,
      "num_input_tokens_seen": 6678080,
      "step": 7705
    },
    {
      "epoch": 3.635077793493635,
      "grad_norm": 0.0023738176096230745,
      "learning_rate": 0.27333576297730255,
      "loss": 0.1934,
      "num_input_tokens_seen": 6682416,
      "step": 7710
    },
    {
      "epoch": 3.6374351720886375,
      "grad_norm": 0.0035761655308306217,
      "learning_rate": 0.2733022281664638,
      "loss": 0.2469,
      "num_input_tokens_seen": 6687072,
      "step": 7715
    },
    {
      "epoch": 3.63979255068364,
      "grad_norm": 0.0020619500428438187,
      "learning_rate": 0.273268674340872,
      "loss": 0.1793,
      "num_input_tokens_seen": 6691824,
      "step": 7720
    },
    {
      "epoch": 3.6421499292786423,
      "grad_norm": 0.033405501395463943,
      "learning_rate": 0.27323510150570146,
      "loss": 0.2553,
      "num_input_tokens_seen": 6696128,
      "step": 7725
    },
    {
      "epoch": 3.6445073078736447,
      "grad_norm": 0.0029146000742912292,
      "learning_rate": 0.27320150966612966,
      "loss": 0.2503,
      "num_input_tokens_seen": 6700192,
      "step": 7730
    },
    {
      "epoch": 3.6468646864686467,
      "grad_norm": 0.004211536142975092,
      "learning_rate": 0.2731678988273368,
      "loss": 0.2344,
      "num_input_tokens_seen": 6704064,
      "step": 7735
    },
    {
      "epoch": 3.649222065063649,
      "grad_norm": 0.0037623511161655188,
      "learning_rate": 0.27313426899450605,
      "loss": 0.2263,
      "num_input_tokens_seen": 6708944,
      "step": 7740
    },
    {
      "epoch": 3.6515794436586515,
      "grad_norm": 0.0022754922974854708,
      "learning_rate": 0.27310062017282366,
      "loss": 0.2291,
      "num_input_tokens_seen": 6713184,
      "step": 7745
    },
    {
      "epoch": 3.653936822253654,
      "grad_norm": 0.15484841167926788,
      "learning_rate": 0.2730669523674787,
      "loss": 0.3082,
      "num_input_tokens_seen": 6717424,
      "step": 7750
    },
    {
      "epoch": 3.6562942008486563,
      "grad_norm": 0.004124440718442202,
      "learning_rate": 0.2730332655836631,
      "loss": 0.1846,
      "num_input_tokens_seen": 6720960,
      "step": 7755
    },
    {
      "epoch": 3.6586515794436587,
      "grad_norm": 0.009590270929038525,
      "learning_rate": 0.2729995598265718,
      "loss": 0.3028,
      "num_input_tokens_seen": 6724800,
      "step": 7760
    },
    {
      "epoch": 3.661008958038661,
      "grad_norm": 0.004252584185451269,
      "learning_rate": 0.2729658351014027,
      "loss": 0.2277,
      "num_input_tokens_seen": 6728928,
      "step": 7765
    },
    {
      "epoch": 3.6633663366336635,
      "grad_norm": 0.004570796620100737,
      "learning_rate": 0.27293209141335656,
      "loss": 0.2213,
      "num_input_tokens_seen": 6732736,
      "step": 7770
    },
    {
      "epoch": 3.6657237152286655,
      "grad_norm": 0.005625327117741108,
      "learning_rate": 0.27289832876763703,
      "loss": 0.2587,
      "num_input_tokens_seen": 6737328,
      "step": 7775
    },
    {
      "epoch": 3.668081093823668,
      "grad_norm": 0.0033283380325883627,
      "learning_rate": 0.27286454716945074,
      "loss": 0.2251,
      "num_input_tokens_seen": 6741760,
      "step": 7780
    },
    {
      "epoch": 3.6704384724186703,
      "grad_norm": 0.0032676574774086475,
      "learning_rate": 0.27283074662400725,
      "loss": 0.2363,
      "num_input_tokens_seen": 6745952,
      "step": 7785
    },
    {
      "epoch": 3.6727958510136727,
      "grad_norm": 0.004196304362267256,
      "learning_rate": 0.2727969271365191,
      "loss": 0.2778,
      "num_input_tokens_seen": 6750096,
      "step": 7790
    },
    {
      "epoch": 3.675153229608675,
      "grad_norm": 0.03748945891857147,
      "learning_rate": 0.2727630887122016,
      "loss": 0.2278,
      "num_input_tokens_seen": 6754704,
      "step": 7795
    },
    {
      "epoch": 3.6775106082036775,
      "grad_norm": 0.035863518714904785,
      "learning_rate": 0.27272923135627314,
      "loss": 0.2897,
      "num_input_tokens_seen": 6759696,
      "step": 7800
    },
    {
      "epoch": 3.6775106082036775,
      "eval_loss": 0.3597075641155243,
      "eval_runtime": 21.8614,
      "eval_samples_per_second": 43.135,
      "eval_steps_per_second": 21.591,
      "num_input_tokens_seen": 6759696,
      "step": 7800
    },
    {
      "epoch": 3.67986798679868,
      "grad_norm": 0.033299729228019714,
      "learning_rate": 0.2726953550739548,
      "loss": 0.3237,
      "num_input_tokens_seen": 6763664,
      "step": 7805
    },
    {
      "epoch": 3.6822253653936823,
      "grad_norm": 0.010386921465396881,
      "learning_rate": 0.27266145987047086,
      "loss": 0.3984,
      "num_input_tokens_seen": 6767296,
      "step": 7810
    },
    {
      "epoch": 3.6845827439886847,
      "grad_norm": 0.018354682251811028,
      "learning_rate": 0.27262754575104836,
      "loss": 0.2373,
      "num_input_tokens_seen": 6771728,
      "step": 7815
    },
    {
      "epoch": 3.686940122583687,
      "grad_norm": 0.12378855794668198,
      "learning_rate": 0.27259361272091726,
      "loss": 0.6709,
      "num_input_tokens_seen": 6775968,
      "step": 7820
    },
    {
      "epoch": 3.6892975011786895,
      "grad_norm": 0.07291828840970993,
      "learning_rate": 0.27255966078531046,
      "loss": 0.2807,
      "num_input_tokens_seen": 6780384,
      "step": 7825
    },
    {
      "epoch": 3.691654879773692,
      "grad_norm": 0.008483180776238441,
      "learning_rate": 0.2725256899494638,
      "loss": 0.2658,
      "num_input_tokens_seen": 6784592,
      "step": 7830
    },
    {
      "epoch": 3.694012258368694,
      "grad_norm": 0.005903953220695257,
      "learning_rate": 0.272491700218616,
      "loss": 0.1911,
      "num_input_tokens_seen": 6788592,
      "step": 7835
    },
    {
      "epoch": 3.6963696369636962,
      "grad_norm": 0.02551133930683136,
      "learning_rate": 0.27245769159800876,
      "loss": 0.2617,
      "num_input_tokens_seen": 6792384,
      "step": 7840
    },
    {
      "epoch": 3.6987270155586986,
      "grad_norm": 0.011712886393070221,
      "learning_rate": 0.2724236640928865,
      "loss": 0.3135,
      "num_input_tokens_seen": 6796480,
      "step": 7845
    },
    {
      "epoch": 3.701084394153701,
      "grad_norm": 0.010079601779580116,
      "learning_rate": 0.27238961770849673,
      "loss": 0.2164,
      "num_input_tokens_seen": 6800400,
      "step": 7850
    },
    {
      "epoch": 3.7034417727487035,
      "grad_norm": 0.007175520062446594,
      "learning_rate": 0.27235555245008997,
      "loss": 0.2937,
      "num_input_tokens_seen": 6804896,
      "step": 7855
    },
    {
      "epoch": 3.705799151343706,
      "grad_norm": 0.009018326178193092,
      "learning_rate": 0.2723214683229193,
      "loss": 0.2472,
      "num_input_tokens_seen": 6809216,
      "step": 7860
    },
    {
      "epoch": 3.7081565299387083,
      "grad_norm": 0.009131485596299171,
      "learning_rate": 0.27228736533224107,
      "loss": 0.2829,
      "num_input_tokens_seen": 6813904,
      "step": 7865
    },
    {
      "epoch": 3.7105139085337107,
      "grad_norm": 0.007141607813537121,
      "learning_rate": 0.27225324348331437,
      "loss": 0.2614,
      "num_input_tokens_seen": 6818128,
      "step": 7870
    },
    {
      "epoch": 3.7128712871287126,
      "grad_norm": 0.0032803788781166077,
      "learning_rate": 0.27221910278140116,
      "loss": 0.2453,
      "num_input_tokens_seen": 6823136,
      "step": 7875
    },
    {
      "epoch": 3.715228665723715,
      "grad_norm": 0.002921412466093898,
      "learning_rate": 0.2721849432317664,
      "loss": 0.2982,
      "num_input_tokens_seen": 6827760,
      "step": 7880
    },
    {
      "epoch": 3.7175860443187174,
      "grad_norm": 0.0025941256899386644,
      "learning_rate": 0.2721507648396779,
      "loss": 0.2711,
      "num_input_tokens_seen": 6832144,
      "step": 7885
    },
    {
      "epoch": 3.71994342291372,
      "grad_norm": 0.005638516042381525,
      "learning_rate": 0.27211656761040653,
      "loss": 0.2318,
      "num_input_tokens_seen": 6835760,
      "step": 7890
    },
    {
      "epoch": 3.7223008015087222,
      "grad_norm": 0.01043060701340437,
      "learning_rate": 0.2720823515492257,
      "loss": 0.3034,
      "num_input_tokens_seen": 6840128,
      "step": 7895
    },
    {
      "epoch": 3.7246581801037246,
      "grad_norm": 0.0058115446008741856,
      "learning_rate": 0.27204811666141215,
      "loss": 0.2203,
      "num_input_tokens_seen": 6843920,
      "step": 7900
    },
    {
      "epoch": 3.727015558698727,
      "grad_norm": 0.005477519705891609,
      "learning_rate": 0.2720138629522452,
      "loss": 0.2485,
      "num_input_tokens_seen": 6848304,
      "step": 7905
    },
    {
      "epoch": 3.7293729372937294,
      "grad_norm": 0.009948208928108215,
      "learning_rate": 0.2719795904270073,
      "loss": 0.2686,
      "num_input_tokens_seen": 6852544,
      "step": 7910
    },
    {
      "epoch": 3.731730315888732,
      "grad_norm": 0.0054732575081288815,
      "learning_rate": 0.2719452990909837,
      "loss": 0.2363,
      "num_input_tokens_seen": 6857696,
      "step": 7915
    },
    {
      "epoch": 3.7340876944837342,
      "grad_norm": 0.00380932935513556,
      "learning_rate": 0.2719109889494625,
      "loss": 0.353,
      "num_input_tokens_seen": 6862496,
      "step": 7920
    },
    {
      "epoch": 3.7364450730787366,
      "grad_norm": 0.005342503078281879,
      "learning_rate": 0.27187666000773475,
      "loss": 0.2175,
      "num_input_tokens_seen": 6867056,
      "step": 7925
    },
    {
      "epoch": 3.738802451673739,
      "grad_norm": 0.00863492488861084,
      "learning_rate": 0.2718423122710944,
      "loss": 0.2141,
      "num_input_tokens_seen": 6872144,
      "step": 7930
    },
    {
      "epoch": 3.741159830268741,
      "grad_norm": 0.009092372842133045,
      "learning_rate": 0.2718079457448384,
      "loss": 0.293,
      "num_input_tokens_seen": 6876464,
      "step": 7935
    },
    {
      "epoch": 3.7435172088637434,
      "grad_norm": 0.004403768107295036,
      "learning_rate": 0.27177356043426637,
      "loss": 0.3639,
      "num_input_tokens_seen": 6882096,
      "step": 7940
    },
    {
      "epoch": 3.745874587458746,
      "grad_norm": 0.008505245670676231,
      "learning_rate": 0.27173915634468104,
      "loss": 0.2598,
      "num_input_tokens_seen": 6886048,
      "step": 7945
    },
    {
      "epoch": 3.748231966053748,
      "grad_norm": 0.0038329889066517353,
      "learning_rate": 0.27170473348138796,
      "loss": 0.258,
      "num_input_tokens_seen": 6890976,
      "step": 7950
    },
    {
      "epoch": 3.7505893446487506,
      "grad_norm": 0.0030589252710342407,
      "learning_rate": 0.27167029184969554,
      "loss": 0.2262,
      "num_input_tokens_seen": 6895440,
      "step": 7955
    },
    {
      "epoch": 3.752946723243753,
      "grad_norm": 0.002406762447208166,
      "learning_rate": 0.27163583145491504,
      "loss": 0.2236,
      "num_input_tokens_seen": 6899824,
      "step": 7960
    },
    {
      "epoch": 3.7553041018387554,
      "grad_norm": 0.0025360353756695986,
      "learning_rate": 0.2716013523023608,
      "loss": 0.2213,
      "num_input_tokens_seen": 6903632,
      "step": 7965
    },
    {
      "epoch": 3.757661480433758,
      "grad_norm": 0.0026745933573693037,
      "learning_rate": 0.27156685439734995,
      "loss": 0.2319,
      "num_input_tokens_seen": 6907648,
      "step": 7970
    },
    {
      "epoch": 3.7600188590287598,
      "grad_norm": 0.0028864555060863495,
      "learning_rate": 0.2715323377452024,
      "loss": 0.2014,
      "num_input_tokens_seen": 6912080,
      "step": 7975
    },
    {
      "epoch": 3.762376237623762,
      "grad_norm": 0.0023386471439152956,
      "learning_rate": 0.2714978023512411,
      "loss": 0.1671,
      "num_input_tokens_seen": 6916400,
      "step": 7980
    },
    {
      "epoch": 3.7647336162187646,
      "grad_norm": 0.0016304695745930076,
      "learning_rate": 0.2714632482207918,
      "loss": 0.258,
      "num_input_tokens_seen": 6920064,
      "step": 7985
    },
    {
      "epoch": 3.767090994813767,
      "grad_norm": 0.003582676174119115,
      "learning_rate": 0.2714286753591833,
      "loss": 0.2236,
      "num_input_tokens_seen": 6924256,
      "step": 7990
    },
    {
      "epoch": 3.7694483734087694,
      "grad_norm": 0.00515202758833766,
      "learning_rate": 0.27139408377174706,
      "loss": 0.2583,
      "num_input_tokens_seen": 6928624,
      "step": 7995
    },
    {
      "epoch": 3.7718057520037718,
      "grad_norm": 0.0033301750663667917,
      "learning_rate": 0.27135947346381756,
      "loss": 0.236,
      "num_input_tokens_seen": 6932384,
      "step": 8000
    },
    {
      "epoch": 3.7718057520037718,
      "eval_loss": 0.22836388647556305,
      "eval_runtime": 21.8833,
      "eval_samples_per_second": 43.092,
      "eval_steps_per_second": 21.569,
      "num_input_tokens_seen": 6932384,
      "step": 8000
    },
    {
      "epoch": 3.774163130598774,
      "grad_norm": 0.004053758457303047,
      "learning_rate": 0.2713248444407322,
      "loss": 0.2354,
      "num_input_tokens_seen": 6937344,
      "step": 8005
    },
    {
      "epoch": 3.7765205091937766,
      "grad_norm": 0.0029634456150233746,
      "learning_rate": 0.27129019670783106,
      "loss": 0.2114,
      "num_input_tokens_seen": 6942160,
      "step": 8010
    },
    {
      "epoch": 3.778877887788779,
      "grad_norm": 0.0019901327323168516,
      "learning_rate": 0.27125553027045746,
      "loss": 0.2104,
      "num_input_tokens_seen": 6946544,
      "step": 8015
    },
    {
      "epoch": 3.7812352663837814,
      "grad_norm": 0.003887371625751257,
      "learning_rate": 0.2712208451339572,
      "loss": 0.1887,
      "num_input_tokens_seen": 6951264,
      "step": 8020
    },
    {
      "epoch": 3.783592644978784,
      "grad_norm": 0.006146032828837633,
      "learning_rate": 0.27118614130367935,
      "loss": 0.2622,
      "num_input_tokens_seen": 6955280,
      "step": 8025
    },
    {
      "epoch": 3.785950023573786,
      "grad_norm": 0.002621586201712489,
      "learning_rate": 0.2711514187849756,
      "loss": 0.1861,
      "num_input_tokens_seen": 6960096,
      "step": 8030
    },
    {
      "epoch": 3.7883074021687886,
      "grad_norm": 0.002455734880641103,
      "learning_rate": 0.27111667758320057,
      "loss": 0.1973,
      "num_input_tokens_seen": 6963936,
      "step": 8035
    },
    {
      "epoch": 3.7906647807637905,
      "grad_norm": 0.002987583866342902,
      "learning_rate": 0.27108191770371176,
      "loss": 0.2013,
      "num_input_tokens_seen": 6968640,
      "step": 8040
    },
    {
      "epoch": 3.793022159358793,
      "grad_norm": 0.0023241937160491943,
      "learning_rate": 0.2710471391518697,
      "loss": 0.2857,
      "num_input_tokens_seen": 6972752,
      "step": 8045
    },
    {
      "epoch": 3.7953795379537953,
      "grad_norm": 0.005478262901306152,
      "learning_rate": 0.2710123419330375,
      "loss": 0.1962,
      "num_input_tokens_seen": 6976576,
      "step": 8050
    },
    {
      "epoch": 3.7977369165487977,
      "grad_norm": 0.0811113715171814,
      "learning_rate": 0.2709775260525816,
      "loss": 0.369,
      "num_input_tokens_seen": 6981408,
      "step": 8055
    },
    {
      "epoch": 3.8000942951438,
      "grad_norm": 0.006993884686380625,
      "learning_rate": 0.27094269151587075,
      "loss": 0.2418,
      "num_input_tokens_seen": 6985840,
      "step": 8060
    },
    {
      "epoch": 3.8024516737388026,
      "grad_norm": 0.0054725706577301025,
      "learning_rate": 0.27090783832827703,
      "loss": 0.229,
      "num_input_tokens_seen": 6989632,
      "step": 8065
    },
    {
      "epoch": 3.804809052333805,
      "grad_norm": 0.0058980234898626804,
      "learning_rate": 0.2708729664951753,
      "loss": 0.1913,
      "num_input_tokens_seen": 6993312,
      "step": 8070
    },
    {
      "epoch": 3.807166430928807,
      "grad_norm": 0.01407616026699543,
      "learning_rate": 0.27083807602194304,
      "loss": 0.2563,
      "num_input_tokens_seen": 6997824,
      "step": 8075
    },
    {
      "epoch": 3.8095238095238093,
      "grad_norm": 0.0160497035831213,
      "learning_rate": 0.270803166913961,
      "loss": 0.2232,
      "num_input_tokens_seen": 7002272,
      "step": 8080
    },
    {
      "epoch": 3.8118811881188117,
      "grad_norm": 0.01089609507471323,
      "learning_rate": 0.27076823917661247,
      "loss": 0.3546,
      "num_input_tokens_seen": 7006672,
      "step": 8085
    },
    {
      "epoch": 3.814238566713814,
      "grad_norm": 0.014443318359553814,
      "learning_rate": 0.2707332928152838,
      "loss": 0.2792,
      "num_input_tokens_seen": 7011024,
      "step": 8090
    },
    {
      "epoch": 3.8165959453088165,
      "grad_norm": 0.016217920929193497,
      "learning_rate": 0.2706983278353641,
      "loss": 0.2565,
      "num_input_tokens_seen": 7015264,
      "step": 8095
    },
    {
      "epoch": 3.818953323903819,
      "grad_norm": 0.015500267967581749,
      "learning_rate": 0.27066334424224553,
      "loss": 0.2381,
      "num_input_tokens_seen": 7019616,
      "step": 8100
    },
    {
      "epoch": 3.8213107024988213,
      "grad_norm": 0.010520357638597488,
      "learning_rate": 0.27062834204132297,
      "loss": 0.4027,
      "num_input_tokens_seen": 7023984,
      "step": 8105
    },
    {
      "epoch": 3.8236680810938237,
      "grad_norm": 0.016029974445700645,
      "learning_rate": 0.27059332123799407,
      "loss": 0.2592,
      "num_input_tokens_seen": 7027648,
      "step": 8110
    },
    {
      "epoch": 3.826025459688826,
      "grad_norm": 0.008358299732208252,
      "learning_rate": 0.27055828183765956,
      "loss": 0.1755,
      "num_input_tokens_seen": 7031744,
      "step": 8115
    },
    {
      "epoch": 3.8283828382838285,
      "grad_norm": 0.03197292983531952,
      "learning_rate": 0.270523223845723,
      "loss": 0.3782,
      "num_input_tokens_seen": 7036624,
      "step": 8120
    },
    {
      "epoch": 3.830740216878831,
      "grad_norm": 0.01838686503469944,
      "learning_rate": 0.2704881472675907,
      "loss": 0.2611,
      "num_input_tokens_seen": 7040960,
      "step": 8125
    },
    {
      "epoch": 3.8330975954738333,
      "grad_norm": 0.012036159634590149,
      "learning_rate": 0.270453052108672,
      "loss": 0.2003,
      "num_input_tokens_seen": 7045024,
      "step": 8130
    },
    {
      "epoch": 3.8354549740688357,
      "grad_norm": 0.02660859189927578,
      "learning_rate": 0.2704179383743789,
      "loss": 0.2147,
      "num_input_tokens_seen": 7049360,
      "step": 8135
    },
    {
      "epoch": 3.8378123526638377,
      "grad_norm": 0.023135241121053696,
      "learning_rate": 0.27038280607012644,
      "loss": 0.3311,
      "num_input_tokens_seen": 7053968,
      "step": 8140
    },
    {
      "epoch": 3.84016973125884,
      "grad_norm": 0.011876881122589111,
      "learning_rate": 0.27034765520133247,
      "loss": 0.2808,
      "num_input_tokens_seen": 7057520,
      "step": 8145
    },
    {
      "epoch": 3.8425271098538425,
      "grad_norm": 0.013870195485651493,
      "learning_rate": 0.2703124857734177,
      "loss": 0.2482,
      "num_input_tokens_seen": 7061728,
      "step": 8150
    },
    {
      "epoch": 3.844884488448845,
      "grad_norm": 0.010514350607991219,
      "learning_rate": 0.27027729779180565,
      "loss": 0.2692,
      "num_input_tokens_seen": 7065680,
      "step": 8155
    },
    {
      "epoch": 3.8472418670438473,
      "grad_norm": 0.01690591685473919,
      "learning_rate": 0.27024209126192283,
      "loss": 0.2613,
      "num_input_tokens_seen": 7069904,
      "step": 8160
    },
    {
      "epoch": 3.8495992456388497,
      "grad_norm": 0.009164371527731419,
      "learning_rate": 0.2702068661891984,
      "loss": 0.2686,
      "num_input_tokens_seen": 7074224,
      "step": 8165
    },
    {
      "epoch": 3.851956624233852,
      "grad_norm": 0.00950782559812069,
      "learning_rate": 0.2701716225790647,
      "loss": 0.2371,
      "num_input_tokens_seen": 7077968,
      "step": 8170
    },
    {
      "epoch": 3.854314002828854,
      "grad_norm": 0.00806693360209465,
      "learning_rate": 0.27013636043695655,
      "loss": 0.2105,
      "num_input_tokens_seen": 7081552,
      "step": 8175
    },
    {
      "epoch": 3.8566713814238565,
      "grad_norm": 0.01922430843114853,
      "learning_rate": 0.27010107976831194,
      "loss": 0.2657,
      "num_input_tokens_seen": 7086160,
      "step": 8180
    },
    {
      "epoch": 3.859028760018859,
      "grad_norm": 0.014402499422430992,
      "learning_rate": 0.2700657805785715,
      "loss": 0.229,
      "num_input_tokens_seen": 7090416,
      "step": 8185
    },
    {
      "epoch": 3.8613861386138613,
      "grad_norm": 0.016360608860850334,
      "learning_rate": 0.2700304628731789,
      "loss": 0.2475,
      "num_input_tokens_seen": 7094304,
      "step": 8190
    },
    {
      "epoch": 3.8637435172088637,
      "grad_norm": 0.014377044513821602,
      "learning_rate": 0.26999512665758046,
      "loss": 0.2466,
      "num_input_tokens_seen": 7099264,
      "step": 8195
    },
    {
      "epoch": 3.866100895803866,
      "grad_norm": 0.029197312891483307,
      "learning_rate": 0.2699597719372256,
      "loss": 0.2549,
      "num_input_tokens_seen": 7103328,
      "step": 8200
    },
    {
      "epoch": 3.866100895803866,
      "eval_loss": 0.24831777811050415,
      "eval_runtime": 21.9653,
      "eval_samples_per_second": 42.931,
      "eval_steps_per_second": 21.488,
      "num_input_tokens_seen": 7103328,
      "step": 8200
    },
    {
      "epoch": 3.8684582743988685,
      "grad_norm": 0.03602661192417145,
      "learning_rate": 0.26992439871756635,
      "loss": 0.2549,
      "num_input_tokens_seen": 7107472,
      "step": 8205
    },
    {
      "epoch": 3.870815652993871,
      "grad_norm": 0.06549565494060516,
      "learning_rate": 0.2698890070040578,
      "loss": 0.2201,
      "num_input_tokens_seen": 7111776,
      "step": 8210
    },
    {
      "epoch": 3.8731730315888733,
      "grad_norm": 0.08627652376890182,
      "learning_rate": 0.2698535968021577,
      "loss": 0.3053,
      "num_input_tokens_seen": 7115616,
      "step": 8215
    },
    {
      "epoch": 3.8755304101838757,
      "grad_norm": 0.059709735214710236,
      "learning_rate": 0.26981816811732684,
      "loss": 0.2824,
      "num_input_tokens_seen": 7119808,
      "step": 8220
    },
    {
      "epoch": 3.877887788778878,
      "grad_norm": 0.07356496155261993,
      "learning_rate": 0.26978272095502875,
      "loss": 0.4549,
      "num_input_tokens_seen": 7124064,
      "step": 8225
    },
    {
      "epoch": 3.8802451673738805,
      "grad_norm": 0.06219466030597687,
      "learning_rate": 0.26974725532072974,
      "loss": 0.2641,
      "num_input_tokens_seen": 7128848,
      "step": 8230
    },
    {
      "epoch": 3.882602545968883,
      "grad_norm": 0.1410224735736847,
      "learning_rate": 0.26971177121989914,
      "loss": 0.2735,
      "num_input_tokens_seen": 7134128,
      "step": 8235
    },
    {
      "epoch": 3.884959924563885,
      "grad_norm": 0.02292437106370926,
      "learning_rate": 0.2696762686580091,
      "loss": 0.2825,
      "num_input_tokens_seen": 7138912,
      "step": 8240
    },
    {
      "epoch": 3.8873173031588872,
      "grad_norm": 0.045064736157655716,
      "learning_rate": 0.26964074764053436,
      "loss": 0.3228,
      "num_input_tokens_seen": 7143792,
      "step": 8245
    },
    {
      "epoch": 3.8896746817538896,
      "grad_norm": 0.09522506594657898,
      "learning_rate": 0.2696052081729529,
      "loss": 0.3348,
      "num_input_tokens_seen": 7148368,
      "step": 8250
    },
    {
      "epoch": 3.892032060348892,
      "grad_norm": 0.018252654001116753,
      "learning_rate": 0.2695696502607453,
      "loss": 0.3503,
      "num_input_tokens_seen": 7154208,
      "step": 8255
    },
    {
      "epoch": 3.8943894389438944,
      "grad_norm": 0.013888959772884846,
      "learning_rate": 0.26953407390939504,
      "loss": 0.3341,
      "num_input_tokens_seen": 7158960,
      "step": 8260
    },
    {
      "epoch": 3.896746817538897,
      "grad_norm": 0.1465916484594345,
      "learning_rate": 0.26949847912438835,
      "loss": 0.326,
      "num_input_tokens_seen": 7164688,
      "step": 8265
    },
    {
      "epoch": 3.8991041961338992,
      "grad_norm": 0.007146978750824928,
      "learning_rate": 0.26946286591121454,
      "loss": 0.1702,
      "num_input_tokens_seen": 7168944,
      "step": 8270
    },
    {
      "epoch": 3.901461574728901,
      "grad_norm": 0.007197447121143341,
      "learning_rate": 0.2694272342753655,
      "loss": 0.5296,
      "num_input_tokens_seen": 7173648,
      "step": 8275
    },
    {
      "epoch": 3.9038189533239036,
      "grad_norm": 0.013607836328446865,
      "learning_rate": 0.26939158422233617,
      "loss": 0.2428,
      "num_input_tokens_seen": 7177888,
      "step": 8280
    },
    {
      "epoch": 3.906176331918906,
      "grad_norm": 0.012997593730688095,
      "learning_rate": 0.26935591575762413,
      "loss": 0.2243,
      "num_input_tokens_seen": 7181984,
      "step": 8285
    },
    {
      "epoch": 3.9085337105139084,
      "grad_norm": 0.02256922610104084,
      "learning_rate": 0.26932022888672996,
      "loss": 0.2657,
      "num_input_tokens_seen": 7186528,
      "step": 8290
    },
    {
      "epoch": 3.910891089108911,
      "grad_norm": 0.01380437146872282,
      "learning_rate": 0.26928452361515703,
      "loss": 0.4025,
      "num_input_tokens_seen": 7191296,
      "step": 8295
    },
    {
      "epoch": 3.913248467703913,
      "grad_norm": 0.016456304118037224,
      "learning_rate": 0.26924879994841155,
      "loss": 0.2605,
      "num_input_tokens_seen": 7194944,
      "step": 8300
    },
    {
      "epoch": 3.9156058462989156,
      "grad_norm": 0.011400873772799969,
      "learning_rate": 0.2692130578920025,
      "loss": 0.3116,
      "num_input_tokens_seen": 7198784,
      "step": 8305
    },
    {
      "epoch": 3.917963224893918,
      "grad_norm": 0.00675022229552269,
      "learning_rate": 0.26917729745144187,
      "loss": 0.2345,
      "num_input_tokens_seen": 7202752,
      "step": 8310
    },
    {
      "epoch": 3.9203206034889204,
      "grad_norm": 0.00681073684245348,
      "learning_rate": 0.2691415186322443,
      "loss": 0.2231,
      "num_input_tokens_seen": 7207680,
      "step": 8315
    },
    {
      "epoch": 3.922677982083923,
      "grad_norm": 0.007269169203937054,
      "learning_rate": 0.2691057214399273,
      "loss": 0.2017,
      "num_input_tokens_seen": 7211280,
      "step": 8320
    },
    {
      "epoch": 3.9250353606789252,
      "grad_norm": 0.01001354493200779,
      "learning_rate": 0.2690699058800113,
      "loss": 0.2628,
      "num_input_tokens_seen": 7215696,
      "step": 8325
    },
    {
      "epoch": 3.9273927392739276,
      "grad_norm": 0.017319487407803535,
      "learning_rate": 0.2690340719580194,
      "loss": 0.3087,
      "num_input_tokens_seen": 7219104,
      "step": 8330
    },
    {
      "epoch": 3.92975011786893,
      "grad_norm": 0.0163667444139719,
      "learning_rate": 0.2689982196794778,
      "loss": 0.2799,
      "num_input_tokens_seen": 7223504,
      "step": 8335
    },
    {
      "epoch": 3.932107496463932,
      "grad_norm": 0.008726874366402626,
      "learning_rate": 0.2689623490499153,
      "loss": 0.1567,
      "num_input_tokens_seen": 7227424,
      "step": 8340
    },
    {
      "epoch": 3.9344648750589344,
      "grad_norm": 0.014816123992204666,
      "learning_rate": 0.2689264600748636,
      "loss": 0.2103,
      "num_input_tokens_seen": 7231632,
      "step": 8345
    },
    {
      "epoch": 3.936822253653937,
      "grad_norm": 0.006456134375184774,
      "learning_rate": 0.26889055275985724,
      "loss": 0.1497,
      "num_input_tokens_seen": 7235712,
      "step": 8350
    },
    {
      "epoch": 3.939179632248939,
      "grad_norm": 0.00960117019712925,
      "learning_rate": 0.2688546271104335,
      "loss": 0.2162,
      "num_input_tokens_seen": 7240384,
      "step": 8355
    },
    {
      "epoch": 3.9415370108439416,
      "grad_norm": 0.012548311613500118,
      "learning_rate": 0.26881868313213275,
      "loss": 0.2795,
      "num_input_tokens_seen": 7244320,
      "step": 8360
    },
    {
      "epoch": 3.943894389438944,
      "grad_norm": 0.00907771848142147,
      "learning_rate": 0.2687827208304978,
      "loss": 0.258,
      "num_input_tokens_seen": 7248528,
      "step": 8365
    },
    {
      "epoch": 3.9462517680339464,
      "grad_norm": 0.008674316108226776,
      "learning_rate": 0.26874674021107464,
      "loss": 0.296,
      "num_input_tokens_seen": 7252352,
      "step": 8370
    },
    {
      "epoch": 3.9486091466289484,
      "grad_norm": 0.007676490116864443,
      "learning_rate": 0.2687107412794118,
      "loss": 0.2533,
      "num_input_tokens_seen": 7257152,
      "step": 8375
    },
    {
      "epoch": 3.9509665252239508,
      "grad_norm": 0.008198868483304977,
      "learning_rate": 0.26867472404106096,
      "loss": 0.2688,
      "num_input_tokens_seen": 7261136,
      "step": 8380
    },
    {
      "epoch": 3.953323903818953,
      "grad_norm": 0.01284998469054699,
      "learning_rate": 0.26863868850157624,
      "loss": 0.2246,
      "num_input_tokens_seen": 7265040,
      "step": 8385
    },
    {
      "epoch": 3.9556812824139556,
      "grad_norm": 0.01035616360604763,
      "learning_rate": 0.26860263466651485,
      "loss": 0.2298,
      "num_input_tokens_seen": 7269104,
      "step": 8390
    },
    {
      "epoch": 3.958038661008958,
      "grad_norm": 0.015500301495194435,
      "learning_rate": 0.26856656254143674,
      "loss": 0.2724,
      "num_input_tokens_seen": 7272512,
      "step": 8395
    },
    {
      "epoch": 3.9603960396039604,
      "grad_norm": 0.014138370752334595,
      "learning_rate": 0.2685304721319047,
      "loss": 0.264,
      "num_input_tokens_seen": 7276304,
      "step": 8400
    },
    {
      "epoch": 3.9603960396039604,
      "eval_loss": 0.29675740003585815,
      "eval_runtime": 21.8918,
      "eval_samples_per_second": 43.076,
      "eval_steps_per_second": 21.561,
      "num_input_tokens_seen": 7276304,
      "step": 8400
    },
    {
      "epoch": 3.9627534181989628,
      "grad_norm": 0.014727059751749039,
      "learning_rate": 0.2684943634434843,
      "loss": 0.3367,
      "num_input_tokens_seen": 7281440,
      "step": 8405
    },
    {
      "epoch": 3.965110796793965,
      "grad_norm": 0.007395980414003134,
      "learning_rate": 0.268458236481744,
      "loss": 0.2271,
      "num_input_tokens_seen": 7286192,
      "step": 8410
    },
    {
      "epoch": 3.9674681753889676,
      "grad_norm": 0.007851596921682358,
      "learning_rate": 0.2684220912522549,
      "loss": 0.2327,
      "num_input_tokens_seen": 7290224,
      "step": 8415
    },
    {
      "epoch": 3.96982555398397,
      "grad_norm": 0.005390584003180265,
      "learning_rate": 0.2683859277605913,
      "loss": 0.2506,
      "num_input_tokens_seen": 7294352,
      "step": 8420
    },
    {
      "epoch": 3.9721829325789724,
      "grad_norm": 0.006175574380904436,
      "learning_rate": 0.2683497460123298,
      "loss": 0.1862,
      "num_input_tokens_seen": 7299472,
      "step": 8425
    },
    {
      "epoch": 3.9745403111739748,
      "grad_norm": 0.010258338414132595,
      "learning_rate": 0.26831354601305013,
      "loss": 0.2085,
      "num_input_tokens_seen": 7303280,
      "step": 8430
    },
    {
      "epoch": 3.976897689768977,
      "grad_norm": 0.0073865302838385105,
      "learning_rate": 0.26827732776833496,
      "loss": 0.1748,
      "num_input_tokens_seen": 7307504,
      "step": 8435
    },
    {
      "epoch": 3.979255068363979,
      "grad_norm": 0.013866053894162178,
      "learning_rate": 0.26824109128376944,
      "loss": 0.2461,
      "num_input_tokens_seen": 7312192,
      "step": 8440
    },
    {
      "epoch": 3.9816124469589815,
      "grad_norm": 0.012387150898575783,
      "learning_rate": 0.2682048365649417,
      "loss": 0.2446,
      "num_input_tokens_seen": 7316352,
      "step": 8445
    },
    {
      "epoch": 3.983969825553984,
      "grad_norm": 0.01737266406416893,
      "learning_rate": 0.2681685636174428,
      "loss": 0.2515,
      "num_input_tokens_seen": 7321056,
      "step": 8450
    },
    {
      "epoch": 3.9863272041489863,
      "grad_norm": 0.016208503395318985,
      "learning_rate": 0.2681322724468663,
      "loss": 0.2272,
      "num_input_tokens_seen": 7324880,
      "step": 8455
    },
    {
      "epoch": 3.9886845827439887,
      "grad_norm": 0.013225015252828598,
      "learning_rate": 0.2680959630588089,
      "loss": 0.2416,
      "num_input_tokens_seen": 7328976,
      "step": 8460
    },
    {
      "epoch": 3.991041961338991,
      "grad_norm": 0.006945619825273752,
      "learning_rate": 0.26805963545886985,
      "loss": 0.2131,
      "num_input_tokens_seen": 7334784,
      "step": 8465
    },
    {
      "epoch": 3.9933993399339935,
      "grad_norm": 0.01150052435696125,
      "learning_rate": 0.26802328965265143,
      "loss": 0.3147,
      "num_input_tokens_seen": 7338528,
      "step": 8470
    },
    {
      "epoch": 3.9957567185289955,
      "grad_norm": 0.009082772769033909,
      "learning_rate": 0.26798692564575854,
      "loss": 0.2217,
      "num_input_tokens_seen": 7342560,
      "step": 8475
    },
    {
      "epoch": 3.998114097123998,
      "grad_norm": 0.00817931815981865,
      "learning_rate": 0.26795054344379904,
      "loss": 0.2414,
      "num_input_tokens_seen": 7346944,
      "step": 8480
    },
    {
      "epoch": 4.000471475719,
      "grad_norm": 0.005739480257034302,
      "learning_rate": 0.2679141430523835,
      "loss": 0.2225,
      "num_input_tokens_seen": 7350464,
      "step": 8485
    },
    {
      "epoch": 4.002828854314003,
      "grad_norm": 0.0051119523122906685,
      "learning_rate": 0.2678777244771252,
      "loss": 0.1819,
      "num_input_tokens_seen": 7354608,
      "step": 8490
    },
    {
      "epoch": 4.005186232909005,
      "grad_norm": 0.008346782065927982,
      "learning_rate": 0.2678412877236405,
      "loss": 0.2394,
      "num_input_tokens_seen": 7358480,
      "step": 8495
    },
    {
      "epoch": 4.0075436115040075,
      "grad_norm": 0.00833813101053238,
      "learning_rate": 0.2678048327975484,
      "loss": 0.2242,
      "num_input_tokens_seen": 7362608,
      "step": 8500
    },
    {
      "epoch": 4.00990099009901,
      "grad_norm": 0.005893635097891092,
      "learning_rate": 0.2677683597044706,
      "loss": 0.1832,
      "num_input_tokens_seen": 7367008,
      "step": 8505
    },
    {
      "epoch": 4.012258368694012,
      "grad_norm": 0.012804664671421051,
      "learning_rate": 0.2677318684500318,
      "loss": 0.2407,
      "num_input_tokens_seen": 7371376,
      "step": 8510
    },
    {
      "epoch": 4.014615747289015,
      "grad_norm": 0.008774391375482082,
      "learning_rate": 0.2676953590398593,
      "loss": 0.3487,
      "num_input_tokens_seen": 7375424,
      "step": 8515
    },
    {
      "epoch": 4.016973125884017,
      "grad_norm": 0.007615596055984497,
      "learning_rate": 0.2676588314795834,
      "loss": 0.2801,
      "num_input_tokens_seen": 7379472,
      "step": 8520
    },
    {
      "epoch": 4.0193305044790195,
      "grad_norm": 0.006788135506212711,
      "learning_rate": 0.26762228577483715,
      "loss": 0.2678,
      "num_input_tokens_seen": 7383408,
      "step": 8525
    },
    {
      "epoch": 4.021687883074022,
      "grad_norm": 0.0068372697569429874,
      "learning_rate": 0.2675857219312563,
      "loss": 0.1936,
      "num_input_tokens_seen": 7387280,
      "step": 8530
    },
    {
      "epoch": 4.024045261669024,
      "grad_norm": 0.008132080547511578,
      "learning_rate": 0.2675491399544794,
      "loss": 0.3034,
      "num_input_tokens_seen": 7391808,
      "step": 8535
    },
    {
      "epoch": 4.026402640264027,
      "grad_norm": 0.006972379982471466,
      "learning_rate": 0.2675125398501479,
      "loss": 0.2488,
      "num_input_tokens_seen": 7395728,
      "step": 8540
    },
    {
      "epoch": 4.028760018859029,
      "grad_norm": 0.0056379069574177265,
      "learning_rate": 0.26747592162390604,
      "loss": 0.1933,
      "num_input_tokens_seen": 7399760,
      "step": 8545
    },
    {
      "epoch": 4.0311173974540315,
      "grad_norm": 0.006394523195922375,
      "learning_rate": 0.26743928528140076,
      "loss": 0.2501,
      "num_input_tokens_seen": 7404896,
      "step": 8550
    },
    {
      "epoch": 4.033474776049033,
      "grad_norm": 0.004622991196811199,
      "learning_rate": 0.26740263082828186,
      "loss": 0.2193,
      "num_input_tokens_seen": 7408928,
      "step": 8555
    },
    {
      "epoch": 4.035832154644035,
      "grad_norm": 0.00964755192399025,
      "learning_rate": 0.2673659582702019,
      "loss": 0.2628,
      "num_input_tokens_seen": 7413392,
      "step": 8560
    },
    {
      "epoch": 4.038189533239038,
      "grad_norm": 0.007488161791115999,
      "learning_rate": 0.2673292676128163,
      "loss": 0.284,
      "num_input_tokens_seen": 7417424,
      "step": 8565
    },
    {
      "epoch": 4.04054691183404,
      "grad_norm": 0.007395235355943441,
      "learning_rate": 0.2672925588617831,
      "loss": 0.2138,
      "num_input_tokens_seen": 7422128,
      "step": 8570
    },
    {
      "epoch": 4.042904290429043,
      "grad_norm": 0.003505491651594639,
      "learning_rate": 0.2672558320227634,
      "loss": 0.2952,
      "num_input_tokens_seen": 7426512,
      "step": 8575
    },
    {
      "epoch": 4.045261669024045,
      "grad_norm": 0.005421205423772335,
      "learning_rate": 0.2672190871014209,
      "loss": 0.2716,
      "num_input_tokens_seen": 7430496,
      "step": 8580
    },
    {
      "epoch": 4.0476190476190474,
      "grad_norm": 0.005314737558364868,
      "learning_rate": 0.267182324103422,
      "loss": 0.2261,
      "num_input_tokens_seen": 7435552,
      "step": 8585
    },
    {
      "epoch": 4.04997642621405,
      "grad_norm": 0.006585229653865099,
      "learning_rate": 0.2671455430344362,
      "loss": 0.2543,
      "num_input_tokens_seen": 7439360,
      "step": 8590
    },
    {
      "epoch": 4.052333804809052,
      "grad_norm": 0.0045740315690636635,
      "learning_rate": 0.2671087439001355,
      "loss": 0.2301,
      "num_input_tokens_seen": 7443728,
      "step": 8595
    },
    {
      "epoch": 4.054691183404055,
      "grad_norm": 0.0055543044582009315,
      "learning_rate": 0.2670719267061948,
      "loss": 0.238,
      "num_input_tokens_seen": 7448112,
      "step": 8600
    },
    {
      "epoch": 4.054691183404055,
      "eval_loss": 0.22166672348976135,
      "eval_runtime": 21.9215,
      "eval_samples_per_second": 43.017,
      "eval_steps_per_second": 21.531,
      "num_input_tokens_seen": 7448112,
      "step": 8600
    },
    {
      "epoch": 4.057048561999057,
      "grad_norm": 0.003249850356951356,
      "learning_rate": 0.2670350914582918,
      "loss": 0.1931,
      "num_input_tokens_seen": 7452768,
      "step": 8605
    },
    {
      "epoch": 4.0594059405940595,
      "grad_norm": 0.010159089230000973,
      "learning_rate": 0.26699823816210694,
      "loss": 0.222,
      "num_input_tokens_seen": 7456800,
      "step": 8610
    },
    {
      "epoch": 4.061763319189062,
      "grad_norm": 0.004173879977315664,
      "learning_rate": 0.26696136682332344,
      "loss": 0.2533,
      "num_input_tokens_seen": 7460960,
      "step": 8615
    },
    {
      "epoch": 4.064120697784064,
      "grad_norm": 0.0071021756157279015,
      "learning_rate": 0.2669244774476274,
      "loss": 0.2223,
      "num_input_tokens_seen": 7465536,
      "step": 8620
    },
    {
      "epoch": 4.066478076379067,
      "grad_norm": 0.005173949524760246,
      "learning_rate": 0.2668875700407075,
      "loss": 0.2279,
      "num_input_tokens_seen": 7470400,
      "step": 8625
    },
    {
      "epoch": 4.068835454974069,
      "grad_norm": 0.004146794322878122,
      "learning_rate": 0.26685064460825547,
      "loss": 0.2217,
      "num_input_tokens_seen": 7474432,
      "step": 8630
    },
    {
      "epoch": 4.0711928335690715,
      "grad_norm": 0.006195591762661934,
      "learning_rate": 0.26681370115596553,
      "loss": 0.2398,
      "num_input_tokens_seen": 7478576,
      "step": 8635
    },
    {
      "epoch": 4.073550212164074,
      "grad_norm": 0.004962254781275988,
      "learning_rate": 0.26677673968953497,
      "loss": 0.2017,
      "num_input_tokens_seen": 7483120,
      "step": 8640
    },
    {
      "epoch": 4.075907590759076,
      "grad_norm": 0.010424282401800156,
      "learning_rate": 0.2667397602146636,
      "loss": 0.1808,
      "num_input_tokens_seen": 7487120,
      "step": 8645
    },
    {
      "epoch": 4.078264969354079,
      "grad_norm": 0.010652782395482063,
      "learning_rate": 0.2667027627370542,
      "loss": 0.2842,
      "num_input_tokens_seen": 7491776,
      "step": 8650
    },
    {
      "epoch": 4.08062234794908,
      "grad_norm": 0.005829522851854563,
      "learning_rate": 0.26666574726241216,
      "loss": 0.2031,
      "num_input_tokens_seen": 7496320,
      "step": 8655
    },
    {
      "epoch": 4.082979726544083,
      "grad_norm": 0.00436852453276515,
      "learning_rate": 0.2666287137964458,
      "loss": 0.2026,
      "num_input_tokens_seen": 7500144,
      "step": 8660
    },
    {
      "epoch": 4.085337105139085,
      "grad_norm": 0.0058376905508339405,
      "learning_rate": 0.26659166234486614,
      "loss": 0.2418,
      "num_input_tokens_seen": 7504432,
      "step": 8665
    },
    {
      "epoch": 4.087694483734087,
      "grad_norm": 0.002846485236659646,
      "learning_rate": 0.2665545929133869,
      "loss": 0.2202,
      "num_input_tokens_seen": 7508864,
      "step": 8670
    },
    {
      "epoch": 4.09005186232909,
      "grad_norm": 0.004907501861453056,
      "learning_rate": 0.2665175055077248,
      "loss": 0.2182,
      "num_input_tokens_seen": 7512544,
      "step": 8675
    },
    {
      "epoch": 4.092409240924092,
      "grad_norm": 0.0071626207791268826,
      "learning_rate": 0.2664804001335991,
      "loss": 0.2472,
      "num_input_tokens_seen": 7517328,
      "step": 8680
    },
    {
      "epoch": 4.094766619519095,
      "grad_norm": 0.004120876081287861,
      "learning_rate": 0.26644327679673185,
      "loss": 0.2148,
      "num_input_tokens_seen": 7521264,
      "step": 8685
    },
    {
      "epoch": 4.097123998114097,
      "grad_norm": 0.004130437038838863,
      "learning_rate": 0.26640613550284803,
      "loss": 0.2607,
      "num_input_tokens_seen": 7525824,
      "step": 8690
    },
    {
      "epoch": 4.099481376709099,
      "grad_norm": 0.0052009327337145805,
      "learning_rate": 0.26636897625767525,
      "loss": 0.2361,
      "num_input_tokens_seen": 7530672,
      "step": 8695
    },
    {
      "epoch": 4.101838755304102,
      "grad_norm": 0.005387585144490004,
      "learning_rate": 0.266331799066944,
      "loss": 0.2445,
      "num_input_tokens_seen": 7536336,
      "step": 8700
    },
    {
      "epoch": 4.104196133899104,
      "grad_norm": 0.007412449922412634,
      "learning_rate": 0.2662946039363874,
      "loss": 0.2239,
      "num_input_tokens_seen": 7540256,
      "step": 8705
    },
    {
      "epoch": 4.106553512494107,
      "grad_norm": 0.00426905881613493,
      "learning_rate": 0.2662573908717414,
      "loss": 0.2304,
      "num_input_tokens_seen": 7544160,
      "step": 8710
    },
    {
      "epoch": 4.108910891089109,
      "grad_norm": 0.0033299934584647417,
      "learning_rate": 0.2662201598787447,
      "loss": 0.2283,
      "num_input_tokens_seen": 7547904,
      "step": 8715
    },
    {
      "epoch": 4.111268269684111,
      "grad_norm": 0.007426364812999964,
      "learning_rate": 0.2661829109631389,
      "loss": 0.2298,
      "num_input_tokens_seen": 7551840,
      "step": 8720
    },
    {
      "epoch": 4.113625648279114,
      "grad_norm": 0.005416298285126686,
      "learning_rate": 0.26614564413066816,
      "loss": 0.2085,
      "num_input_tokens_seen": 7556896,
      "step": 8725
    },
    {
      "epoch": 4.115983026874116,
      "grad_norm": 0.005559032782912254,
      "learning_rate": 0.2661083593870795,
      "loss": 0.2377,
      "num_input_tokens_seen": 7561712,
      "step": 8730
    },
    {
      "epoch": 4.118340405469119,
      "grad_norm": 0.005237642675638199,
      "learning_rate": 0.26607105673812276,
      "loss": 0.2378,
      "num_input_tokens_seen": 7566528,
      "step": 8735
    },
    {
      "epoch": 4.120697784064121,
      "grad_norm": 0.0057446761056780815,
      "learning_rate": 0.2660337361895504,
      "loss": 0.2234,
      "num_input_tokens_seen": 7571664,
      "step": 8740
    },
    {
      "epoch": 4.123055162659123,
      "grad_norm": 0.0062489863485097885,
      "learning_rate": 0.26599639774711775,
      "loss": 0.2449,
      "num_input_tokens_seen": 7576048,
      "step": 8745
    },
    {
      "epoch": 4.125412541254126,
      "grad_norm": 0.004194039385765791,
      "learning_rate": 0.2659590414165829,
      "loss": 0.2086,
      "num_input_tokens_seen": 7581424,
      "step": 8750
    },
    {
      "epoch": 4.127769919849127,
      "grad_norm": 0.00473930174484849,
      "learning_rate": 0.2659216672037066,
      "loss": 0.2,
      "num_input_tokens_seen": 7585232,
      "step": 8755
    },
    {
      "epoch": 4.13012729844413,
      "grad_norm": 0.0042151473462581635,
      "learning_rate": 0.26588427511425244,
      "loss": 0.2943,
      "num_input_tokens_seen": 7588976,
      "step": 8760
    },
    {
      "epoch": 4.132484677039132,
      "grad_norm": 0.004632531199604273,
      "learning_rate": 0.26584686515398676,
      "loss": 0.2196,
      "num_input_tokens_seen": 7593120,
      "step": 8765
    },
    {
      "epoch": 4.1348420556341345,
      "grad_norm": 0.0064179552718997,
      "learning_rate": 0.2658094373286787,
      "loss": 0.2504,
      "num_input_tokens_seen": 7597488,
      "step": 8770
    },
    {
      "epoch": 4.137199434229137,
      "grad_norm": 0.003998386207967997,
      "learning_rate": 0.2657719916441,
      "loss": 0.2224,
      "num_input_tokens_seen": 7602464,
      "step": 8775
    },
    {
      "epoch": 4.139556812824139,
      "grad_norm": 0.00582402478903532,
      "learning_rate": 0.2657345281060253,
      "loss": 0.2189,
      "num_input_tokens_seen": 7606032,
      "step": 8780
    },
    {
      "epoch": 4.141914191419142,
      "grad_norm": 0.006836670450866222,
      "learning_rate": 0.26569704672023203,
      "loss": 0.3364,
      "num_input_tokens_seen": 7610416,
      "step": 8785
    },
    {
      "epoch": 4.144271570014144,
      "grad_norm": 0.009202091954648495,
      "learning_rate": 0.26565954749250015,
      "loss": 0.1929,
      "num_input_tokens_seen": 7614896,
      "step": 8790
    },
    {
      "epoch": 4.1466289486091465,
      "grad_norm": 0.013122500851750374,
      "learning_rate": 0.2656220304286126,
      "loss": 0.3037,
      "num_input_tokens_seen": 7619440,
      "step": 8795
    },
    {
      "epoch": 4.148986327204149,
      "grad_norm": 0.006890695542097092,
      "learning_rate": 0.265584495534355,
      "loss": 0.2153,
      "num_input_tokens_seen": 7623632,
      "step": 8800
    },
    {
      "epoch": 4.148986327204149,
      "eval_loss": 0.23249080777168274,
      "eval_runtime": 21.8923,
      "eval_samples_per_second": 43.075,
      "eval_steps_per_second": 21.56,
      "num_input_tokens_seen": 7623632,
      "step": 8800
    },
    {
      "epoch": 4.151343705799151,
      "grad_norm": 0.005266821477562189,
      "learning_rate": 0.2655469428155156,
      "loss": 0.2366,
      "num_input_tokens_seen": 7627584,
      "step": 8805
    },
    {
      "epoch": 4.153701084394154,
      "grad_norm": 0.004813721869140863,
      "learning_rate": 0.2655093722778856,
      "loss": 0.2195,
      "num_input_tokens_seen": 7631888,
      "step": 8810
    },
    {
      "epoch": 4.156058462989156,
      "grad_norm": 0.003676759544759989,
      "learning_rate": 0.2654717839272588,
      "loss": 0.2348,
      "num_input_tokens_seen": 7636096,
      "step": 8815
    },
    {
      "epoch": 4.158415841584159,
      "grad_norm": 0.006451107095927,
      "learning_rate": 0.2654341777694318,
      "loss": 0.2527,
      "num_input_tokens_seen": 7639424,
      "step": 8820
    },
    {
      "epoch": 4.160773220179161,
      "grad_norm": 0.0058168284595012665,
      "learning_rate": 0.265396553810204,
      "loss": 0.2548,
      "num_input_tokens_seen": 7643392,
      "step": 8825
    },
    {
      "epoch": 4.163130598774163,
      "grad_norm": 0.023541085422039032,
      "learning_rate": 0.26535891205537737,
      "loss": 0.2505,
      "num_input_tokens_seen": 7648496,
      "step": 8830
    },
    {
      "epoch": 4.165487977369166,
      "grad_norm": 0.005607908125966787,
      "learning_rate": 0.26532125251075683,
      "loss": 0.2364,
      "num_input_tokens_seen": 7652768,
      "step": 8835
    },
    {
      "epoch": 4.167845355964168,
      "grad_norm": 0.007601615507155657,
      "learning_rate": 0.26528357518214996,
      "loss": 0.2246,
      "num_input_tokens_seen": 7657312,
      "step": 8840
    },
    {
      "epoch": 4.170202734559171,
      "grad_norm": 0.0064124311320483685,
      "learning_rate": 0.26524588007536704,
      "loss": 0.2606,
      "num_input_tokens_seen": 7661824,
      "step": 8845
    },
    {
      "epoch": 4.172560113154173,
      "grad_norm": 0.0065261246636509895,
      "learning_rate": 0.26520816719622115,
      "loss": 0.2169,
      "num_input_tokens_seen": 7665680,
      "step": 8850
    },
    {
      "epoch": 4.174917491749175,
      "grad_norm": 0.007443987298756838,
      "learning_rate": 0.2651704365505281,
      "loss": 0.2304,
      "num_input_tokens_seen": 7670240,
      "step": 8855
    },
    {
      "epoch": 4.177274870344177,
      "grad_norm": 0.007404308300465345,
      "learning_rate": 0.26513268814410634,
      "loss": 0.2396,
      "num_input_tokens_seen": 7674256,
      "step": 8860
    },
    {
      "epoch": 4.179632248939179,
      "grad_norm": 0.004552786238491535,
      "learning_rate": 0.2650949219827773,
      "loss": 0.2485,
      "num_input_tokens_seen": 7678768,
      "step": 8865
    },
    {
      "epoch": 4.181989627534182,
      "grad_norm": 0.006495942827314138,
      "learning_rate": 0.26505713807236486,
      "loss": 0.2535,
      "num_input_tokens_seen": 7682800,
      "step": 8870
    },
    {
      "epoch": 4.184347006129184,
      "grad_norm": 0.043057870119810104,
      "learning_rate": 0.26501933641869585,
      "loss": 0.2321,
      "num_input_tokens_seen": 7687296,
      "step": 8875
    },
    {
      "epoch": 4.1867043847241865,
      "grad_norm": 0.0072715566493570805,
      "learning_rate": 0.26498151702759976,
      "loss": 0.2309,
      "num_input_tokens_seen": 7691568,
      "step": 8880
    },
    {
      "epoch": 4.189061763319189,
      "grad_norm": 0.0038313947152346373,
      "learning_rate": 0.2649436799049088,
      "loss": 0.2061,
      "num_input_tokens_seen": 7695616,
      "step": 8885
    },
    {
      "epoch": 4.191419141914191,
      "grad_norm": 0.006389549467712641,
      "learning_rate": 0.2649058250564579,
      "loss": 0.2271,
      "num_input_tokens_seen": 7700144,
      "step": 8890
    },
    {
      "epoch": 4.193776520509194,
      "grad_norm": 0.013330805115401745,
      "learning_rate": 0.26486795248808476,
      "loss": 0.2588,
      "num_input_tokens_seen": 7705216,
      "step": 8895
    },
    {
      "epoch": 4.196133899104196,
      "grad_norm": 0.008232085965573788,
      "learning_rate": 0.2648300622056298,
      "loss": 0.2501,
      "num_input_tokens_seen": 7709936,
      "step": 8900
    },
    {
      "epoch": 4.1984912776991985,
      "grad_norm": 0.008924928493797779,
      "learning_rate": 0.2647921542149363,
      "loss": 0.2461,
      "num_input_tokens_seen": 7714416,
      "step": 8905
    },
    {
      "epoch": 4.200848656294201,
      "grad_norm": 0.006211942061781883,
      "learning_rate": 0.26475422852185,
      "loss": 0.2365,
      "num_input_tokens_seen": 7718208,
      "step": 8910
    },
    {
      "epoch": 4.203206034889203,
      "grad_norm": 0.005423965398222208,
      "learning_rate": 0.2647162851322196,
      "loss": 0.2159,
      "num_input_tokens_seen": 7723008,
      "step": 8915
    },
    {
      "epoch": 4.205563413484206,
      "grad_norm": 0.0064941635355353355,
      "learning_rate": 0.2646783240518964,
      "loss": 0.2626,
      "num_input_tokens_seen": 7726864,
      "step": 8920
    },
    {
      "epoch": 4.207920792079208,
      "grad_norm": 0.005787142086774111,
      "learning_rate": 0.26464034528673447,
      "loss": 0.2282,
      "num_input_tokens_seen": 7731872,
      "step": 8925
    },
    {
      "epoch": 4.2102781706742105,
      "grad_norm": 0.007065621670335531,
      "learning_rate": 0.26460234884259065,
      "loss": 0.242,
      "num_input_tokens_seen": 7736864,
      "step": 8930
    },
    {
      "epoch": 4.212635549269213,
      "grad_norm": 0.005654407199472189,
      "learning_rate": 0.2645643347253245,
      "loss": 0.1892,
      "num_input_tokens_seen": 7740992,
      "step": 8935
    },
    {
      "epoch": 4.214992927864215,
      "grad_norm": 0.003399886889383197,
      "learning_rate": 0.2645263029407982,
      "loss": 0.1955,
      "num_input_tokens_seen": 7745168,
      "step": 8940
    },
    {
      "epoch": 4.217350306459218,
      "grad_norm": 0.00637801131233573,
      "learning_rate": 0.2644882534948767,
      "loss": 0.2852,
      "num_input_tokens_seen": 7749008,
      "step": 8945
    },
    {
      "epoch": 4.21970768505422,
      "grad_norm": 0.007995901629328728,
      "learning_rate": 0.2644501863934278,
      "loss": 0.2239,
      "num_input_tokens_seen": 7754176,
      "step": 8950
    },
    {
      "epoch": 4.222065063649222,
      "grad_norm": 0.006450950633734465,
      "learning_rate": 0.26441210164232193,
      "loss": 0.2259,
      "num_input_tokens_seen": 7759216,
      "step": 8955
    },
    {
      "epoch": 4.224422442244224,
      "grad_norm": 0.011865093372762203,
      "learning_rate": 0.26437399924743216,
      "loss": 0.2312,
      "num_input_tokens_seen": 7763376,
      "step": 8960
    },
    {
      "epoch": 4.226779820839226,
      "grad_norm": 0.004237552173435688,
      "learning_rate": 0.26433587921463436,
      "loss": 0.2051,
      "num_input_tokens_seen": 7767584,
      "step": 8965
    },
    {
      "epoch": 4.229137199434229,
      "grad_norm": 0.006064979825168848,
      "learning_rate": 0.2642977415498072,
      "loss": 0.2225,
      "num_input_tokens_seen": 7772192,
      "step": 8970
    },
    {
      "epoch": 4.231494578029231,
      "grad_norm": 0.003420487279072404,
      "learning_rate": 0.26425958625883195,
      "loss": 0.2387,
      "num_input_tokens_seen": 7776000,
      "step": 8975
    },
    {
      "epoch": 4.233851956624234,
      "grad_norm": 0.006010773591697216,
      "learning_rate": 0.2642214133475926,
      "loss": 0.2304,
      "num_input_tokens_seen": 7780272,
      "step": 8980
    },
    {
      "epoch": 4.236209335219236,
      "grad_norm": 0.006011554505676031,
      "learning_rate": 0.26418322282197587,
      "loss": 0.2068,
      "num_input_tokens_seen": 7784992,
      "step": 8985
    },
    {
      "epoch": 4.238566713814238,
      "grad_norm": 0.006112040486186743,
      "learning_rate": 0.2641450146878714,
      "loss": 0.2522,
      "num_input_tokens_seen": 7789888,
      "step": 8990
    },
    {
      "epoch": 4.240924092409241,
      "grad_norm": 0.005789645481854677,
      "learning_rate": 0.26410678895117107,
      "loss": 0.2106,
      "num_input_tokens_seen": 7794464,
      "step": 8995
    },
    {
      "epoch": 4.243281471004243,
      "grad_norm": 0.003842594102025032,
      "learning_rate": 0.26406854561777,
      "loss": 0.1812,
      "num_input_tokens_seen": 7799248,
      "step": 9000
    },
    {
      "epoch": 4.243281471004243,
      "eval_loss": 0.2273428738117218,
      "eval_runtime": 21.9331,
      "eval_samples_per_second": 42.994,
      "eval_steps_per_second": 21.52,
      "num_input_tokens_seen": 7799248,
      "step": 9000
    },
    {
      "epoch": 4.245638849599246,
      "grad_norm": 0.003607124090194702,
      "learning_rate": 0.26403028469356576,
      "loss": 0.2979,
      "num_input_tokens_seen": 7803184,
      "step": 9005
    },
    {
      "epoch": 4.247996228194248,
      "grad_norm": 0.007619247771799564,
      "learning_rate": 0.2639920061844585,
      "loss": 0.2008,
      "num_input_tokens_seen": 7808064,
      "step": 9010
    },
    {
      "epoch": 4.2503536067892504,
      "grad_norm": 0.00533232931047678,
      "learning_rate": 0.2639537100963515,
      "loss": 0.2172,
      "num_input_tokens_seen": 7812352,
      "step": 9015
    },
    {
      "epoch": 4.252710985384253,
      "grad_norm": 0.0036311037838459015,
      "learning_rate": 0.26391539643515033,
      "loss": 0.239,
      "num_input_tokens_seen": 7816400,
      "step": 9020
    },
    {
      "epoch": 4.255068363979255,
      "grad_norm": 0.005975732579827309,
      "learning_rate": 0.26387706520676346,
      "loss": 0.2182,
      "num_input_tokens_seen": 7819936,
      "step": 9025
    },
    {
      "epoch": 4.257425742574258,
      "grad_norm": 0.004812866915017366,
      "learning_rate": 0.26383871641710205,
      "loss": 0.25,
      "num_input_tokens_seen": 7823984,
      "step": 9030
    },
    {
      "epoch": 4.25978312116926,
      "grad_norm": 0.004944560118019581,
      "learning_rate": 0.26380035007208,
      "loss": 0.233,
      "num_input_tokens_seen": 7828624,
      "step": 9035
    },
    {
      "epoch": 4.2621404997642625,
      "grad_norm": 0.014386586844921112,
      "learning_rate": 0.26376196617761394,
      "loss": 0.2312,
      "num_input_tokens_seen": 7833136,
      "step": 9040
    },
    {
      "epoch": 4.264497878359265,
      "grad_norm": 0.005070526152849197,
      "learning_rate": 0.263723564739623,
      "loss": 0.2349,
      "num_input_tokens_seen": 7837360,
      "step": 9045
    },
    {
      "epoch": 4.266855256954267,
      "grad_norm": 0.0032284704502671957,
      "learning_rate": 0.2636851457640293,
      "loss": 0.2279,
      "num_input_tokens_seen": 7841408,
      "step": 9050
    },
    {
      "epoch": 4.26921263554927,
      "grad_norm": 0.0033750662114471197,
      "learning_rate": 0.26364670925675737,
      "loss": 0.1933,
      "num_input_tokens_seen": 7845056,
      "step": 9055
    },
    {
      "epoch": 4.271570014144271,
      "grad_norm": 0.0025347196497023106,
      "learning_rate": 0.2636082552237347,
      "loss": 0.2461,
      "num_input_tokens_seen": 7850144,
      "step": 9060
    },
    {
      "epoch": 4.273927392739274,
      "grad_norm": 0.0034134548623114824,
      "learning_rate": 0.26356978367089146,
      "loss": 0.2183,
      "num_input_tokens_seen": 7855024,
      "step": 9065
    },
    {
      "epoch": 4.276284771334276,
      "grad_norm": 0.009256832301616669,
      "learning_rate": 0.26353129460416036,
      "loss": 0.2574,
      "num_input_tokens_seen": 7859680,
      "step": 9070
    },
    {
      "epoch": 4.278642149929278,
      "grad_norm": 0.006364412605762482,
      "learning_rate": 0.2634927880294769,
      "loss": 0.2227,
      "num_input_tokens_seen": 7863776,
      "step": 9075
    },
    {
      "epoch": 4.280999528524281,
      "grad_norm": 0.006698859389871359,
      "learning_rate": 0.26345426395277927,
      "loss": 0.2382,
      "num_input_tokens_seen": 7868192,
      "step": 9080
    },
    {
      "epoch": 4.283356907119283,
      "grad_norm": 0.0031029193196445704,
      "learning_rate": 0.2634157223800084,
      "loss": 0.2241,
      "num_input_tokens_seen": 7872720,
      "step": 9085
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.0047925915569067,
      "learning_rate": 0.26337716331710787,
      "loss": 0.2237,
      "num_input_tokens_seen": 7876880,
      "step": 9090
    },
    {
      "epoch": 4.288071664309288,
      "grad_norm": 0.0026141153648495674,
      "learning_rate": 0.2633385867700239,
      "loss": 0.2333,
      "num_input_tokens_seen": 7881184,
      "step": 9095
    },
    {
      "epoch": 4.29042904290429,
      "grad_norm": 0.005250149872153997,
      "learning_rate": 0.2632999927447056,
      "loss": 0.2601,
      "num_input_tokens_seen": 7886144,
      "step": 9100
    },
    {
      "epoch": 4.292786421499293,
      "grad_norm": 0.004535315558314323,
      "learning_rate": 0.2632613812471046,
      "loss": 0.2493,
      "num_input_tokens_seen": 7890272,
      "step": 9105
    },
    {
      "epoch": 4.295143800094295,
      "grad_norm": 0.0031777212861925364,
      "learning_rate": 0.2632227522831753,
      "loss": 0.207,
      "num_input_tokens_seen": 7894112,
      "step": 9110
    },
    {
      "epoch": 4.297501178689298,
      "grad_norm": 0.003022312419489026,
      "learning_rate": 0.26318410585887475,
      "loss": 0.2334,
      "num_input_tokens_seen": 7897920,
      "step": 9115
    },
    {
      "epoch": 4.2998585572843,
      "grad_norm": 0.004160049371421337,
      "learning_rate": 0.2631454419801627,
      "loss": 0.2082,
      "num_input_tokens_seen": 7901888,
      "step": 9120
    },
    {
      "epoch": 4.302215935879302,
      "grad_norm": 0.004063785076141357,
      "learning_rate": 0.2631067606530016,
      "loss": 0.3817,
      "num_input_tokens_seen": 7906512,
      "step": 9125
    },
    {
      "epoch": 4.304573314474305,
      "grad_norm": 0.0037633036263287067,
      "learning_rate": 0.2630680618833567,
      "loss": 0.2283,
      "num_input_tokens_seen": 7910848,
      "step": 9130
    },
    {
      "epoch": 4.306930693069307,
      "grad_norm": 0.006465422920882702,
      "learning_rate": 0.26302934567719566,
      "loss": 0.2735,
      "num_input_tokens_seen": 7915536,
      "step": 9135
    },
    {
      "epoch": 4.30928807166431,
      "grad_norm": 0.002964041195809841,
      "learning_rate": 0.2629906120404892,
      "loss": 0.2263,
      "num_input_tokens_seen": 7919904,
      "step": 9140
    },
    {
      "epoch": 4.311645450259312,
      "grad_norm": 0.0063500129617750645,
      "learning_rate": 0.26295186097921036,
      "loss": 0.237,
      "num_input_tokens_seen": 7925472,
      "step": 9145
    },
    {
      "epoch": 4.314002828854314,
      "grad_norm": 0.007083000149577856,
      "learning_rate": 0.2629130924993351,
      "loss": 0.2339,
      "num_input_tokens_seen": 7929152,
      "step": 9150
    },
    {
      "epoch": 4.316360207449316,
      "grad_norm": 0.005854646675288677,
      "learning_rate": 0.2628743066068421,
      "loss": 0.2349,
      "num_input_tokens_seen": 7933408,
      "step": 9155
    },
    {
      "epoch": 4.318717586044318,
      "grad_norm": 0.0035294583067297935,
      "learning_rate": 0.26283550330771244,
      "loss": 0.2103,
      "num_input_tokens_seen": 7937584,
      "step": 9160
    },
    {
      "epoch": 4.321074964639321,
      "grad_norm": 0.00582813611254096,
      "learning_rate": 0.2627966826079303,
      "loss": 0.2359,
      "num_input_tokens_seen": 7941616,
      "step": 9165
    },
    {
      "epoch": 4.323432343234323,
      "grad_norm": 0.004016651306301355,
      "learning_rate": 0.26275784451348216,
      "loss": 0.2085,
      "num_input_tokens_seen": 7946288,
      "step": 9170
    },
    {
      "epoch": 4.3257897218293255,
      "grad_norm": 0.0025169069413095713,
      "learning_rate": 0.2627189890303574,
      "loss": 0.1711,
      "num_input_tokens_seen": 7950496,
      "step": 9175
    },
    {
      "epoch": 4.328147100424328,
      "grad_norm": 0.002617758233100176,
      "learning_rate": 0.262680116164548,
      "loss": 0.2152,
      "num_input_tokens_seen": 7955056,
      "step": 9180
    },
    {
      "epoch": 4.33050447901933,
      "grad_norm": 0.004490070976316929,
      "learning_rate": 0.2626412259220487,
      "loss": 0.3224,
      "num_input_tokens_seen": 7959520,
      "step": 9185
    },
    {
      "epoch": 4.332861857614333,
      "grad_norm": 0.0029117155354470015,
      "learning_rate": 0.2626023183088568,
      "loss": 0.232,
      "num_input_tokens_seen": 7964944,
      "step": 9190
    },
    {
      "epoch": 4.335219236209335,
      "grad_norm": 0.005904405377805233,
      "learning_rate": 0.26256339333097234,
      "loss": 0.257,
      "num_input_tokens_seen": 7969600,
      "step": 9195
    },
    {
      "epoch": 4.3375766148043375,
      "grad_norm": 0.0056585026904940605,
      "learning_rate": 0.2625244509943981,
      "loss": 0.2833,
      "num_input_tokens_seen": 7974368,
      "step": 9200
    },
    {
      "epoch": 4.3375766148043375,
      "eval_loss": 0.2250385880470276,
      "eval_runtime": 21.8981,
      "eval_samples_per_second": 43.063,
      "eval_steps_per_second": 21.554,
      "num_input_tokens_seen": 7974368,
      "step": 9200
    },
    {
      "epoch": 4.33993399339934,
      "grad_norm": 0.0029976810328662395,
      "learning_rate": 0.2624854913051395,
      "loss": 0.2323,
      "num_input_tokens_seen": 7979056,
      "step": 9205
    },
    {
      "epoch": 4.342291371994342,
      "grad_norm": 0.0024618261959403753,
      "learning_rate": 0.26244651426920446,
      "loss": 0.1943,
      "num_input_tokens_seen": 7982976,
      "step": 9210
    },
    {
      "epoch": 4.344648750589345,
      "grad_norm": 0.003350294893607497,
      "learning_rate": 0.26240751989260386,
      "loss": 0.2648,
      "num_input_tokens_seen": 7987152,
      "step": 9215
    },
    {
      "epoch": 4.347006129184347,
      "grad_norm": 0.00448601134121418,
      "learning_rate": 0.2623685081813511,
      "loss": 0.2773,
      "num_input_tokens_seen": 7991024,
      "step": 9220
    },
    {
      "epoch": 4.3493635077793495,
      "grad_norm": 0.003992486745119095,
      "learning_rate": 0.2623294791414623,
      "loss": 0.2521,
      "num_input_tokens_seen": 7995344,
      "step": 9225
    },
    {
      "epoch": 4.351720886374352,
      "grad_norm": 0.0038789138197898865,
      "learning_rate": 0.26229043277895614,
      "loss": 0.2408,
      "num_input_tokens_seen": 7999584,
      "step": 9230
    },
    {
      "epoch": 4.354078264969354,
      "grad_norm": 0.003746961010619998,
      "learning_rate": 0.2622513690998542,
      "loss": 0.2147,
      "num_input_tokens_seen": 8003920,
      "step": 9235
    },
    {
      "epoch": 4.356435643564357,
      "grad_norm": 0.004681269638240337,
      "learning_rate": 0.26221228811018044,
      "loss": 0.2417,
      "num_input_tokens_seen": 8008032,
      "step": 9240
    },
    {
      "epoch": 4.358793022159359,
      "grad_norm": 0.005208021495491266,
      "learning_rate": 0.2621731898159617,
      "loss": 0.2336,
      "num_input_tokens_seen": 8013088,
      "step": 9245
    },
    {
      "epoch": 4.3611504007543616,
      "grad_norm": 0.0037083022762089968,
      "learning_rate": 0.26213407422322743,
      "loss": 0.2163,
      "num_input_tokens_seen": 8018240,
      "step": 9250
    },
    {
      "epoch": 4.363507779349364,
      "grad_norm": 0.0032967613078653812,
      "learning_rate": 0.2620949413380098,
      "loss": 0.2131,
      "num_input_tokens_seen": 8022032,
      "step": 9255
    },
    {
      "epoch": 4.3658651579443655,
      "grad_norm": 0.003296947106719017,
      "learning_rate": 0.26205579116634353,
      "loss": 0.2466,
      "num_input_tokens_seen": 8026400,
      "step": 9260
    },
    {
      "epoch": 4.368222536539368,
      "grad_norm": 0.0055676912888884544,
      "learning_rate": 0.26201662371426604,
      "loss": 0.191,
      "num_input_tokens_seen": 8030672,
      "step": 9265
    },
    {
      "epoch": 4.37057991513437,
      "grad_norm": 0.003398091997951269,
      "learning_rate": 0.2619774389878175,
      "loss": 0.1649,
      "num_input_tokens_seen": 8035136,
      "step": 9270
    },
    {
      "epoch": 4.372937293729373,
      "grad_norm": 0.001919765374623239,
      "learning_rate": 0.2619382369930407,
      "loss": 0.279,
      "num_input_tokens_seen": 8039152,
      "step": 9275
    },
    {
      "epoch": 4.375294672324375,
      "grad_norm": 0.00324031850323081,
      "learning_rate": 0.261899017735981,
      "loss": 0.2124,
      "num_input_tokens_seen": 8043344,
      "step": 9280
    },
    {
      "epoch": 4.3776520509193775,
      "grad_norm": 0.005271621979773045,
      "learning_rate": 0.2618597812226866,
      "loss": 0.2193,
      "num_input_tokens_seen": 8047536,
      "step": 9285
    },
    {
      "epoch": 4.38000942951438,
      "grad_norm": 0.005476982332766056,
      "learning_rate": 0.2618205274592082,
      "loss": 0.2479,
      "num_input_tokens_seen": 8051920,
      "step": 9290
    },
    {
      "epoch": 4.382366808109382,
      "grad_norm": 0.004577161278575659,
      "learning_rate": 0.2617812564515992,
      "loss": 0.2462,
      "num_input_tokens_seen": 8056416,
      "step": 9295
    },
    {
      "epoch": 4.384724186704385,
      "grad_norm": 0.002623429987579584,
      "learning_rate": 0.2617419682059158,
      "loss": 0.2454,
      "num_input_tokens_seen": 8060576,
      "step": 9300
    },
    {
      "epoch": 4.387081565299387,
      "grad_norm": 0.003817977849394083,
      "learning_rate": 0.26170266272821663,
      "loss": 0.2118,
      "num_input_tokens_seen": 8065024,
      "step": 9305
    },
    {
      "epoch": 4.3894389438943895,
      "grad_norm": 0.0037288011517375708,
      "learning_rate": 0.26166334002456315,
      "loss": 0.2342,
      "num_input_tokens_seen": 8070416,
      "step": 9310
    },
    {
      "epoch": 4.391796322489392,
      "grad_norm": 0.0017615173710510135,
      "learning_rate": 0.2616240001010194,
      "loss": 0.2295,
      "num_input_tokens_seen": 8074608,
      "step": 9315
    },
    {
      "epoch": 4.394153701084394,
      "grad_norm": 0.004655566532164812,
      "learning_rate": 0.26158464296365197,
      "loss": 0.2591,
      "num_input_tokens_seen": 8078976,
      "step": 9320
    },
    {
      "epoch": 4.396511079679397,
      "grad_norm": 0.0037334789521992207,
      "learning_rate": 0.2615452686185304,
      "loss": 0.2048,
      "num_input_tokens_seen": 8082880,
      "step": 9325
    },
    {
      "epoch": 4.398868458274399,
      "grad_norm": 0.0060518281534314156,
      "learning_rate": 0.26150587707172673,
      "loss": 0.2288,
      "num_input_tokens_seen": 8086144,
      "step": 9330
    },
    {
      "epoch": 4.4012258368694015,
      "grad_norm": 0.004281120374798775,
      "learning_rate": 0.2614664683293154,
      "loss": 0.2206,
      "num_input_tokens_seen": 8090960,
      "step": 9335
    },
    {
      "epoch": 4.403583215464404,
      "grad_norm": 0.00545575050637126,
      "learning_rate": 0.26142704239737397,
      "loss": 0.2332,
      "num_input_tokens_seen": 8095072,
      "step": 9340
    },
    {
      "epoch": 4.405940594059406,
      "grad_norm": 0.002180506708100438,
      "learning_rate": 0.26138759928198235,
      "loss": 0.2409,
      "num_input_tokens_seen": 8099248,
      "step": 9345
    },
    {
      "epoch": 4.408297972654409,
      "grad_norm": 0.0035082832910120487,
      "learning_rate": 0.26134813898922304,
      "loss": 0.2314,
      "num_input_tokens_seen": 8103616,
      "step": 9350
    },
    {
      "epoch": 4.41065535124941,
      "grad_norm": 0.00253653759136796,
      "learning_rate": 0.26130866152518145,
      "loss": 0.2184,
      "num_input_tokens_seen": 8107648,
      "step": 9355
    },
    {
      "epoch": 4.413012729844413,
      "grad_norm": 0.0032756056170910597,
      "learning_rate": 0.2612691668959455,
      "loss": 0.2087,
      "num_input_tokens_seen": 8112080,
      "step": 9360
    },
    {
      "epoch": 4.415370108439415,
      "grad_norm": 0.00418089097365737,
      "learning_rate": 0.2612296551076057,
      "loss": 0.2375,
      "num_input_tokens_seen": 8116784,
      "step": 9365
    },
    {
      "epoch": 4.417727487034417,
      "grad_norm": 0.0038985793944448233,
      "learning_rate": 0.26119012616625525,
      "loss": 0.1829,
      "num_input_tokens_seen": 8120880,
      "step": 9370
    },
    {
      "epoch": 4.42008486562942,
      "grad_norm": 0.002612902782857418,
      "learning_rate": 0.26115058007799,
      "loss": 0.2374,
      "num_input_tokens_seen": 8125120,
      "step": 9375
    },
    {
      "epoch": 4.422442244224422,
      "grad_norm": 0.003643402364104986,
      "learning_rate": 0.26111101684890864,
      "loss": 0.2426,
      "num_input_tokens_seen": 8129088,
      "step": 9380
    },
    {
      "epoch": 4.424799622819425,
      "grad_norm": 0.00430586701259017,
      "learning_rate": 0.26107143648511205,
      "loss": 0.231,
      "num_input_tokens_seen": 8134192,
      "step": 9385
    },
    {
      "epoch": 4.427157001414427,
      "grad_norm": 0.004517610650509596,
      "learning_rate": 0.2610318389927042,
      "loss": 0.2043,
      "num_input_tokens_seen": 8138080,
      "step": 9390
    },
    {
      "epoch": 4.429514380009429,
      "grad_norm": 0.005822936072945595,
      "learning_rate": 0.26099222437779146,
      "loss": 0.2243,
      "num_input_tokens_seen": 8142304,
      "step": 9395
    },
    {
      "epoch": 4.431871758604432,
      "grad_norm": 0.005944953765720129,
      "learning_rate": 0.26095259264648285,
      "loss": 0.2609,
      "num_input_tokens_seen": 8146384,
      "step": 9400
    },
    {
      "epoch": 4.431871758604432,
      "eval_loss": 0.22689031064510345,
      "eval_runtime": 21.9136,
      "eval_samples_per_second": 43.033,
      "eval_steps_per_second": 21.539,
      "num_input_tokens_seen": 8146384,
      "step": 9400
    },
    {
      "epoch": 4.434229137199434,
      "grad_norm": 0.004055943340063095,
      "learning_rate": 0.2609129438048902,
      "loss": 0.2265,
      "num_input_tokens_seen": 8150448,
      "step": 9405
    },
    {
      "epoch": 4.436586515794437,
      "grad_norm": 0.0041501265950500965,
      "learning_rate": 0.2608732778591278,
      "loss": 0.2223,
      "num_input_tokens_seen": 8154784,
      "step": 9410
    },
    {
      "epoch": 4.438943894389439,
      "grad_norm": 0.004107818007469177,
      "learning_rate": 0.2608335948153126,
      "loss": 0.2159,
      "num_input_tokens_seen": 8159008,
      "step": 9415
    },
    {
      "epoch": 4.441301272984441,
      "grad_norm": 0.004072488751262426,
      "learning_rate": 0.26079389467956426,
      "loss": 0.2152,
      "num_input_tokens_seen": 8163488,
      "step": 9420
    },
    {
      "epoch": 4.443658651579444,
      "grad_norm": 0.003146030707284808,
      "learning_rate": 0.26075417745800505,
      "loss": 0.2487,
      "num_input_tokens_seen": 8167440,
      "step": 9425
    },
    {
      "epoch": 4.446016030174446,
      "grad_norm": 0.004799408372491598,
      "learning_rate": 0.26071444315675985,
      "loss": 0.2174,
      "num_input_tokens_seen": 8172336,
      "step": 9430
    },
    {
      "epoch": 4.448373408769449,
      "grad_norm": 0.0035962937399744987,
      "learning_rate": 0.2606746917819562,
      "loss": 0.1783,
      "num_input_tokens_seen": 8176160,
      "step": 9435
    },
    {
      "epoch": 4.450730787364451,
      "grad_norm": 0.00557676050812006,
      "learning_rate": 0.2606349233397242,
      "loss": 0.2094,
      "num_input_tokens_seen": 8179952,
      "step": 9440
    },
    {
      "epoch": 4.4530881659594534,
      "grad_norm": 0.00624077906832099,
      "learning_rate": 0.26059513783619676,
      "loss": 0.2379,
      "num_input_tokens_seen": 8184048,
      "step": 9445
    },
    {
      "epoch": 4.455445544554456,
      "grad_norm": 0.010251578874886036,
      "learning_rate": 0.26055533527750924,
      "loss": 0.255,
      "num_input_tokens_seen": 8188368,
      "step": 9450
    },
    {
      "epoch": 4.457802923149458,
      "grad_norm": 0.039820559322834015,
      "learning_rate": 0.26051551566979964,
      "loss": 0.2355,
      "num_input_tokens_seen": 8194000,
      "step": 9455
    },
    {
      "epoch": 4.46016030174446,
      "grad_norm": 0.009643997997045517,
      "learning_rate": 0.26047567901920876,
      "loss": 0.4538,
      "num_input_tokens_seen": 8201072,
      "step": 9460
    },
    {
      "epoch": 4.462517680339462,
      "grad_norm": 0.005420724395662546,
      "learning_rate": 0.2604358253318798,
      "loss": 0.1938,
      "num_input_tokens_seen": 8205680,
      "step": 9465
    },
    {
      "epoch": 4.464875058934465,
      "grad_norm": 0.01084974780678749,
      "learning_rate": 0.26039595461395876,
      "loss": 0.3105,
      "num_input_tokens_seen": 8210464,
      "step": 9470
    },
    {
      "epoch": 4.467232437529467,
      "grad_norm": 0.01789839379489422,
      "learning_rate": 0.26035606687159424,
      "loss": 0.2718,
      "num_input_tokens_seen": 8214752,
      "step": 9475
    },
    {
      "epoch": 4.469589816124469,
      "grad_norm": 0.005781130399554968,
      "learning_rate": 0.26031616211093733,
      "loss": 0.2405,
      "num_input_tokens_seen": 8218848,
      "step": 9480
    },
    {
      "epoch": 4.471947194719472,
      "grad_norm": 0.006741950288414955,
      "learning_rate": 0.26027624033814195,
      "loss": 0.2172,
      "num_input_tokens_seen": 8223200,
      "step": 9485
    },
    {
      "epoch": 4.474304573314474,
      "grad_norm": 0.005049307364970446,
      "learning_rate": 0.2602363015593645,
      "loss": 0.1959,
      "num_input_tokens_seen": 8227280,
      "step": 9490
    },
    {
      "epoch": 4.476661951909477,
      "grad_norm": 0.009826654568314552,
      "learning_rate": 0.26019634578076395,
      "loss": 0.2397,
      "num_input_tokens_seen": 8232304,
      "step": 9495
    },
    {
      "epoch": 4.479019330504479,
      "grad_norm": 0.009060743264853954,
      "learning_rate": 0.26015637300850214,
      "loss": 0.2526,
      "num_input_tokens_seen": 8236592,
      "step": 9500
    },
    {
      "epoch": 4.481376709099481,
      "grad_norm": 0.0037591268774122,
      "learning_rate": 0.26011638324874325,
      "loss": 0.1722,
      "num_input_tokens_seen": 8240800,
      "step": 9505
    },
    {
      "epoch": 4.483734087694484,
      "grad_norm": 0.003495309967547655,
      "learning_rate": 0.2600763765076543,
      "loss": 0.2709,
      "num_input_tokens_seen": 8244816,
      "step": 9510
    },
    {
      "epoch": 4.486091466289486,
      "grad_norm": 0.005895252339541912,
      "learning_rate": 0.2600363527914048,
      "loss": 0.2274,
      "num_input_tokens_seen": 8249136,
      "step": 9515
    },
    {
      "epoch": 4.488448844884489,
      "grad_norm": 0.004823778290301561,
      "learning_rate": 0.25999631210616686,
      "loss": 0.2103,
      "num_input_tokens_seen": 8253536,
      "step": 9520
    },
    {
      "epoch": 4.490806223479491,
      "grad_norm": 0.005318429786711931,
      "learning_rate": 0.25995625445811527,
      "loss": 0.2288,
      "num_input_tokens_seen": 8257504,
      "step": 9525
    },
    {
      "epoch": 4.493163602074493,
      "grad_norm": 0.005329417996108532,
      "learning_rate": 0.2599161798534275,
      "loss": 0.237,
      "num_input_tokens_seen": 8261728,
      "step": 9530
    },
    {
      "epoch": 4.495520980669496,
      "grad_norm": 0.006978516932576895,
      "learning_rate": 0.25987608829828346,
      "loss": 0.2462,
      "num_input_tokens_seen": 8266592,
      "step": 9535
    },
    {
      "epoch": 4.497878359264498,
      "grad_norm": 0.00986330397427082,
      "learning_rate": 0.25983597979886586,
      "loss": 0.2445,
      "num_input_tokens_seen": 8270112,
      "step": 9540
    },
    {
      "epoch": 4.500235737859501,
      "grad_norm": 0.004314525052905083,
      "learning_rate": 0.2597958543613599,
      "loss": 0.2615,
      "num_input_tokens_seen": 8274656,
      "step": 9545
    },
    {
      "epoch": 4.502593116454502,
      "grad_norm": 0.006108290050178766,
      "learning_rate": 0.25975571199195335,
      "loss": 0.2265,
      "num_input_tokens_seen": 8279040,
      "step": 9550
    },
    {
      "epoch": 4.5049504950495045,
      "grad_norm": 0.02270430140197277,
      "learning_rate": 0.25971555269683677,
      "loss": 0.252,
      "num_input_tokens_seen": 8283088,
      "step": 9555
    },
    {
      "epoch": 4.507307873644507,
      "grad_norm": 0.004617371596395969,
      "learning_rate": 0.25967537648220324,
      "loss": 0.2596,
      "num_input_tokens_seen": 8287664,
      "step": 9560
    },
    {
      "epoch": 4.509665252239509,
      "grad_norm": 0.005759835243225098,
      "learning_rate": 0.2596351833542483,
      "loss": 0.2208,
      "num_input_tokens_seen": 8291872,
      "step": 9565
    },
    {
      "epoch": 4.512022630834512,
      "grad_norm": 0.006046622525900602,
      "learning_rate": 0.25959497331917036,
      "loss": 0.2352,
      "num_input_tokens_seen": 8295744,
      "step": 9570
    },
    {
      "epoch": 4.514380009429514,
      "grad_norm": 0.00503824558109045,
      "learning_rate": 0.2595547463831703,
      "loss": 0.2361,
      "num_input_tokens_seen": 8300016,
      "step": 9575
    },
    {
      "epoch": 4.5167373880245165,
      "grad_norm": 0.007392666768282652,
      "learning_rate": 0.25951450255245156,
      "loss": 0.2336,
      "num_input_tokens_seen": 8303840,
      "step": 9580
    },
    {
      "epoch": 4.519094766619519,
      "grad_norm": 0.007078335154801607,
      "learning_rate": 0.2594742418332203,
      "loss": 0.2507,
      "num_input_tokens_seen": 8307472,
      "step": 9585
    },
    {
      "epoch": 4.521452145214521,
      "grad_norm": 0.0046226088888943195,
      "learning_rate": 0.2594339642316852,
      "loss": 0.197,
      "num_input_tokens_seen": 8312128,
      "step": 9590
    },
    {
      "epoch": 4.523809523809524,
      "grad_norm": 0.010770881548523903,
      "learning_rate": 0.2593936697540576,
      "loss": 0.3129,
      "num_input_tokens_seen": 8316672,
      "step": 9595
    },
    {
      "epoch": 4.526166902404526,
      "grad_norm": 0.004187664482742548,
      "learning_rate": 0.2593533584065514,
      "loss": 0.2603,
      "num_input_tokens_seen": 8321456,
      "step": 9600
    },
    {
      "epoch": 4.526166902404526,
      "eval_loss": 0.22926250100135803,
      "eval_runtime": 21.8861,
      "eval_samples_per_second": 43.087,
      "eval_steps_per_second": 21.566,
      "num_input_tokens_seen": 8321456,
      "step": 9600
    },
    {
      "epoch": 4.5285242809995285,
      "grad_norm": 0.004887247458100319,
      "learning_rate": 0.2593130301953831,
      "loss": 0.2147,
      "num_input_tokens_seen": 8326384,
      "step": 9605
    },
    {
      "epoch": 4.530881659594531,
      "grad_norm": 0.0026440327055752277,
      "learning_rate": 0.2592726851267718,
      "loss": 0.2067,
      "num_input_tokens_seen": 8329920,
      "step": 9610
    },
    {
      "epoch": 4.533239038189533,
      "grad_norm": 0.004833553917706013,
      "learning_rate": 0.2592323232069393,
      "loss": 0.2381,
      "num_input_tokens_seen": 8334832,
      "step": 9615
    },
    {
      "epoch": 4.535596416784536,
      "grad_norm": 0.005247662775218487,
      "learning_rate": 0.25919194444210986,
      "loss": 0.2769,
      "num_input_tokens_seen": 8339584,
      "step": 9620
    },
    {
      "epoch": 4.537953795379538,
      "grad_norm": 0.003497653640806675,
      "learning_rate": 0.2591515488385103,
      "loss": 0.2397,
      "num_input_tokens_seen": 8343552,
      "step": 9625
    },
    {
      "epoch": 4.5403111739745405,
      "grad_norm": 0.003319311188533902,
      "learning_rate": 0.2591111364023704,
      "loss": 0.2314,
      "num_input_tokens_seen": 8348160,
      "step": 9630
    },
    {
      "epoch": 4.542668552569543,
      "grad_norm": 0.0033221289049834013,
      "learning_rate": 0.259070707139922,
      "loss": 0.2282,
      "num_input_tokens_seen": 8352080,
      "step": 9635
    },
    {
      "epoch": 4.545025931164545,
      "grad_norm": 0.0030070925131440163,
      "learning_rate": 0.25903026105739985,
      "loss": 0.2247,
      "num_input_tokens_seen": 8356576,
      "step": 9640
    },
    {
      "epoch": 4.547383309759548,
      "grad_norm": 0.003090352052822709,
      "learning_rate": 0.2589897981610413,
      "loss": 0.2339,
      "num_input_tokens_seen": 8361328,
      "step": 9645
    },
    {
      "epoch": 4.54974068835455,
      "grad_norm": 0.0038421889767050743,
      "learning_rate": 0.2589493184570863,
      "loss": 0.2487,
      "num_input_tokens_seen": 8365744,
      "step": 9650
    },
    {
      "epoch": 4.5520980669495525,
      "grad_norm": 0.005688461475074291,
      "learning_rate": 0.25890882195177717,
      "loss": 0.2455,
      "num_input_tokens_seen": 8369984,
      "step": 9655
    },
    {
      "epoch": 4.554455445544555,
      "grad_norm": 0.005062142387032509,
      "learning_rate": 0.25886830865135907,
      "loss": 0.2418,
      "num_input_tokens_seen": 8374432,
      "step": 9660
    },
    {
      "epoch": 4.5568128241395565,
      "grad_norm": 0.00412049749866128,
      "learning_rate": 0.25882777856207967,
      "loss": 0.2452,
      "num_input_tokens_seen": 8378144,
      "step": 9665
    },
    {
      "epoch": 4.559170202734559,
      "grad_norm": 0.003634281689301133,
      "learning_rate": 0.2587872316901892,
      "loss": 0.2149,
      "num_input_tokens_seen": 8383056,
      "step": 9670
    },
    {
      "epoch": 4.561527581329561,
      "grad_norm": 0.006366658490151167,
      "learning_rate": 0.25874666804194046,
      "loss": 0.2425,
      "num_input_tokens_seen": 8387696,
      "step": 9675
    },
    {
      "epoch": 4.563884959924564,
      "grad_norm": 0.003567873500287533,
      "learning_rate": 0.258706087623589,
      "loss": 0.1813,
      "num_input_tokens_seen": 8391712,
      "step": 9680
    },
    {
      "epoch": 4.566242338519566,
      "grad_norm": 0.0028828075155615807,
      "learning_rate": 0.25866549044139264,
      "loss": 0.1598,
      "num_input_tokens_seen": 8395360,
      "step": 9685
    },
    {
      "epoch": 4.5685997171145685,
      "grad_norm": 0.003961341921240091,
      "learning_rate": 0.25862487650161214,
      "loss": 0.26,
      "num_input_tokens_seen": 8399152,
      "step": 9690
    },
    {
      "epoch": 4.570957095709571,
      "grad_norm": 0.003979521803557873,
      "learning_rate": 0.2585842458105106,
      "loss": 0.2226,
      "num_input_tokens_seen": 8403232,
      "step": 9695
    },
    {
      "epoch": 4.573314474304573,
      "grad_norm": 0.0027804041747003794,
      "learning_rate": 0.2585435983743538,
      "loss": 0.1773,
      "num_input_tokens_seen": 8407712,
      "step": 9700
    },
    {
      "epoch": 4.575671852899576,
      "grad_norm": 0.0028531397692859173,
      "learning_rate": 0.2585029341994101,
      "loss": 0.1727,
      "num_input_tokens_seen": 8412192,
      "step": 9705
    },
    {
      "epoch": 4.578029231494578,
      "grad_norm": 0.0029682388994842768,
      "learning_rate": 0.2584622532919504,
      "loss": 0.2049,
      "num_input_tokens_seen": 8416144,
      "step": 9710
    },
    {
      "epoch": 4.5803866100895805,
      "grad_norm": 0.003179004881531,
      "learning_rate": 0.2584215556582482,
      "loss": 0.2312,
      "num_input_tokens_seen": 8420096,
      "step": 9715
    },
    {
      "epoch": 4.582743988684583,
      "grad_norm": 0.002416998613625765,
      "learning_rate": 0.25838084130457967,
      "loss": 0.1769,
      "num_input_tokens_seen": 8423664,
      "step": 9720
    },
    {
      "epoch": 4.585101367279585,
      "grad_norm": 0.003860981436446309,
      "learning_rate": 0.2583401102372234,
      "loss": 0.2452,
      "num_input_tokens_seen": 8428112,
      "step": 9725
    },
    {
      "epoch": 4.587458745874588,
      "grad_norm": 0.005476855207234621,
      "learning_rate": 0.2582993624624606,
      "loss": 0.2191,
      "num_input_tokens_seen": 8432224,
      "step": 9730
    },
    {
      "epoch": 4.58981612446959,
      "grad_norm": 0.002516490640118718,
      "learning_rate": 0.25825859798657513,
      "loss": 0.2393,
      "num_input_tokens_seen": 8436816,
      "step": 9735
    },
    {
      "epoch": 4.5921735030645925,
      "grad_norm": 0.0033048244658857584,
      "learning_rate": 0.25821781681585343,
      "loss": 0.2307,
      "num_input_tokens_seen": 8441200,
      "step": 9740
    },
    {
      "epoch": 4.594530881659595,
      "grad_norm": 0.0030626184307038784,
      "learning_rate": 0.2581770189565844,
      "loss": 0.2231,
      "num_input_tokens_seen": 8446096,
      "step": 9745
    },
    {
      "epoch": 4.596888260254596,
      "grad_norm": 0.003267318708822131,
      "learning_rate": 0.25813620441505963,
      "loss": 0.1796,
      "num_input_tokens_seen": 8450800,
      "step": 9750
    },
    {
      "epoch": 4.599245638849599,
      "grad_norm": 0.0036412018816918135,
      "learning_rate": 0.2580953731975732,
      "loss": 0.236,
      "num_input_tokens_seen": 8455296,
      "step": 9755
    },
    {
      "epoch": 4.601603017444601,
      "grad_norm": 0.00294292438775301,
      "learning_rate": 0.2580545253104218,
      "loss": 0.2127,
      "num_input_tokens_seen": 8459472,
      "step": 9760
    },
    {
      "epoch": 4.603960396039604,
      "grad_norm": 0.0035328813828527927,
      "learning_rate": 0.2580136607599047,
      "loss": 0.248,
      "num_input_tokens_seen": 8463280,
      "step": 9765
    },
    {
      "epoch": 4.606317774634606,
      "grad_norm": 0.0023099060636013746,
      "learning_rate": 0.2579727795523238,
      "loss": 0.2591,
      "num_input_tokens_seen": 8467360,
      "step": 9770
    },
    {
      "epoch": 4.608675153229608,
      "grad_norm": 0.004458182957023382,
      "learning_rate": 0.25793188169398334,
      "loss": 0.2515,
      "num_input_tokens_seen": 8471152,
      "step": 9775
    },
    {
      "epoch": 4.611032531824611,
      "grad_norm": 0.0029415234457701445,
      "learning_rate": 0.25789096719119037,
      "loss": 0.2369,
      "num_input_tokens_seen": 8474912,
      "step": 9780
    },
    {
      "epoch": 4.613389910419613,
      "grad_norm": 0.004502985626459122,
      "learning_rate": 0.2578500360502544,
      "loss": 0.2517,
      "num_input_tokens_seen": 8478464,
      "step": 9785
    },
    {
      "epoch": 4.615747289014616,
      "grad_norm": 0.0023702296894043684,
      "learning_rate": 0.2578090882774876,
      "loss": 0.193,
      "num_input_tokens_seen": 8482656,
      "step": 9790
    },
    {
      "epoch": 4.618104667609618,
      "grad_norm": 0.003934910520911217,
      "learning_rate": 0.25776812387920456,
      "loss": 0.2942,
      "num_input_tokens_seen": 8486848,
      "step": 9795
    },
    {
      "epoch": 4.62046204620462,
      "grad_norm": 0.0023481135722249746,
      "learning_rate": 0.2577271428617225,
      "loss": 0.2278,
      "num_input_tokens_seen": 8490096,
      "step": 9800
    },
    {
      "epoch": 4.62046204620462,
      "eval_loss": 0.22128428518772125,
      "eval_runtime": 21.9274,
      "eval_samples_per_second": 43.006,
      "eval_steps_per_second": 21.526,
      "num_input_tokens_seen": 8490096,
      "step": 9800
    },
    {
      "epoch": 4.622819424799623,
      "grad_norm": 0.003357951296493411,
      "learning_rate": 0.25768614523136124,
      "loss": 0.2353,
      "num_input_tokens_seen": 8494800,
      "step": 9805
    },
    {
      "epoch": 4.625176803394625,
      "grad_norm": 0.0028887705411762,
      "learning_rate": 0.25764513099444314,
      "loss": 0.2115,
      "num_input_tokens_seen": 8498928,
      "step": 9810
    },
    {
      "epoch": 4.627534181989628,
      "grad_norm": 0.0038435584865510464,
      "learning_rate": 0.25760410015729307,
      "loss": 0.2105,
      "num_input_tokens_seen": 8503264,
      "step": 9815
    },
    {
      "epoch": 4.62989156058463,
      "grad_norm": 0.003056455636397004,
      "learning_rate": 0.2575630527262385,
      "loss": 0.1997,
      "num_input_tokens_seen": 8506624,
      "step": 9820
    },
    {
      "epoch": 4.632248939179632,
      "grad_norm": 0.006272217258810997,
      "learning_rate": 0.25752198870760945,
      "loss": 0.2357,
      "num_input_tokens_seen": 8511440,
      "step": 9825
    },
    {
      "epoch": 4.634606317774635,
      "grad_norm": 0.0039057552348822355,
      "learning_rate": 0.2574809081077386,
      "loss": 0.2347,
      "num_input_tokens_seen": 8515584,
      "step": 9830
    },
    {
      "epoch": 4.636963696369637,
      "grad_norm": 0.003493529511615634,
      "learning_rate": 0.257439810932961,
      "loss": 0.2357,
      "num_input_tokens_seen": 8519280,
      "step": 9835
    },
    {
      "epoch": 4.63932107496464,
      "grad_norm": 0.008781437762081623,
      "learning_rate": 0.2573986971896144,
      "loss": 0.2397,
      "num_input_tokens_seen": 8522816,
      "step": 9840
    },
    {
      "epoch": 4.641678453559642,
      "grad_norm": 0.0033617110457271338,
      "learning_rate": 0.257357566884039,
      "loss": 0.2296,
      "num_input_tokens_seen": 8526720,
      "step": 9845
    },
    {
      "epoch": 4.644035832154644,
      "grad_norm": 0.0022055399604141712,
      "learning_rate": 0.25731642002257765,
      "loss": 0.2069,
      "num_input_tokens_seen": 8531808,
      "step": 9850
    },
    {
      "epoch": 4.646393210749647,
      "grad_norm": 0.00842844694852829,
      "learning_rate": 0.25727525661157574,
      "loss": 0.2492,
      "num_input_tokens_seen": 8536000,
      "step": 9855
    },
    {
      "epoch": 4.648750589344649,
      "grad_norm": 0.0037738282699137926,
      "learning_rate": 0.2572340766573811,
      "loss": 0.2444,
      "num_input_tokens_seen": 8539776,
      "step": 9860
    },
    {
      "epoch": 4.651107967939651,
      "grad_norm": 0.0038234624080359936,
      "learning_rate": 0.25719288016634434,
      "loss": 0.1719,
      "num_input_tokens_seen": 8544384,
      "step": 9865
    },
    {
      "epoch": 4.653465346534653,
      "grad_norm": 0.003316260641440749,
      "learning_rate": 0.25715166714481835,
      "loss": 0.1947,
      "num_input_tokens_seen": 8548464,
      "step": 9870
    },
    {
      "epoch": 4.655822725129656,
      "grad_norm": 0.006922050379216671,
      "learning_rate": 0.2571104375991587,
      "loss": 0.2491,
      "num_input_tokens_seen": 8552496,
      "step": 9875
    },
    {
      "epoch": 4.658180103724658,
      "grad_norm": 0.002857015235349536,
      "learning_rate": 0.2570691915357236,
      "loss": 0.2446,
      "num_input_tokens_seen": 8557104,
      "step": 9880
    },
    {
      "epoch": 4.66053748231966,
      "grad_norm": 0.0035318995360285044,
      "learning_rate": 0.2570279289608736,
      "loss": 0.2364,
      "num_input_tokens_seen": 8561568,
      "step": 9885
    },
    {
      "epoch": 4.662894860914663,
      "grad_norm": 0.0036992570385336876,
      "learning_rate": 0.256986649880972,
      "loss": 0.2249,
      "num_input_tokens_seen": 8565760,
      "step": 9890
    },
    {
      "epoch": 4.665252239509665,
      "grad_norm": 0.003426705254241824,
      "learning_rate": 0.25694535430238447,
      "loss": 0.2438,
      "num_input_tokens_seen": 8571344,
      "step": 9895
    },
    {
      "epoch": 4.667609618104668,
      "grad_norm": 0.0063272821716964245,
      "learning_rate": 0.25690404223147933,
      "loss": 0.2213,
      "num_input_tokens_seen": 8576432,
      "step": 9900
    },
    {
      "epoch": 4.66996699669967,
      "grad_norm": 0.0038960715755820274,
      "learning_rate": 0.2568627136746275,
      "loss": 0.2623,
      "num_input_tokens_seen": 8580640,
      "step": 9905
    },
    {
      "epoch": 4.672324375294672,
      "grad_norm": 0.004203151445835829,
      "learning_rate": 0.25682136863820226,
      "loss": 0.2446,
      "num_input_tokens_seen": 8585392,
      "step": 9910
    },
    {
      "epoch": 4.674681753889675,
      "grad_norm": 0.004907167516648769,
      "learning_rate": 0.25678000712857957,
      "loss": 0.2304,
      "num_input_tokens_seen": 8589488,
      "step": 9915
    },
    {
      "epoch": 4.677039132484677,
      "grad_norm": 0.004104039166122675,
      "learning_rate": 0.2567386291521379,
      "loss": 0.2174,
      "num_input_tokens_seen": 8593936,
      "step": 9920
    },
    {
      "epoch": 4.67939651107968,
      "grad_norm": 0.002158834831789136,
      "learning_rate": 0.2566972347152583,
      "loss": 0.2293,
      "num_input_tokens_seen": 8598976,
      "step": 9925
    },
    {
      "epoch": 4.681753889674682,
      "grad_norm": 0.003145094495266676,
      "learning_rate": 0.2566558238243242,
      "loss": 0.2081,
      "num_input_tokens_seen": 8603312,
      "step": 9930
    },
    {
      "epoch": 4.684111268269684,
      "grad_norm": 0.0036779381334781647,
      "learning_rate": 0.25661439648572176,
      "loss": 0.2235,
      "num_input_tokens_seen": 8607680,
      "step": 9935
    },
    {
      "epoch": 4.686468646864687,
      "grad_norm": 0.0023090678732842207,
      "learning_rate": 0.25657295270583963,
      "loss": 0.2296,
      "num_input_tokens_seen": 8611216,
      "step": 9940
    },
    {
      "epoch": 4.688826025459689,
      "grad_norm": 0.0047711776569485664,
      "learning_rate": 0.25653149249106894,
      "loss": 0.2636,
      "num_input_tokens_seen": 8615744,
      "step": 9945
    },
    {
      "epoch": 4.691183404054691,
      "grad_norm": 0.0027853306382894516,
      "learning_rate": 0.25649001584780323,
      "loss": 0.1936,
      "num_input_tokens_seen": 8620416,
      "step": 9950
    },
    {
      "epoch": 4.693540782649693,
      "grad_norm": 0.00333112059161067,
      "learning_rate": 0.2564485227824389,
      "loss": 0.2227,
      "num_input_tokens_seen": 8624480,
      "step": 9955
    },
    {
      "epoch": 4.6958981612446955,
      "grad_norm": 0.003930002450942993,
      "learning_rate": 0.25640701330137466,
      "loss": 0.1872,
      "num_input_tokens_seen": 8629120,
      "step": 9960
    },
    {
      "epoch": 4.698255539839698,
      "grad_norm": 0.0029448510613292456,
      "learning_rate": 0.2563654874110117,
      "loss": 0.1782,
      "num_input_tokens_seen": 8633744,
      "step": 9965
    },
    {
      "epoch": 4.7006129184347,
      "grad_norm": 0.004129595123231411,
      "learning_rate": 0.256323945117754,
      "loss": 0.3511,
      "num_input_tokens_seen": 8638912,
      "step": 9970
    },
    {
      "epoch": 4.702970297029703,
      "grad_norm": 0.00583377992734313,
      "learning_rate": 0.2562823864280078,
      "loss": 0.2579,
      "num_input_tokens_seen": 8643360,
      "step": 9975
    },
    {
      "epoch": 4.705327675624705,
      "grad_norm": 0.0033459847327321768,
      "learning_rate": 0.25624081134818194,
      "loss": 0.2358,
      "num_input_tokens_seen": 8647680,
      "step": 9980
    },
    {
      "epoch": 4.7076850542197075,
      "grad_norm": 0.010874238796532154,
      "learning_rate": 0.2561992198846879,
      "loss": 0.2614,
      "num_input_tokens_seen": 8652288,
      "step": 9985
    },
    {
      "epoch": 4.71004243281471,
      "grad_norm": 0.006391587201505899,
      "learning_rate": 0.25615761204393955,
      "loss": 0.2564,
      "num_input_tokens_seen": 8656768,
      "step": 9990
    },
    {
      "epoch": 4.712399811409712,
      "grad_norm": 0.014513351023197174,
      "learning_rate": 0.2561159878323534,
      "loss": 0.2325,
      "num_input_tokens_seen": 8661920,
      "step": 9995
    },
    {
      "epoch": 4.714757190004715,
      "grad_norm": 0.009979763068258762,
      "learning_rate": 0.2560743472563483,
      "loss": 0.2126,
      "num_input_tokens_seen": 8665904,
      "step": 10000
    },
    {
      "epoch": 4.714757190004715,
      "eval_loss": 0.22703884541988373,
      "eval_runtime": 21.9521,
      "eval_samples_per_second": 42.957,
      "eval_steps_per_second": 21.501,
      "num_input_tokens_seen": 8665904,
      "step": 10000
    },
    {
      "epoch": 4.717114568599717,
      "grad_norm": 0.0031102376524358988,
      "learning_rate": 0.25603269032234593,
      "loss": 0.206,
      "num_input_tokens_seen": 8669696,
      "step": 10005
    },
    {
      "epoch": 4.7194719471947195,
      "grad_norm": 0.002624074462801218,
      "learning_rate": 0.2559910170367702,
      "loss": 0.2055,
      "num_input_tokens_seen": 8673440,
      "step": 10010
    },
    {
      "epoch": 4.721829325789722,
      "grad_norm": 0.0036257850006222725,
      "learning_rate": 0.2559493274060477,
      "loss": 0.1968,
      "num_input_tokens_seen": 8677504,
      "step": 10015
    },
    {
      "epoch": 4.724186704384724,
      "grad_norm": 0.010346390306949615,
      "learning_rate": 0.2559076214366074,
      "loss": 0.2598,
      "num_input_tokens_seen": 8682544,
      "step": 10020
    },
    {
      "epoch": 4.726544082979727,
      "grad_norm": 0.004050659481436014,
      "learning_rate": 0.25586589913488106,
      "loss": 0.2369,
      "num_input_tokens_seen": 8686576,
      "step": 10025
    },
    {
      "epoch": 4.728901461574729,
      "grad_norm": 0.007511364761739969,
      "learning_rate": 0.2558241605073026,
      "loss": 0.2446,
      "num_input_tokens_seen": 8690032,
      "step": 10030
    },
    {
      "epoch": 4.7312588401697315,
      "grad_norm": 0.006256521679461002,
      "learning_rate": 0.25578240556030873,
      "loss": 0.2592,
      "num_input_tokens_seen": 8694320,
      "step": 10035
    },
    {
      "epoch": 4.733616218764734,
      "grad_norm": 0.006115077529102564,
      "learning_rate": 0.2557406343003386,
      "loss": 0.2225,
      "num_input_tokens_seen": 8698736,
      "step": 10040
    },
    {
      "epoch": 4.735973597359736,
      "grad_norm": 0.006495185662060976,
      "learning_rate": 0.25569884673383375,
      "loss": 0.235,
      "num_input_tokens_seen": 8702816,
      "step": 10045
    },
    {
      "epoch": 4.738330975954739,
      "grad_norm": 0.006397186312824488,
      "learning_rate": 0.25565704286723856,
      "loss": 0.2272,
      "num_input_tokens_seen": 8707568,
      "step": 10050
    },
    {
      "epoch": 4.740688354549741,
      "grad_norm": 0.006227480247616768,
      "learning_rate": 0.25561522270699955,
      "loss": 0.2294,
      "num_input_tokens_seen": 8711792,
      "step": 10055
    },
    {
      "epoch": 4.7430457331447435,
      "grad_norm": 0.012340952642261982,
      "learning_rate": 0.25557338625956594,
      "loss": 0.2274,
      "num_input_tokens_seen": 8715872,
      "step": 10060
    },
    {
      "epoch": 4.745403111739745,
      "grad_norm": 0.005235287360846996,
      "learning_rate": 0.25553153353138947,
      "loss": 0.2481,
      "num_input_tokens_seen": 8719776,
      "step": 10065
    },
    {
      "epoch": 4.7477604903347475,
      "grad_norm": 0.005572305992245674,
      "learning_rate": 0.2554896645289243,
      "loss": 0.2194,
      "num_input_tokens_seen": 8723888,
      "step": 10070
    },
    {
      "epoch": 4.75011786892975,
      "grad_norm": 0.0073296381160616875,
      "learning_rate": 0.2554477792586272,
      "loss": 0.1983,
      "num_input_tokens_seen": 8728048,
      "step": 10075
    },
    {
      "epoch": 4.752475247524752,
      "grad_norm": 0.002989573869854212,
      "learning_rate": 0.25540587772695744,
      "loss": 0.1702,
      "num_input_tokens_seen": 8733280,
      "step": 10080
    },
    {
      "epoch": 4.754832626119755,
      "grad_norm": 0.005363411735743284,
      "learning_rate": 0.2553639599403767,
      "loss": 0.3059,
      "num_input_tokens_seen": 8737872,
      "step": 10085
    },
    {
      "epoch": 4.757190004714757,
      "grad_norm": 0.004296573810279369,
      "learning_rate": 0.2553220259053493,
      "loss": 0.2908,
      "num_input_tokens_seen": 8742208,
      "step": 10090
    },
    {
      "epoch": 4.7595473833097595,
      "grad_norm": 0.004309081472456455,
      "learning_rate": 0.2552800756283419,
      "loss": 0.2834,
      "num_input_tokens_seen": 8746144,
      "step": 10095
    },
    {
      "epoch": 4.761904761904762,
      "grad_norm": 0.0028175327461212873,
      "learning_rate": 0.25523810911582373,
      "loss": 0.2627,
      "num_input_tokens_seen": 8750000,
      "step": 10100
    },
    {
      "epoch": 4.764262140499764,
      "grad_norm": 0.002214773092418909,
      "learning_rate": 0.25519612637426675,
      "loss": 0.234,
      "num_input_tokens_seen": 8754080,
      "step": 10105
    },
    {
      "epoch": 4.766619519094767,
      "grad_norm": 0.00444077979773283,
      "learning_rate": 0.25515412741014504,
      "loss": 0.2597,
      "num_input_tokens_seen": 8758176,
      "step": 10110
    },
    {
      "epoch": 4.768976897689769,
      "grad_norm": 0.0037436443381011486,
      "learning_rate": 0.2551121122299355,
      "loss": 0.2135,
      "num_input_tokens_seen": 8762496,
      "step": 10115
    },
    {
      "epoch": 4.7713342762847715,
      "grad_norm": 0.0017685518832877278,
      "learning_rate": 0.2550700808401173,
      "loss": 0.2585,
      "num_input_tokens_seen": 8766992,
      "step": 10120
    },
    {
      "epoch": 4.773691654879774,
      "grad_norm": 0.001823101774789393,
      "learning_rate": 0.2550280332471722,
      "loss": 0.232,
      "num_input_tokens_seen": 8771360,
      "step": 10125
    },
    {
      "epoch": 4.776049033474776,
      "grad_norm": 0.0034805668983608484,
      "learning_rate": 0.2549859694575845,
      "loss": 0.231,
      "num_input_tokens_seen": 8775952,
      "step": 10130
    },
    {
      "epoch": 4.778406412069779,
      "grad_norm": 0.0030522567685693502,
      "learning_rate": 0.254943889477841,
      "loss": 0.2085,
      "num_input_tokens_seen": 8780464,
      "step": 10135
    },
    {
      "epoch": 4.780763790664781,
      "grad_norm": 0.0026623026933521032,
      "learning_rate": 0.25490179331443097,
      "loss": 0.2525,
      "num_input_tokens_seen": 8785264,
      "step": 10140
    },
    {
      "epoch": 4.7831211692597835,
      "grad_norm": 0.00334662152454257,
      "learning_rate": 0.25485968097384615,
      "loss": 0.2419,
      "num_input_tokens_seen": 8789904,
      "step": 10145
    },
    {
      "epoch": 4.785478547854785,
      "grad_norm": 0.003158440813422203,
      "learning_rate": 0.25481755246258075,
      "loss": 0.2221,
      "num_input_tokens_seen": 8794832,
      "step": 10150
    },
    {
      "epoch": 4.787835926449787,
      "grad_norm": 0.002311042742803693,
      "learning_rate": 0.2547754077871315,
      "loss": 0.2286,
      "num_input_tokens_seen": 8799680,
      "step": 10155
    },
    {
      "epoch": 4.79019330504479,
      "grad_norm": 0.0026144045405089855,
      "learning_rate": 0.25473324695399774,
      "loss": 0.2185,
      "num_input_tokens_seen": 8804256,
      "step": 10160
    },
    {
      "epoch": 4.792550683639792,
      "grad_norm": 0.0015132286353036761,
      "learning_rate": 0.25469106996968105,
      "loss": 0.2467,
      "num_input_tokens_seen": 8808496,
      "step": 10165
    },
    {
      "epoch": 4.794908062234795,
      "grad_norm": 0.005861798767000437,
      "learning_rate": 0.2546488768406858,
      "loss": 0.2271,
      "num_input_tokens_seen": 8812880,
      "step": 10170
    },
    {
      "epoch": 4.797265440829797,
      "grad_norm": 0.003185089910402894,
      "learning_rate": 0.25460666757351863,
      "loss": 0.2423,
      "num_input_tokens_seen": 8816720,
      "step": 10175
    },
    {
      "epoch": 4.799622819424799,
      "grad_norm": 0.0015774901257827878,
      "learning_rate": 0.25456444217468877,
      "loss": 0.2387,
      "num_input_tokens_seen": 8821040,
      "step": 10180
    },
    {
      "epoch": 4.801980198019802,
      "grad_norm": 0.002794100670143962,
      "learning_rate": 0.25452220065070785,
      "loss": 0.2122,
      "num_input_tokens_seen": 8825536,
      "step": 10185
    },
    {
      "epoch": 4.804337576614804,
      "grad_norm": 0.0017370900604873896,
      "learning_rate": 0.2544799430080901,
      "loss": 0.2491,
      "num_input_tokens_seen": 8829584,
      "step": 10190
    },
    {
      "epoch": 4.806694955209807,
      "grad_norm": 0.00214237323962152,
      "learning_rate": 0.2544376692533522,
      "loss": 0.3126,
      "num_input_tokens_seen": 8833904,
      "step": 10195
    },
    {
      "epoch": 4.809052333804809,
      "grad_norm": 0.001375768450088799,
      "learning_rate": 0.2543953793930132,
      "loss": 0.2166,
      "num_input_tokens_seen": 8837712,
      "step": 10200
    },
    {
      "epoch": 4.809052333804809,
      "eval_loss": 0.22068026661872864,
      "eval_runtime": 21.8971,
      "eval_samples_per_second": 43.065,
      "eval_steps_per_second": 21.555,
      "num_input_tokens_seen": 8837712,
      "step": 10200
    },
    {
      "epoch": 4.811409712399811,
      "grad_norm": 0.0020371403079479933,
      "learning_rate": 0.2543530734335948,
      "loss": 0.2026,
      "num_input_tokens_seen": 8842464,
      "step": 10205
    },
    {
      "epoch": 4.813767090994814,
      "grad_norm": 0.004257515538483858,
      "learning_rate": 0.2543107513816211,
      "loss": 0.2787,
      "num_input_tokens_seen": 8846768,
      "step": 10210
    },
    {
      "epoch": 4.816124469589816,
      "grad_norm": 0.0015303095569834113,
      "learning_rate": 0.25426841324361865,
      "loss": 0.2431,
      "num_input_tokens_seen": 8850944,
      "step": 10215
    },
    {
      "epoch": 4.818481848184819,
      "grad_norm": 0.002723381156101823,
      "learning_rate": 0.2542260590261166,
      "loss": 0.2516,
      "num_input_tokens_seen": 8855264,
      "step": 10220
    },
    {
      "epoch": 4.820839226779821,
      "grad_norm": 0.001737845130264759,
      "learning_rate": 0.2541836887356465,
      "loss": 0.242,
      "num_input_tokens_seen": 8859424,
      "step": 10225
    },
    {
      "epoch": 4.823196605374823,
      "grad_norm": 0.0038533592596650124,
      "learning_rate": 0.2541413023787423,
      "loss": 0.2771,
      "num_input_tokens_seen": 8864272,
      "step": 10230
    },
    {
      "epoch": 4.825553983969826,
      "grad_norm": 0.0027567155193537474,
      "learning_rate": 0.2540988999619405,
      "loss": 0.2575,
      "num_input_tokens_seen": 8868304,
      "step": 10235
    },
    {
      "epoch": 4.827911362564828,
      "grad_norm": 0.004842326510697603,
      "learning_rate": 0.25405648149178023,
      "loss": 0.2454,
      "num_input_tokens_seen": 8872832,
      "step": 10240
    },
    {
      "epoch": 4.830268741159831,
      "grad_norm": 0.001968837110325694,
      "learning_rate": 0.2540140469748028,
      "loss": 0.2221,
      "num_input_tokens_seen": 8877472,
      "step": 10245
    },
    {
      "epoch": 4.832626119754833,
      "grad_norm": 0.0018205288797616959,
      "learning_rate": 0.25397159641755224,
      "loss": 0.2036,
      "num_input_tokens_seen": 8882304,
      "step": 10250
    },
    {
      "epoch": 4.834983498349835,
      "grad_norm": 0.0015791808255016804,
      "learning_rate": 0.2539291298265749,
      "loss": 0.2284,
      "num_input_tokens_seen": 8886272,
      "step": 10255
    },
    {
      "epoch": 4.837340876944838,
      "grad_norm": 0.004217464942485094,
      "learning_rate": 0.2538866472084197,
      "loss": 0.2075,
      "num_input_tokens_seen": 8890368,
      "step": 10260
    },
    {
      "epoch": 4.839698255539839,
      "grad_norm": 0.001711296266876161,
      "learning_rate": 0.25384414856963794,
      "loss": 0.2283,
      "num_input_tokens_seen": 8894480,
      "step": 10265
    },
    {
      "epoch": 4.842055634134842,
      "grad_norm": 0.0027310035657137632,
      "learning_rate": 0.25380163391678356,
      "loss": 0.2166,
      "num_input_tokens_seen": 8898560,
      "step": 10270
    },
    {
      "epoch": 4.844413012729844,
      "grad_norm": 0.0016945303650572896,
      "learning_rate": 0.2537591032564127,
      "loss": 0.235,
      "num_input_tokens_seen": 8902608,
      "step": 10275
    },
    {
      "epoch": 4.8467703913248465,
      "grad_norm": 0.003419552929699421,
      "learning_rate": 0.25371655659508424,
      "loss": 0.2296,
      "num_input_tokens_seen": 8907040,
      "step": 10280
    },
    {
      "epoch": 4.849127769919849,
      "grad_norm": 0.0016422135522589087,
      "learning_rate": 0.25367399393935935,
      "loss": 0.2255,
      "num_input_tokens_seen": 8910704,
      "step": 10285
    },
    {
      "epoch": 4.851485148514851,
      "grad_norm": 0.002946835942566395,
      "learning_rate": 0.25363141529580174,
      "loss": 0.2231,
      "num_input_tokens_seen": 8915072,
      "step": 10290
    },
    {
      "epoch": 4.853842527109854,
      "grad_norm": 0.00183379917871207,
      "learning_rate": 0.2535888206709776,
      "loss": 0.2194,
      "num_input_tokens_seen": 8918976,
      "step": 10295
    },
    {
      "epoch": 4.856199905704856,
      "grad_norm": 0.001559894299134612,
      "learning_rate": 0.2535462100714555,
      "loss": 0.1925,
      "num_input_tokens_seen": 8923568,
      "step": 10300
    },
    {
      "epoch": 4.858557284299859,
      "grad_norm": 0.0023183412849903107,
      "learning_rate": 0.2535035835038066,
      "loss": 0.2546,
      "num_input_tokens_seen": 8927968,
      "step": 10305
    },
    {
      "epoch": 4.860914662894861,
      "grad_norm": 0.0023057523649185896,
      "learning_rate": 0.2534609409746044,
      "loss": 0.2439,
      "num_input_tokens_seen": 8932480,
      "step": 10310
    },
    {
      "epoch": 4.863272041489863,
      "grad_norm": 0.0016307109035551548,
      "learning_rate": 0.253418282490425,
      "loss": 0.216,
      "num_input_tokens_seen": 8936624,
      "step": 10315
    },
    {
      "epoch": 4.865629420084866,
      "grad_norm": 0.0017056745709851384,
      "learning_rate": 0.2533756080578467,
      "loss": 0.2387,
      "num_input_tokens_seen": 8941264,
      "step": 10320
    },
    {
      "epoch": 4.867986798679868,
      "grad_norm": 0.0024380641989409924,
      "learning_rate": 0.25333291768345056,
      "loss": 0.2222,
      "num_input_tokens_seen": 8945600,
      "step": 10325
    },
    {
      "epoch": 4.870344177274871,
      "grad_norm": 0.003554953495040536,
      "learning_rate": 0.25329021137381996,
      "loss": 0.2312,
      "num_input_tokens_seen": 8949616,
      "step": 10330
    },
    {
      "epoch": 4.872701555869873,
      "grad_norm": 0.0019154661567881703,
      "learning_rate": 0.25324748913554074,
      "loss": 0.2353,
      "num_input_tokens_seen": 8954672,
      "step": 10335
    },
    {
      "epoch": 4.875058934464875,
      "grad_norm": 0.0030683274380862713,
      "learning_rate": 0.2532047509752013,
      "loss": 0.2227,
      "num_input_tokens_seen": 8958576,
      "step": 10340
    },
    {
      "epoch": 4.877416313059878,
      "grad_norm": 0.0014877381036058068,
      "learning_rate": 0.25316199689939217,
      "loss": 0.2154,
      "num_input_tokens_seen": 8962880,
      "step": 10345
    },
    {
      "epoch": 4.879773691654879,
      "grad_norm": 0.0021655242890119553,
      "learning_rate": 0.2531192269147068,
      "loss": 0.2098,
      "num_input_tokens_seen": 8967952,
      "step": 10350
    },
    {
      "epoch": 4.882131070249882,
      "grad_norm": 0.0028664893470704556,
      "learning_rate": 0.2530764410277407,
      "loss": 0.2058,
      "num_input_tokens_seen": 8972160,
      "step": 10355
    },
    {
      "epoch": 4.884488448844884,
      "grad_norm": 0.0036216829903423786,
      "learning_rate": 0.25303363924509203,
      "loss": 0.2506,
      "num_input_tokens_seen": 8975888,
      "step": 10360
    },
    {
      "epoch": 4.8868458274398865,
      "grad_norm": 0.0016360852168872952,
      "learning_rate": 0.25299082157336145,
      "loss": 0.2178,
      "num_input_tokens_seen": 8980752,
      "step": 10365
    },
    {
      "epoch": 4.889203206034889,
      "grad_norm": 0.0016494934679940343,
      "learning_rate": 0.2529479880191519,
      "loss": 0.2383,
      "num_input_tokens_seen": 8984704,
      "step": 10370
    },
    {
      "epoch": 4.891560584629891,
      "grad_norm": 0.0020402274094522,
      "learning_rate": 0.2529051385890689,
      "loss": 0.2389,
      "num_input_tokens_seen": 8988432,
      "step": 10375
    },
    {
      "epoch": 4.893917963224894,
      "grad_norm": 0.0018654990708455443,
      "learning_rate": 0.2528622732897203,
      "loss": 0.2138,
      "num_input_tokens_seen": 8992704,
      "step": 10380
    },
    {
      "epoch": 4.896275341819896,
      "grad_norm": 0.0018442028667777777,
      "learning_rate": 0.25281939212771654,
      "loss": 0.2047,
      "num_input_tokens_seen": 8996560,
      "step": 10385
    },
    {
      "epoch": 4.8986327204148985,
      "grad_norm": 0.0025038737803697586,
      "learning_rate": 0.2527764951096704,
      "loss": 0.2352,
      "num_input_tokens_seen": 9001280,
      "step": 10390
    },
    {
      "epoch": 4.900990099009901,
      "grad_norm": 0.0023145771119743586,
      "learning_rate": 0.2527335822421971,
      "loss": 0.255,
      "num_input_tokens_seen": 9005616,
      "step": 10395
    },
    {
      "epoch": 4.903347477604903,
      "grad_norm": 0.0055069103837013245,
      "learning_rate": 0.25269065353191444,
      "loss": 0.2579,
      "num_input_tokens_seen": 9010400,
      "step": 10400
    },
    {
      "epoch": 4.903347477604903,
      "eval_loss": 0.24866719543933868,
      "eval_runtime": 21.911,
      "eval_samples_per_second": 43.038,
      "eval_steps_per_second": 21.542,
      "num_input_tokens_seen": 9010400,
      "step": 10400
    },
    {
      "epoch": 4.905704856199906,
      "grad_norm": 0.004870031028985977,
      "learning_rate": 0.2526477089854425,
      "loss": 0.2441,
      "num_input_tokens_seen": 9014832,
      "step": 10405
    },
    {
      "epoch": 4.908062234794908,
      "grad_norm": 0.002974500646814704,
      "learning_rate": 0.25260474860940385,
      "loss": 0.2375,
      "num_input_tokens_seen": 9019648,
      "step": 10410
    },
    {
      "epoch": 4.9104196133899105,
      "grad_norm": 0.001540199969895184,
      "learning_rate": 0.2525617724104236,
      "loss": 0.2121,
      "num_input_tokens_seen": 9024000,
      "step": 10415
    },
    {
      "epoch": 4.912776991984913,
      "grad_norm": 0.0016707731410861015,
      "learning_rate": 0.25251878039512915,
      "loss": 0.2241,
      "num_input_tokens_seen": 9028352,
      "step": 10420
    },
    {
      "epoch": 4.915134370579915,
      "grad_norm": 0.0013953062007203698,
      "learning_rate": 0.25247577257015047,
      "loss": 0.1846,
      "num_input_tokens_seen": 9032080,
      "step": 10425
    },
    {
      "epoch": 4.917491749174918,
      "grad_norm": 0.0015885273460298777,
      "learning_rate": 0.2524327489421198,
      "loss": 0.2745,
      "num_input_tokens_seen": 9036864,
      "step": 10430
    },
    {
      "epoch": 4.91984912776992,
      "grad_norm": 0.0033220627810806036,
      "learning_rate": 0.25238970951767203,
      "loss": 0.233,
      "num_input_tokens_seen": 9040800,
      "step": 10435
    },
    {
      "epoch": 4.9222065063649225,
      "grad_norm": 0.002091937931254506,
      "learning_rate": 0.25234665430344433,
      "loss": 0.2386,
      "num_input_tokens_seen": 9044864,
      "step": 10440
    },
    {
      "epoch": 4.924563884959925,
      "grad_norm": 0.004154612310230732,
      "learning_rate": 0.2523035833060764,
      "loss": 0.229,
      "num_input_tokens_seen": 9049024,
      "step": 10445
    },
    {
      "epoch": 4.926921263554927,
      "grad_norm": 0.003014535177499056,
      "learning_rate": 0.2522604965322103,
      "loss": 0.2057,
      "num_input_tokens_seen": 9053248,
      "step": 10450
    },
    {
      "epoch": 4.92927864214993,
      "grad_norm": 0.002263062633574009,
      "learning_rate": 0.25221739398849047,
      "loss": 0.2329,
      "num_input_tokens_seen": 9057376,
      "step": 10455
    },
    {
      "epoch": 4.931636020744932,
      "grad_norm": 0.0018450884381309152,
      "learning_rate": 0.252174275681564,
      "loss": 0.2425,
      "num_input_tokens_seen": 9061024,
      "step": 10460
    },
    {
      "epoch": 4.933993399339934,
      "grad_norm": 0.0014395025791600347,
      "learning_rate": 0.2521311416180802,
      "loss": 0.2082,
      "num_input_tokens_seen": 9065664,
      "step": 10465
    },
    {
      "epoch": 4.936350777934936,
      "grad_norm": 0.003026581835001707,
      "learning_rate": 0.25208799180469094,
      "loss": 0.2503,
      "num_input_tokens_seen": 9069968,
      "step": 10470
    },
    {
      "epoch": 4.938708156529938,
      "grad_norm": 0.0015050627989694476,
      "learning_rate": 0.2520448262480504,
      "loss": 0.2301,
      "num_input_tokens_seen": 9074432,
      "step": 10475
    },
    {
      "epoch": 4.941065535124941,
      "grad_norm": 0.0021908835042268038,
      "learning_rate": 0.25200164495481525,
      "loss": 0.2265,
      "num_input_tokens_seen": 9078816,
      "step": 10480
    },
    {
      "epoch": 4.943422913719943,
      "grad_norm": 0.003680185414850712,
      "learning_rate": 0.25195844793164474,
      "loss": 0.2227,
      "num_input_tokens_seen": 9083712,
      "step": 10485
    },
    {
      "epoch": 4.945780292314946,
      "grad_norm": 0.0018576316069811583,
      "learning_rate": 0.2519152351852001,
      "loss": 0.2193,
      "num_input_tokens_seen": 9088880,
      "step": 10490
    },
    {
      "epoch": 4.948137670909948,
      "grad_norm": 0.0018354365602135658,
      "learning_rate": 0.25187200672214555,
      "loss": 0.2327,
      "num_input_tokens_seen": 9093760,
      "step": 10495
    },
    {
      "epoch": 4.9504950495049505,
      "grad_norm": 0.0014261736068874598,
      "learning_rate": 0.2518287625491473,
      "loss": 0.1943,
      "num_input_tokens_seen": 9098944,
      "step": 10500
    },
    {
      "epoch": 4.952852428099953,
      "grad_norm": 0.0025744158774614334,
      "learning_rate": 0.25178550267287425,
      "loss": 0.1975,
      "num_input_tokens_seen": 9102640,
      "step": 10505
    },
    {
      "epoch": 4.955209806694955,
      "grad_norm": 0.00256257108412683,
      "learning_rate": 0.2517422270999976,
      "loss": 0.2771,
      "num_input_tokens_seen": 9107072,
      "step": 10510
    },
    {
      "epoch": 4.957567185289958,
      "grad_norm": 0.0013107958948239684,
      "learning_rate": 0.2516989358371909,
      "loss": 0.2114,
      "num_input_tokens_seen": 9111216,
      "step": 10515
    },
    {
      "epoch": 4.95992456388496,
      "grad_norm": 0.0016697305254638195,
      "learning_rate": 0.25165562889113025,
      "loss": 0.2262,
      "num_input_tokens_seen": 9116352,
      "step": 10520
    },
    {
      "epoch": 4.9622819424799625,
      "grad_norm": 0.0015130338724702597,
      "learning_rate": 0.2516123062684942,
      "loss": 0.2282,
      "num_input_tokens_seen": 9120528,
      "step": 10525
    },
    {
      "epoch": 4.964639321074965,
      "grad_norm": 0.0024297789204865694,
      "learning_rate": 0.25156896797596356,
      "loss": 0.2233,
      "num_input_tokens_seen": 9124560,
      "step": 10530
    },
    {
      "epoch": 4.966996699669967,
      "grad_norm": 0.0021261225920170546,
      "learning_rate": 0.2515256140202216,
      "loss": 0.2211,
      "num_input_tokens_seen": 9129120,
      "step": 10535
    },
    {
      "epoch": 4.96935407826497,
      "grad_norm": 0.001633642241358757,
      "learning_rate": 0.25148224440795425,
      "loss": 0.2366,
      "num_input_tokens_seen": 9132960,
      "step": 10540
    },
    {
      "epoch": 4.971711456859972,
      "grad_norm": 0.001610517967492342,
      "learning_rate": 0.2514388591458494,
      "loss": 0.212,
      "num_input_tokens_seen": 9137216,
      "step": 10545
    },
    {
      "epoch": 4.974068835454974,
      "grad_norm": 0.0018851879285648465,
      "learning_rate": 0.2513954582405977,
      "loss": 0.2295,
      "num_input_tokens_seen": 9141776,
      "step": 10550
    },
    {
      "epoch": 4.976426214049976,
      "grad_norm": 0.0026965998113155365,
      "learning_rate": 0.2513520416988922,
      "loss": 0.2289,
      "num_input_tokens_seen": 9145696,
      "step": 10555
    },
    {
      "epoch": 4.978783592644978,
      "grad_norm": 0.00170559820253402,
      "learning_rate": 0.2513086095274281,
      "loss": 0.2249,
      "num_input_tokens_seen": 9149904,
      "step": 10560
    },
    {
      "epoch": 4.981140971239981,
      "grad_norm": 0.001909121172502637,
      "learning_rate": 0.25126516173290336,
      "loss": 0.2299,
      "num_input_tokens_seen": 9154544,
      "step": 10565
    },
    {
      "epoch": 4.983498349834983,
      "grad_norm": 0.0016581149538978934,
      "learning_rate": 0.2512216983220181,
      "loss": 0.2198,
      "num_input_tokens_seen": 9158768,
      "step": 10570
    },
    {
      "epoch": 4.985855728429986,
      "grad_norm": 0.0031685964204370975,
      "learning_rate": 0.25117821930147494,
      "loss": 0.2033,
      "num_input_tokens_seen": 9163232,
      "step": 10575
    },
    {
      "epoch": 4.988213107024988,
      "grad_norm": 0.002248473232612014,
      "learning_rate": 0.2511347246779788,
      "loss": 0.1533,
      "num_input_tokens_seen": 9167456,
      "step": 10580
    },
    {
      "epoch": 4.99057048561999,
      "grad_norm": 0.005142302718013525,
      "learning_rate": 0.25109121445823723,
      "loss": 0.3131,
      "num_input_tokens_seen": 9172288,
      "step": 10585
    },
    {
      "epoch": 4.992927864214993,
      "grad_norm": 0.001775240758433938,
      "learning_rate": 0.25104768864896004,
      "loss": 0.2691,
      "num_input_tokens_seen": 9176736,
      "step": 10590
    },
    {
      "epoch": 4.995285242809995,
      "grad_norm": 0.0018103191396221519,
      "learning_rate": 0.2510041472568594,
      "loss": 0.2095,
      "num_input_tokens_seen": 9180880,
      "step": 10595
    },
    {
      "epoch": 4.997642621404998,
      "grad_norm": 0.001738940947689116,
      "learning_rate": 0.25096059028864987,
      "loss": 0.2256,
      "num_input_tokens_seen": 9185584,
      "step": 10600
    },
    {
      "epoch": 4.997642621404998,
      "eval_loss": 0.22060146927833557,
      "eval_runtime": 21.9577,
      "eval_samples_per_second": 42.946,
      "eval_steps_per_second": 21.496,
      "num_input_tokens_seen": 9185584,
      "step": 10600
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0017867760034278035,
      "learning_rate": 0.25091701775104863,
      "loss": 0.2195,
      "num_input_tokens_seen": 9189824,
      "step": 10605
    },
    {
      "epoch": 5.002357378595002,
      "grad_norm": 0.0036467607133090496,
      "learning_rate": 0.250873429650775,
      "loss": 0.2468,
      "num_input_tokens_seen": 9194368,
      "step": 10610
    },
    {
      "epoch": 5.004714757190005,
      "grad_norm": 0.0017336803721264005,
      "learning_rate": 0.25082982599455095,
      "loss": 0.2228,
      "num_input_tokens_seen": 9199328,
      "step": 10615
    },
    {
      "epoch": 5.007072135785007,
      "grad_norm": 0.0011912789195775986,
      "learning_rate": 0.2507862067891006,
      "loss": 0.2274,
      "num_input_tokens_seen": 9202896,
      "step": 10620
    },
    {
      "epoch": 5.00942951438001,
      "grad_norm": 0.0023062515538185835,
      "learning_rate": 0.25074257204115064,
      "loss": 0.2302,
      "num_input_tokens_seen": 9207088,
      "step": 10625
    },
    {
      "epoch": 5.011786892975012,
      "grad_norm": 0.001548964879475534,
      "learning_rate": 0.25069892175742997,
      "loss": 0.2188,
      "num_input_tokens_seen": 9211488,
      "step": 10630
    },
    {
      "epoch": 5.014144271570014,
      "grad_norm": 0.0010844528442248702,
      "learning_rate": 0.25065525594467014,
      "loss": 0.2214,
      "num_input_tokens_seen": 9215024,
      "step": 10635
    },
    {
      "epoch": 5.016501650165017,
      "grad_norm": 0.0024387016892433167,
      "learning_rate": 0.2506115746096049,
      "loss": 0.1965,
      "num_input_tokens_seen": 9218928,
      "step": 10640
    },
    {
      "epoch": 5.018859028760019,
      "grad_norm": 0.0014010679442435503,
      "learning_rate": 0.25056787775897055,
      "loss": 0.2221,
      "num_input_tokens_seen": 9222992,
      "step": 10645
    },
    {
      "epoch": 5.021216407355022,
      "grad_norm": 0.0014597059926018119,
      "learning_rate": 0.2505241653995056,
      "loss": 0.1899,
      "num_input_tokens_seen": 9227264,
      "step": 10650
    },
    {
      "epoch": 5.023573785950024,
      "grad_norm": 0.001042926567606628,
      "learning_rate": 0.25048043753795113,
      "loss": 0.1737,
      "num_input_tokens_seen": 9231808,
      "step": 10655
    },
    {
      "epoch": 5.0259311645450255,
      "grad_norm": 0.0010163764236494899,
      "learning_rate": 0.2504366941810504,
      "loss": 0.1785,
      "num_input_tokens_seen": 9237072,
      "step": 10660
    },
    {
      "epoch": 5.028288543140028,
      "grad_norm": 0.003876360598951578,
      "learning_rate": 0.2503929353355493,
      "loss": 0.2262,
      "num_input_tokens_seen": 9241808,
      "step": 10665
    },
    {
      "epoch": 5.03064592173503,
      "grad_norm": 0.002344526583328843,
      "learning_rate": 0.250349161008196,
      "loss": 0.1764,
      "num_input_tokens_seen": 9245840,
      "step": 10670
    },
    {
      "epoch": 5.033003300330033,
      "grad_norm": 0.0014534560032188892,
      "learning_rate": 0.2503053712057409,
      "loss": 0.2453,
      "num_input_tokens_seen": 9249728,
      "step": 10675
    },
    {
      "epoch": 5.035360678925035,
      "grad_norm": 0.002834377344697714,
      "learning_rate": 0.25026156593493715,
      "loss": 0.2956,
      "num_input_tokens_seen": 9254080,
      "step": 10680
    },
    {
      "epoch": 5.0377180575200375,
      "grad_norm": 0.0026421749498695135,
      "learning_rate": 0.2502177452025399,
      "loss": 0.2512,
      "num_input_tokens_seen": 9257952,
      "step": 10685
    },
    {
      "epoch": 5.04007543611504,
      "grad_norm": 0.0031543264631181955,
      "learning_rate": 0.25017390901530695,
      "loss": 0.265,
      "num_input_tokens_seen": 9261568,
      "step": 10690
    },
    {
      "epoch": 5.042432814710042,
      "grad_norm": 0.0021345820277929306,
      "learning_rate": 0.2501300573799984,
      "loss": 0.2388,
      "num_input_tokens_seen": 9266656,
      "step": 10695
    },
    {
      "epoch": 5.044790193305045,
      "grad_norm": 0.0032463984098285437,
      "learning_rate": 0.2500861903033766,
      "loss": 0.2296,
      "num_input_tokens_seen": 9271424,
      "step": 10700
    },
    {
      "epoch": 5.047147571900047,
      "grad_norm": 0.004770686384290457,
      "learning_rate": 0.25004230779220654,
      "loss": 0.2875,
      "num_input_tokens_seen": 9275696,
      "step": 10705
    },
    {
      "epoch": 5.0495049504950495,
      "grad_norm": 0.0016110694268718362,
      "learning_rate": 0.24999840985325542,
      "loss": 0.1924,
      "num_input_tokens_seen": 9280176,
      "step": 10710
    },
    {
      "epoch": 5.051862329090052,
      "grad_norm": 0.002559440676122904,
      "learning_rate": 0.24995449649329285,
      "loss": 0.2475,
      "num_input_tokens_seen": 9284528,
      "step": 10715
    },
    {
      "epoch": 5.054219707685054,
      "grad_norm": 0.0018230616115033627,
      "learning_rate": 0.2499105677190908,
      "loss": 0.2346,
      "num_input_tokens_seen": 9288912,
      "step": 10720
    },
    {
      "epoch": 5.056577086280057,
      "grad_norm": 0.0021518957801163197,
      "learning_rate": 0.24986662353742364,
      "loss": 0.2208,
      "num_input_tokens_seen": 9293120,
      "step": 10725
    },
    {
      "epoch": 5.058934464875059,
      "grad_norm": 0.0015826355665922165,
      "learning_rate": 0.24982266395506814,
      "loss": 0.2214,
      "num_input_tokens_seen": 9297088,
      "step": 10730
    },
    {
      "epoch": 5.061291843470062,
      "grad_norm": 0.001485955435782671,
      "learning_rate": 0.2497786889788034,
      "loss": 0.2091,
      "num_input_tokens_seen": 9300784,
      "step": 10735
    },
    {
      "epoch": 5.063649222065064,
      "grad_norm": 0.002488410333171487,
      "learning_rate": 0.24973469861541095,
      "loss": 0.2272,
      "num_input_tokens_seen": 9305200,
      "step": 10740
    },
    {
      "epoch": 5.066006600660066,
      "grad_norm": 0.0023279262240976095,
      "learning_rate": 0.24969069287167456,
      "loss": 0.213,
      "num_input_tokens_seen": 9309568,
      "step": 10745
    },
    {
      "epoch": 5.068363979255069,
      "grad_norm": 0.0015863998560234904,
      "learning_rate": 0.2496466717543806,
      "loss": 0.1581,
      "num_input_tokens_seen": 9314048,
      "step": 10750
    },
    {
      "epoch": 5.07072135785007,
      "grad_norm": 0.001054070657119155,
      "learning_rate": 0.24960263527031762,
      "loss": 0.2223,
      "num_input_tokens_seen": 9318416,
      "step": 10755
    },
    {
      "epoch": 5.073078736445073,
      "grad_norm": 0.0022177959326654673,
      "learning_rate": 0.24955858342627657,
      "loss": 0.2321,
      "num_input_tokens_seen": 9322880,
      "step": 10760
    },
    {
      "epoch": 5.075436115040075,
      "grad_norm": 0.005402293987572193,
      "learning_rate": 0.24951451622905083,
      "loss": 0.2653,
      "num_input_tokens_seen": 9327584,
      "step": 10765
    },
    {
      "epoch": 5.0777934936350775,
      "grad_norm": 0.004284504801034927,
      "learning_rate": 0.24947043368543612,
      "loss": 0.2304,
      "num_input_tokens_seen": 9331984,
      "step": 10770
    },
    {
      "epoch": 5.08015087223008,
      "grad_norm": 0.0022657192312180996,
      "learning_rate": 0.2494263358022305,
      "loss": 0.2208,
      "num_input_tokens_seen": 9336384,
      "step": 10775
    },
    {
      "epoch": 5.082508250825082,
      "grad_norm": 0.002195975976064801,
      "learning_rate": 0.24938222258623444,
      "loss": 0.2254,
      "num_input_tokens_seen": 9340352,
      "step": 10780
    },
    {
      "epoch": 5.084865629420085,
      "grad_norm": 0.0014843599637970328,
      "learning_rate": 0.24933809404425075,
      "loss": 0.2218,
      "num_input_tokens_seen": 9344656,
      "step": 10785
    },
    {
      "epoch": 5.087223008015087,
      "grad_norm": 0.0013890491100028157,
      "learning_rate": 0.24929395018308453,
      "loss": 0.1925,
      "num_input_tokens_seen": 9349920,
      "step": 10790
    },
    {
      "epoch": 5.0895803866100895,
      "grad_norm": 0.0023382019717246294,
      "learning_rate": 0.24924979100954348,
      "loss": 0.2345,
      "num_input_tokens_seen": 9354032,
      "step": 10795
    },
    {
      "epoch": 5.091937765205092,
      "grad_norm": 0.0033197090961039066,
      "learning_rate": 0.24920561653043735,
      "loss": 0.2219,
      "num_input_tokens_seen": 9358160,
      "step": 10800
    },
    {
      "epoch": 5.091937765205092,
      "eval_loss": 0.22373640537261963,
      "eval_runtime": 21.9153,
      "eval_samples_per_second": 43.029,
      "eval_steps_per_second": 21.537,
      "num_input_tokens_seen": 9358160,
      "step": 10800
    },
    {
      "epoch": 5.094295143800094,
      "grad_norm": 0.0014296055305749178,
      "learning_rate": 0.24916142675257846,
      "loss": 0.2433,
      "num_input_tokens_seen": 9362512,
      "step": 10805
    },
    {
      "epoch": 5.096652522395097,
      "grad_norm": 0.002428998937830329,
      "learning_rate": 0.24911722168278144,
      "loss": 0.2581,
      "num_input_tokens_seen": 9366928,
      "step": 10810
    },
    {
      "epoch": 5.099009900990099,
      "grad_norm": 0.0014241700991988182,
      "learning_rate": 0.24907300132786328,
      "loss": 0.2217,
      "num_input_tokens_seen": 9371136,
      "step": 10815
    },
    {
      "epoch": 5.1013672795851015,
      "grad_norm": 0.001293994253501296,
      "learning_rate": 0.24902876569464322,
      "loss": 0.2266,
      "num_input_tokens_seen": 9375264,
      "step": 10820
    },
    {
      "epoch": 5.103724658180104,
      "grad_norm": 0.0021044372115284204,
      "learning_rate": 0.24898451478994305,
      "loss": 0.2349,
      "num_input_tokens_seen": 9380128,
      "step": 10825
    },
    {
      "epoch": 5.106082036775106,
      "grad_norm": 0.0018773652845993638,
      "learning_rate": 0.2489402486205868,
      "loss": 0.2334,
      "num_input_tokens_seen": 9384512,
      "step": 10830
    },
    {
      "epoch": 5.108439415370109,
      "grad_norm": 0.001599509734660387,
      "learning_rate": 0.24889596719340085,
      "loss": 0.2391,
      "num_input_tokens_seen": 9389424,
      "step": 10835
    },
    {
      "epoch": 5.110796793965111,
      "grad_norm": 0.001800917205400765,
      "learning_rate": 0.24885167051521392,
      "loss": 0.2362,
      "num_input_tokens_seen": 9393344,
      "step": 10840
    },
    {
      "epoch": 5.1131541725601135,
      "grad_norm": 0.0018363536801189184,
      "learning_rate": 0.24880735859285716,
      "loss": 0.2372,
      "num_input_tokens_seen": 9397904,
      "step": 10845
    },
    {
      "epoch": 5.115511551155116,
      "grad_norm": 0.0012843572767451406,
      "learning_rate": 0.24876303143316406,
      "loss": 0.226,
      "num_input_tokens_seen": 9401776,
      "step": 10850
    },
    {
      "epoch": 5.117868929750118,
      "grad_norm": 0.0014027270954102278,
      "learning_rate": 0.24871868904297031,
      "loss": 0.2172,
      "num_input_tokens_seen": 9405936,
      "step": 10855
    },
    {
      "epoch": 5.12022630834512,
      "grad_norm": 0.0011954716173931956,
      "learning_rate": 0.24867433142911416,
      "loss": 0.189,
      "num_input_tokens_seen": 9410640,
      "step": 10860
    },
    {
      "epoch": 5.122583686940122,
      "grad_norm": 0.003320219460874796,
      "learning_rate": 0.24862995859843612,
      "loss": 0.2892,
      "num_input_tokens_seen": 9415120,
      "step": 10865
    },
    {
      "epoch": 5.124941065535125,
      "grad_norm": 0.0019702287390828133,
      "learning_rate": 0.24858557055777897,
      "loss": 0.208,
      "num_input_tokens_seen": 9419248,
      "step": 10870
    },
    {
      "epoch": 5.127298444130127,
      "grad_norm": 0.0027095286641269922,
      "learning_rate": 0.24854116731398793,
      "loss": 0.22,
      "num_input_tokens_seen": 9424032,
      "step": 10875
    },
    {
      "epoch": 5.129655822725129,
      "grad_norm": 0.0013799879234284163,
      "learning_rate": 0.24849674887391052,
      "loss": 0.2408,
      "num_input_tokens_seen": 9428400,
      "step": 10880
    },
    {
      "epoch": 5.132013201320132,
      "grad_norm": 0.002387586748227477,
      "learning_rate": 0.2484523152443967,
      "loss": 0.2271,
      "num_input_tokens_seen": 9432944,
      "step": 10885
    },
    {
      "epoch": 5.134370579915134,
      "grad_norm": 0.0014568858314305544,
      "learning_rate": 0.24840786643229862,
      "loss": 0.2392,
      "num_input_tokens_seen": 9437792,
      "step": 10890
    },
    {
      "epoch": 5.136727958510137,
      "grad_norm": 0.001729584881104529,
      "learning_rate": 0.2483634024444709,
      "loss": 0.2245,
      "num_input_tokens_seen": 9441600,
      "step": 10895
    },
    {
      "epoch": 5.139085337105139,
      "grad_norm": 0.0015277467900887132,
      "learning_rate": 0.24831892328777033,
      "loss": 0.2117,
      "num_input_tokens_seen": 9446064,
      "step": 10900
    },
    {
      "epoch": 5.141442715700141,
      "grad_norm": 0.0015412027714774013,
      "learning_rate": 0.2482744289690563,
      "loss": 0.257,
      "num_input_tokens_seen": 9451536,
      "step": 10905
    },
    {
      "epoch": 5.143800094295144,
      "grad_norm": 0.0018606801750138402,
      "learning_rate": 0.2482299194951903,
      "loss": 0.2441,
      "num_input_tokens_seen": 9455904,
      "step": 10910
    },
    {
      "epoch": 5.146157472890146,
      "grad_norm": 0.0035080555826425552,
      "learning_rate": 0.2481853948730363,
      "loss": 0.2345,
      "num_input_tokens_seen": 9461824,
      "step": 10915
    },
    {
      "epoch": 5.148514851485149,
      "grad_norm": 0.0033204746432602406,
      "learning_rate": 0.24814085510946052,
      "loss": 0.2334,
      "num_input_tokens_seen": 9465440,
      "step": 10920
    },
    {
      "epoch": 5.150872230080151,
      "grad_norm": 0.002022932982072234,
      "learning_rate": 0.24809630021133158,
      "loss": 0.2345,
      "num_input_tokens_seen": 9470400,
      "step": 10925
    },
    {
      "epoch": 5.1532296086751535,
      "grad_norm": 0.0012817138340324163,
      "learning_rate": 0.24805173018552037,
      "loss": 0.2248,
      "num_input_tokens_seen": 9474496,
      "step": 10930
    },
    {
      "epoch": 5.155586987270156,
      "grad_norm": 0.004304759204387665,
      "learning_rate": 0.2480071450389002,
      "loss": 0.2364,
      "num_input_tokens_seen": 9478784,
      "step": 10935
    },
    {
      "epoch": 5.157944365865158,
      "grad_norm": 0.0015767533332109451,
      "learning_rate": 0.24796254477834662,
      "loss": 0.2313,
      "num_input_tokens_seen": 9482448,
      "step": 10940
    },
    {
      "epoch": 5.160301744460161,
      "grad_norm": 0.0025463809724897146,
      "learning_rate": 0.24791792941073754,
      "loss": 0.2257,
      "num_input_tokens_seen": 9486784,
      "step": 10945
    },
    {
      "epoch": 5.162659123055163,
      "grad_norm": 0.0018110288074240088,
      "learning_rate": 0.2478732989429533,
      "loss": 0.2466,
      "num_input_tokens_seen": 9491216,
      "step": 10950
    },
    {
      "epoch": 5.165016501650165,
      "grad_norm": 0.0022011820692569017,
      "learning_rate": 0.24782865338187632,
      "loss": 0.2259,
      "num_input_tokens_seen": 9495536,
      "step": 10955
    },
    {
      "epoch": 5.167373880245167,
      "grad_norm": 0.0015602281782776117,
      "learning_rate": 0.2477839927343916,
      "loss": 0.2068,
      "num_input_tokens_seen": 9499552,
      "step": 10960
    },
    {
      "epoch": 5.169731258840169,
      "grad_norm": 0.002355400938540697,
      "learning_rate": 0.2477393170073864,
      "loss": 0.1968,
      "num_input_tokens_seen": 9503712,
      "step": 10965
    },
    {
      "epoch": 5.172088637435172,
      "grad_norm": 0.0022267696913331747,
      "learning_rate": 0.2476946262077503,
      "loss": 0.1574,
      "num_input_tokens_seen": 9508544,
      "step": 10970
    },
    {
      "epoch": 5.174446016030174,
      "grad_norm": 0.00301539758220315,
      "learning_rate": 0.24764992034237507,
      "loss": 0.1935,
      "num_input_tokens_seen": 9512704,
      "step": 10975
    },
    {
      "epoch": 5.176803394625177,
      "grad_norm": 0.001113447593525052,
      "learning_rate": 0.24760519941815498,
      "loss": 0.1519,
      "num_input_tokens_seen": 9517216,
      "step": 10980
    },
    {
      "epoch": 5.179160773220179,
      "grad_norm": 0.003405479248613119,
      "learning_rate": 0.2475604634419866,
      "loss": 0.3768,
      "num_input_tokens_seen": 9522416,
      "step": 10985
    },
    {
      "epoch": 5.181518151815181,
      "grad_norm": 0.0032278024591505527,
      "learning_rate": 0.24751571242076872,
      "loss": 0.2224,
      "num_input_tokens_seen": 9526400,
      "step": 10990
    },
    {
      "epoch": 5.183875530410184,
      "grad_norm": 0.0029130878392606974,
      "learning_rate": 0.2474709463614025,
      "loss": 0.2318,
      "num_input_tokens_seen": 9531536,
      "step": 10995
    },
    {
      "epoch": 5.186232909005186,
      "grad_norm": 0.001300496980547905,
      "learning_rate": 0.24742616527079145,
      "loss": 0.2263,
      "num_input_tokens_seen": 9535520,
      "step": 11000
    },
    {
      "epoch": 5.186232909005186,
      "eval_loss": 0.22649720311164856,
      "eval_runtime": 21.9432,
      "eval_samples_per_second": 42.975,
      "eval_steps_per_second": 21.51,
      "num_input_tokens_seen": 9535520,
      "step": 11000
    },
    {
      "epoch": 5.188590287600189,
      "grad_norm": 0.0025213537737727165,
      "learning_rate": 0.24738136915584139,
      "loss": 0.2392,
      "num_input_tokens_seen": 9539696,
      "step": 11005
    },
    {
      "epoch": 5.190947666195191,
      "grad_norm": 0.003846134524792433,
      "learning_rate": 0.24733655802346047,
      "loss": 0.2325,
      "num_input_tokens_seen": 9543792,
      "step": 11010
    },
    {
      "epoch": 5.193305044790193,
      "grad_norm": 0.0015698407078161836,
      "learning_rate": 0.24729173188055906,
      "loss": 0.2052,
      "num_input_tokens_seen": 9548896,
      "step": 11015
    },
    {
      "epoch": 5.195662423385196,
      "grad_norm": 0.0019078021869063377,
      "learning_rate": 0.24724689073404996,
      "loss": 0.1879,
      "num_input_tokens_seen": 9552944,
      "step": 11020
    },
    {
      "epoch": 5.198019801980198,
      "grad_norm": 0.0019758448470383883,
      "learning_rate": 0.24720203459084822,
      "loss": 0.2087,
      "num_input_tokens_seen": 9557840,
      "step": 11025
    },
    {
      "epoch": 5.200377180575201,
      "grad_norm": 0.0024560161400586367,
      "learning_rate": 0.24715716345787123,
      "loss": 0.2588,
      "num_input_tokens_seen": 9562064,
      "step": 11030
    },
    {
      "epoch": 5.202734559170203,
      "grad_norm": 0.0019600773230195045,
      "learning_rate": 0.2471122773420387,
      "loss": 0.2117,
      "num_input_tokens_seen": 9566832,
      "step": 11035
    },
    {
      "epoch": 5.205091937765205,
      "grad_norm": 0.0015101187163963914,
      "learning_rate": 0.24706737625027259,
      "loss": 0.2467,
      "num_input_tokens_seen": 9571520,
      "step": 11040
    },
    {
      "epoch": 5.207449316360208,
      "grad_norm": 0.0017958359094336629,
      "learning_rate": 0.24702246018949725,
      "loss": 0.2279,
      "num_input_tokens_seen": 9576272,
      "step": 11045
    },
    {
      "epoch": 5.20980669495521,
      "grad_norm": 0.0020629449281841516,
      "learning_rate": 0.2469775291666393,
      "loss": 0.23,
      "num_input_tokens_seen": 9580704,
      "step": 11050
    },
    {
      "epoch": 5.212164073550213,
      "grad_norm": 0.002542410744354129,
      "learning_rate": 0.24693258318862765,
      "loss": 0.2315,
      "num_input_tokens_seen": 9585712,
      "step": 11055
    },
    {
      "epoch": 5.214521452145214,
      "grad_norm": 0.0016957283951342106,
      "learning_rate": 0.2468876222623935,
      "loss": 0.2166,
      "num_input_tokens_seen": 9590208,
      "step": 11060
    },
    {
      "epoch": 5.2168788307402165,
      "grad_norm": 0.0017622999148443341,
      "learning_rate": 0.2468426463948705,
      "loss": 0.2073,
      "num_input_tokens_seen": 9594672,
      "step": 11065
    },
    {
      "epoch": 5.219236209335219,
      "grad_norm": 0.001091672107577324,
      "learning_rate": 0.24679765559299438,
      "loss": 0.2344,
      "num_input_tokens_seen": 9599856,
      "step": 11070
    },
    {
      "epoch": 5.221593587930221,
      "grad_norm": 0.002150794258341193,
      "learning_rate": 0.24675264986370332,
      "loss": 0.1879,
      "num_input_tokens_seen": 9603440,
      "step": 11075
    },
    {
      "epoch": 5.223950966525224,
      "grad_norm": 0.0035143110435456038,
      "learning_rate": 0.2467076292139378,
      "loss": 0.2845,
      "num_input_tokens_seen": 9606960,
      "step": 11080
    },
    {
      "epoch": 5.226308345120226,
      "grad_norm": 0.0012609228724613786,
      "learning_rate": 0.24666259365064055,
      "loss": 0.2166,
      "num_input_tokens_seen": 9610752,
      "step": 11085
    },
    {
      "epoch": 5.2286657237152285,
      "grad_norm": 0.0017185568576678634,
      "learning_rate": 0.24661754318075663,
      "loss": 0.232,
      "num_input_tokens_seen": 9614880,
      "step": 11090
    },
    {
      "epoch": 5.231023102310231,
      "grad_norm": 0.0016587821301072836,
      "learning_rate": 0.2465724778112334,
      "loss": 0.2389,
      "num_input_tokens_seen": 9619408,
      "step": 11095
    },
    {
      "epoch": 5.233380480905233,
      "grad_norm": 0.002515598898753524,
      "learning_rate": 0.24652739754902042,
      "loss": 0.2414,
      "num_input_tokens_seen": 9623216,
      "step": 11100
    },
    {
      "epoch": 5.235737859500236,
      "grad_norm": 0.00206698733381927,
      "learning_rate": 0.24648230240106975,
      "loss": 0.2596,
      "num_input_tokens_seen": 9628032,
      "step": 11105
    },
    {
      "epoch": 5.238095238095238,
      "grad_norm": 0.0016193678602576256,
      "learning_rate": 0.2464371923743356,
      "loss": 0.2367,
      "num_input_tokens_seen": 9633072,
      "step": 11110
    },
    {
      "epoch": 5.2404526166902405,
      "grad_norm": 0.0012147077359259129,
      "learning_rate": 0.24639206747577444,
      "loss": 0.181,
      "num_input_tokens_seen": 9637600,
      "step": 11115
    },
    {
      "epoch": 5.242809995285243,
      "grad_norm": 0.001052511390298605,
      "learning_rate": 0.24634692771234515,
      "loss": 0.2052,
      "num_input_tokens_seen": 9641504,
      "step": 11120
    },
    {
      "epoch": 5.245167373880245,
      "grad_norm": 0.0011815240141004324,
      "learning_rate": 0.2463017730910088,
      "loss": 0.2075,
      "num_input_tokens_seen": 9645696,
      "step": 11125
    },
    {
      "epoch": 5.247524752475248,
      "grad_norm": 0.004300659988075495,
      "learning_rate": 0.2462566036187289,
      "loss": 0.2364,
      "num_input_tokens_seen": 9650528,
      "step": 11130
    },
    {
      "epoch": 5.24988213107025,
      "grad_norm": 0.002963155275210738,
      "learning_rate": 0.24621141930247106,
      "loss": 0.3011,
      "num_input_tokens_seen": 9653888,
      "step": 11135
    },
    {
      "epoch": 5.2522395096652525,
      "grad_norm": 0.0019008241361007094,
      "learning_rate": 0.2461662201492033,
      "loss": 0.2131,
      "num_input_tokens_seen": 9658768,
      "step": 11140
    },
    {
      "epoch": 5.254596888260255,
      "grad_norm": 0.0018571895780041814,
      "learning_rate": 0.24612100616589586,
      "loss": 0.2131,
      "num_input_tokens_seen": 9662976,
      "step": 11145
    },
    {
      "epoch": 5.256954266855257,
      "grad_norm": 0.0022801333107054234,
      "learning_rate": 0.24607577735952135,
      "loss": 0.17,
      "num_input_tokens_seen": 9667824,
      "step": 11150
    },
    {
      "epoch": 5.259311645450259,
      "grad_norm": 0.003788426984101534,
      "learning_rate": 0.24603053373705464,
      "loss": 0.2448,
      "num_input_tokens_seen": 9672864,
      "step": 11155
    },
    {
      "epoch": 5.261669024045261,
      "grad_norm": 0.002434577327221632,
      "learning_rate": 0.2459852753054728,
      "loss": 0.2744,
      "num_input_tokens_seen": 9676944,
      "step": 11160
    },
    {
      "epoch": 5.264026402640264,
      "grad_norm": 0.0022088466212153435,
      "learning_rate": 0.24594000207175526,
      "loss": 0.2326,
      "num_input_tokens_seen": 9681024,
      "step": 11165
    },
    {
      "epoch": 5.266383781235266,
      "grad_norm": 0.0017202841117978096,
      "learning_rate": 0.2458947140428838,
      "loss": 0.2408,
      "num_input_tokens_seen": 9685936,
      "step": 11170
    },
    {
      "epoch": 5.2687411598302685,
      "grad_norm": 0.004185760859400034,
      "learning_rate": 0.24584941122584233,
      "loss": 0.2475,
      "num_input_tokens_seen": 9689376,
      "step": 11175
    },
    {
      "epoch": 5.271098538425271,
      "grad_norm": 0.0024647661484777927,
      "learning_rate": 0.24580409362761713,
      "loss": 0.2344,
      "num_input_tokens_seen": 9693728,
      "step": 11180
    },
    {
      "epoch": 5.273455917020273,
      "grad_norm": 0.001956983469426632,
      "learning_rate": 0.2457587612551967,
      "loss": 0.2366,
      "num_input_tokens_seen": 9697952,
      "step": 11185
    },
    {
      "epoch": 5.275813295615276,
      "grad_norm": 0.0020128004252910614,
      "learning_rate": 0.24571341411557193,
      "loss": 0.192,
      "num_input_tokens_seen": 9701840,
      "step": 11190
    },
    {
      "epoch": 5.278170674210278,
      "grad_norm": 0.0027684196829795837,
      "learning_rate": 0.2456680522157359,
      "loss": 0.2245,
      "num_input_tokens_seen": 9705248,
      "step": 11195
    },
    {
      "epoch": 5.2805280528052805,
      "grad_norm": 0.0020245984196662903,
      "learning_rate": 0.245622675562684,
      "loss": 0.185,
      "num_input_tokens_seen": 9709232,
      "step": 11200
    },
    {
      "epoch": 5.2805280528052805,
      "eval_loss": 0.22369828820228577,
      "eval_runtime": 21.9379,
      "eval_samples_per_second": 42.985,
      "eval_steps_per_second": 21.515,
      "num_input_tokens_seen": 9709232,
      "step": 11200
    },
    {
      "epoch": 5.282885431400283,
      "grad_norm": 0.0013713166117668152,
      "learning_rate": 0.24557728416341384,
      "loss": 0.1959,
      "num_input_tokens_seen": 9712928,
      "step": 11205
    },
    {
      "epoch": 5.285242809995285,
      "grad_norm": 0.004494520835578442,
      "learning_rate": 0.24553187802492538,
      "loss": 0.2637,
      "num_input_tokens_seen": 9717616,
      "step": 11210
    },
    {
      "epoch": 5.287600188590288,
      "grad_norm": 0.0011971939820796251,
      "learning_rate": 0.24548645715422074,
      "loss": 0.1907,
      "num_input_tokens_seen": 9721968,
      "step": 11215
    },
    {
      "epoch": 5.28995756718529,
      "grad_norm": 0.0020063628908246756,
      "learning_rate": 0.2454410215583045,
      "loss": 0.1842,
      "num_input_tokens_seen": 9726688,
      "step": 11220
    },
    {
      "epoch": 5.2923149457802925,
      "grad_norm": 0.0018448256887495518,
      "learning_rate": 0.24539557124418332,
      "loss": 0.2395,
      "num_input_tokens_seen": 9730912,
      "step": 11225
    },
    {
      "epoch": 5.294672324375295,
      "grad_norm": 0.004227141849696636,
      "learning_rate": 0.24535010621886624,
      "loss": 0.2137,
      "num_input_tokens_seen": 9735136,
      "step": 11230
    },
    {
      "epoch": 5.297029702970297,
      "grad_norm": 0.003618549322709441,
      "learning_rate": 0.2453046264893646,
      "loss": 0.2505,
      "num_input_tokens_seen": 9739568,
      "step": 11235
    },
    {
      "epoch": 5.2993870815653,
      "grad_norm": 0.003519725287333131,
      "learning_rate": 0.24525913206269184,
      "loss": 0.2172,
      "num_input_tokens_seen": 9744976,
      "step": 11240
    },
    {
      "epoch": 5.301744460160302,
      "grad_norm": 0.00260105705820024,
      "learning_rate": 0.2452136229458638,
      "loss": 0.2164,
      "num_input_tokens_seen": 9748864,
      "step": 11245
    },
    {
      "epoch": 5.3041018387553045,
      "grad_norm": 0.0012512417742982507,
      "learning_rate": 0.24516809914589857,
      "loss": 0.2193,
      "num_input_tokens_seen": 9753840,
      "step": 11250
    },
    {
      "epoch": 5.306459217350307,
      "grad_norm": 0.0020031060557812452,
      "learning_rate": 0.2451225606698165,
      "loss": 0.214,
      "num_input_tokens_seen": 9758048,
      "step": 11255
    },
    {
      "epoch": 5.308816595945308,
      "grad_norm": 0.0031816039700061083,
      "learning_rate": 0.2450770075246402,
      "loss": 0.209,
      "num_input_tokens_seen": 9762000,
      "step": 11260
    },
    {
      "epoch": 5.311173974540311,
      "grad_norm": 0.0019977495539933443,
      "learning_rate": 0.24503143971739455,
      "loss": 0.2172,
      "num_input_tokens_seen": 9766112,
      "step": 11265
    },
    {
      "epoch": 5.313531353135313,
      "grad_norm": 0.0014568602200597525,
      "learning_rate": 0.24498585725510663,
      "loss": 0.2284,
      "num_input_tokens_seen": 9770160,
      "step": 11270
    },
    {
      "epoch": 5.315888731730316,
      "grad_norm": 0.0016768709756433964,
      "learning_rate": 0.24494026014480583,
      "loss": 0.2067,
      "num_input_tokens_seen": 9774816,
      "step": 11275
    },
    {
      "epoch": 5.318246110325318,
      "grad_norm": 0.0015292579773813486,
      "learning_rate": 0.24489464839352387,
      "loss": 0.2267,
      "num_input_tokens_seen": 9778672,
      "step": 11280
    },
    {
      "epoch": 5.32060348892032,
      "grad_norm": 0.0012707986170426011,
      "learning_rate": 0.2448490220082946,
      "loss": 0.2421,
      "num_input_tokens_seen": 9782000,
      "step": 11285
    },
    {
      "epoch": 5.322960867515323,
      "grad_norm": 0.0020803394727408886,
      "learning_rate": 0.24480338099615415,
      "loss": 0.2306,
      "num_input_tokens_seen": 9786048,
      "step": 11290
    },
    {
      "epoch": 5.325318246110325,
      "grad_norm": 0.0019657425582408905,
      "learning_rate": 0.244757725364141,
      "loss": 0.225,
      "num_input_tokens_seen": 9790592,
      "step": 11295
    },
    {
      "epoch": 5.327675624705328,
      "grad_norm": 0.001957591623067856,
      "learning_rate": 0.24471205511929583,
      "loss": 0.2278,
      "num_input_tokens_seen": 9794752,
      "step": 11300
    },
    {
      "epoch": 5.33003300330033,
      "grad_norm": 0.0019496369641274214,
      "learning_rate": 0.24466637026866145,
      "loss": 0.208,
      "num_input_tokens_seen": 9799664,
      "step": 11305
    },
    {
      "epoch": 5.332390381895332,
      "grad_norm": 0.0014565993333235383,
      "learning_rate": 0.2446206708192832,
      "loss": 0.2128,
      "num_input_tokens_seen": 9804048,
      "step": 11310
    },
    {
      "epoch": 5.334747760490335,
      "grad_norm": 0.0026904584374278784,
      "learning_rate": 0.2445749567782084,
      "loss": 0.2055,
      "num_input_tokens_seen": 9808992,
      "step": 11315
    },
    {
      "epoch": 5.337105139085337,
      "grad_norm": 0.0033633927814662457,
      "learning_rate": 0.2445292281524868,
      "loss": 0.2635,
      "num_input_tokens_seen": 9813104,
      "step": 11320
    },
    {
      "epoch": 5.33946251768034,
      "grad_norm": 0.008481680415570736,
      "learning_rate": 0.24448348494917022,
      "loss": 0.2266,
      "num_input_tokens_seen": 9817456,
      "step": 11325
    },
    {
      "epoch": 5.341819896275342,
      "grad_norm": 0.002741200150921941,
      "learning_rate": 0.24443772717531295,
      "loss": 0.2308,
      "num_input_tokens_seen": 9822336,
      "step": 11330
    },
    {
      "epoch": 5.344177274870344,
      "grad_norm": 0.0037526795640587807,
      "learning_rate": 0.24439195483797138,
      "loss": 0.2327,
      "num_input_tokens_seen": 9827168,
      "step": 11335
    },
    {
      "epoch": 5.346534653465347,
      "grad_norm": 0.0030065695755183697,
      "learning_rate": 0.24434616794420416,
      "loss": 0.2331,
      "num_input_tokens_seen": 9830928,
      "step": 11340
    },
    {
      "epoch": 5.348892032060349,
      "grad_norm": 0.0017801984213292599,
      "learning_rate": 0.24430036650107223,
      "loss": 0.2388,
      "num_input_tokens_seen": 9834848,
      "step": 11345
    },
    {
      "epoch": 5.351249410655352,
      "grad_norm": 0.0017228694632649422,
      "learning_rate": 0.2442545505156387,
      "loss": 0.1973,
      "num_input_tokens_seen": 9838512,
      "step": 11350
    },
    {
      "epoch": 5.353606789250353,
      "grad_norm": 0.003936372697353363,
      "learning_rate": 0.24420871999496904,
      "loss": 0.2755,
      "num_input_tokens_seen": 9842640,
      "step": 11355
    },
    {
      "epoch": 5.355964167845356,
      "grad_norm": 0.003623110009357333,
      "learning_rate": 0.24416287494613084,
      "loss": 0.2492,
      "num_input_tokens_seen": 9847904,
      "step": 11360
    },
    {
      "epoch": 5.358321546440358,
      "grad_norm": 0.0025485383812338114,
      "learning_rate": 0.24411701537619399,
      "loss": 0.2245,
      "num_input_tokens_seen": 9851712,
      "step": 11365
    },
    {
      "epoch": 5.36067892503536,
      "grad_norm": 0.001961366506293416,
      "learning_rate": 0.24407114129223062,
      "loss": 0.2012,
      "num_input_tokens_seen": 9854816,
      "step": 11370
    },
    {
      "epoch": 5.363036303630363,
      "grad_norm": 0.002851983066648245,
      "learning_rate": 0.2440252527013151,
      "loss": 0.2313,
      "num_input_tokens_seen": 9859664,
      "step": 11375
    },
    {
      "epoch": 5.365393682225365,
      "grad_norm": 0.004325107205659151,
      "learning_rate": 0.24397934961052403,
      "loss": 0.2275,
      "num_input_tokens_seen": 9863440,
      "step": 11380
    },
    {
      "epoch": 5.367751060820368,
      "grad_norm": 0.0026469980366528034,
      "learning_rate": 0.24393343202693618,
      "loss": 0.2088,
      "num_input_tokens_seen": 9867600,
      "step": 11385
    },
    {
      "epoch": 5.37010843941537,
      "grad_norm": 0.003807930974289775,
      "learning_rate": 0.2438874999576327,
      "loss": 0.2024,
      "num_input_tokens_seen": 9872208,
      "step": 11390
    },
    {
      "epoch": 5.372465818010372,
      "grad_norm": 0.002371760318055749,
      "learning_rate": 0.24384155340969688,
      "loss": 0.1965,
      "num_input_tokens_seen": 9876544,
      "step": 11395
    },
    {
      "epoch": 5.374823196605375,
      "grad_norm": 0.0037363239098340273,
      "learning_rate": 0.24379559239021423,
      "loss": 0.1847,
      "num_input_tokens_seen": 9880896,
      "step": 11400
    },
    {
      "epoch": 5.374823196605375,
      "eval_loss": 0.2308521419763565,
      "eval_runtime": 21.9422,
      "eval_samples_per_second": 42.976,
      "eval_steps_per_second": 21.511,
      "num_input_tokens_seen": 9880896,
      "step": 11400
    },
    {
      "epoch": 5.377180575200377,
      "grad_norm": 0.0032352821435779333,
      "learning_rate": 0.2437496169062725,
      "loss": 0.219,
      "num_input_tokens_seen": 9885840,
      "step": 11405
    },
    {
      "epoch": 5.37953795379538,
      "grad_norm": 0.004335922654718161,
      "learning_rate": 0.24370362696496176,
      "loss": 0.2603,
      "num_input_tokens_seen": 9889856,
      "step": 11410
    },
    {
      "epoch": 5.381895332390382,
      "grad_norm": 0.005886504892259836,
      "learning_rate": 0.24365762257337417,
      "loss": 0.2326,
      "num_input_tokens_seen": 9894976,
      "step": 11415
    },
    {
      "epoch": 5.384252710985384,
      "grad_norm": 0.0021525488700717688,
      "learning_rate": 0.2436116037386042,
      "loss": 0.2253,
      "num_input_tokens_seen": 9898720,
      "step": 11420
    },
    {
      "epoch": 5.386610089580387,
      "grad_norm": 0.002636562567204237,
      "learning_rate": 0.24356557046774852,
      "loss": 0.2387,
      "num_input_tokens_seen": 9902144,
      "step": 11425
    },
    {
      "epoch": 5.388967468175389,
      "grad_norm": 0.001826910418458283,
      "learning_rate": 0.24351952276790606,
      "loss": 0.2275,
      "num_input_tokens_seen": 9906016,
      "step": 11430
    },
    {
      "epoch": 5.391324846770392,
      "grad_norm": 0.002310969168320298,
      "learning_rate": 0.24347346064617797,
      "loss": 0.236,
      "num_input_tokens_seen": 9910160,
      "step": 11435
    },
    {
      "epoch": 5.393682225365394,
      "grad_norm": 0.002616526558995247,
      "learning_rate": 0.24342738410966758,
      "loss": 0.2108,
      "num_input_tokens_seen": 9914992,
      "step": 11440
    },
    {
      "epoch": 5.396039603960396,
      "grad_norm": 0.0016470474656671286,
      "learning_rate": 0.24338129316548046,
      "loss": 0.211,
      "num_input_tokens_seen": 9919824,
      "step": 11445
    },
    {
      "epoch": 5.398396982555399,
      "grad_norm": 0.002278881845995784,
      "learning_rate": 0.24333518782072444,
      "loss": 0.2238,
      "num_input_tokens_seen": 9923184,
      "step": 11450
    },
    {
      "epoch": 5.400754361150401,
      "grad_norm": 0.00238321116194129,
      "learning_rate": 0.24328906808250952,
      "loss": 0.2425,
      "num_input_tokens_seen": 9926976,
      "step": 11455
    },
    {
      "epoch": 5.403111739745403,
      "grad_norm": 0.0017931893235072494,
      "learning_rate": 0.243242933957948,
      "loss": 0.2062,
      "num_input_tokens_seen": 9931280,
      "step": 11460
    },
    {
      "epoch": 5.405469118340405,
      "grad_norm": 0.003933845087885857,
      "learning_rate": 0.24319678545415427,
      "loss": 0.2389,
      "num_input_tokens_seen": 9935376,
      "step": 11465
    },
    {
      "epoch": 5.4078264969354075,
      "grad_norm": 0.002394163515418768,
      "learning_rate": 0.24315062257824507,
      "loss": 0.2208,
      "num_input_tokens_seen": 9939616,
      "step": 11470
    },
    {
      "epoch": 5.41018387553041,
      "grad_norm": 0.0014431895688176155,
      "learning_rate": 0.24310444533733921,
      "loss": 0.1998,
      "num_input_tokens_seen": 9943808,
      "step": 11475
    },
    {
      "epoch": 5.412541254125412,
      "grad_norm": 0.0032012120354920626,
      "learning_rate": 0.2430582537385579,
      "loss": 0.2508,
      "num_input_tokens_seen": 9948352,
      "step": 11480
    },
    {
      "epoch": 5.414898632720415,
      "grad_norm": 0.002077660523355007,
      "learning_rate": 0.2430120477890244,
      "loss": 0.1297,
      "num_input_tokens_seen": 9952704,
      "step": 11485
    },
    {
      "epoch": 5.417256011315417,
      "grad_norm": 0.002724386751651764,
      "learning_rate": 0.24296582749586426,
      "loss": 0.2199,
      "num_input_tokens_seen": 9957440,
      "step": 11490
    },
    {
      "epoch": 5.4196133899104195,
      "grad_norm": 0.0017438612412661314,
      "learning_rate": 0.24291959286620526,
      "loss": 0.1892,
      "num_input_tokens_seen": 9961248,
      "step": 11495
    },
    {
      "epoch": 5.421970768505422,
      "grad_norm": 0.001281241187825799,
      "learning_rate": 0.24287334390717738,
      "loss": 0.2052,
      "num_input_tokens_seen": 9965520,
      "step": 11500
    },
    {
      "epoch": 5.424328147100424,
      "grad_norm": 0.00312383403070271,
      "learning_rate": 0.24282708062591268,
      "loss": 0.2256,
      "num_input_tokens_seen": 9970192,
      "step": 11505
    },
    {
      "epoch": 5.426685525695427,
      "grad_norm": 0.002079016761854291,
      "learning_rate": 0.24278080302954563,
      "loss": 0.2035,
      "num_input_tokens_seen": 9974320,
      "step": 11510
    },
    {
      "epoch": 5.429042904290429,
      "grad_norm": 0.0014294630382210016,
      "learning_rate": 0.24273451112521283,
      "loss": 0.1906,
      "num_input_tokens_seen": 9978752,
      "step": 11515
    },
    {
      "epoch": 5.4314002828854315,
      "grad_norm": 0.0014103431021794677,
      "learning_rate": 0.242688204920053,
      "loss": 0.2584,
      "num_input_tokens_seen": 9982768,
      "step": 11520
    },
    {
      "epoch": 5.433757661480434,
      "grad_norm": 0.001641866285353899,
      "learning_rate": 0.24264188442120715,
      "loss": 0.2205,
      "num_input_tokens_seen": 9987088,
      "step": 11525
    },
    {
      "epoch": 5.436115040075436,
      "grad_norm": 0.0017979093827307224,
      "learning_rate": 0.24259554963581853,
      "loss": 0.2195,
      "num_input_tokens_seen": 9991600,
      "step": 11530
    },
    {
      "epoch": 5.438472418670439,
      "grad_norm": 0.0029355494771152735,
      "learning_rate": 0.24254920057103257,
      "loss": 0.2077,
      "num_input_tokens_seen": 9996464,
      "step": 11535
    },
    {
      "epoch": 5.440829797265441,
      "grad_norm": 0.007207668386399746,
      "learning_rate": 0.24250283723399685,
      "loss": 0.2368,
      "num_input_tokens_seen": 10001024,
      "step": 11540
    },
    {
      "epoch": 5.4431871758604435,
      "grad_norm": 0.001982905901968479,
      "learning_rate": 0.24245645963186108,
      "loss": 0.2347,
      "num_input_tokens_seen": 10005488,
      "step": 11545
    },
    {
      "epoch": 5.445544554455446,
      "grad_norm": 0.002146466402336955,
      "learning_rate": 0.2424100677717774,
      "loss": 0.2111,
      "num_input_tokens_seen": 10009808,
      "step": 11550
    },
    {
      "epoch": 5.4479019330504475,
      "grad_norm": 0.0023647614289075136,
      "learning_rate": 0.24236366166090004,
      "loss": 0.1846,
      "num_input_tokens_seen": 10013664,
      "step": 11555
    },
    {
      "epoch": 5.45025931164545,
      "grad_norm": 0.0027295362669974566,
      "learning_rate": 0.24231724130638527,
      "loss": 0.2002,
      "num_input_tokens_seen": 10018080,
      "step": 11560
    },
    {
      "epoch": 5.452616690240452,
      "grad_norm": 0.003507294226437807,
      "learning_rate": 0.2422708067153917,
      "loss": 0.1185,
      "num_input_tokens_seen": 10022272,
      "step": 11565
    },
    {
      "epoch": 5.454974068835455,
      "grad_norm": 0.00338916527107358,
      "learning_rate": 0.24222435789508026,
      "loss": 0.1491,
      "num_input_tokens_seen": 10027568,
      "step": 11570
    },
    {
      "epoch": 5.457331447430457,
      "grad_norm": 0.001665207790210843,
      "learning_rate": 0.24217789485261387,
      "loss": 0.3225,
      "num_input_tokens_seen": 10032192,
      "step": 11575
    },
    {
      "epoch": 5.4596888260254595,
      "grad_norm": 0.0014746248489245772,
      "learning_rate": 0.2421314175951577,
      "loss": 0.1864,
      "num_input_tokens_seen": 10036048,
      "step": 11580
    },
    {
      "epoch": 5.462046204620462,
      "grad_norm": 0.0025962763465940952,
      "learning_rate": 0.2420849261298791,
      "loss": 0.2641,
      "num_input_tokens_seen": 10041088,
      "step": 11585
    },
    {
      "epoch": 5.464403583215464,
      "grad_norm": 0.0024121664464473724,
      "learning_rate": 0.24203842046394775,
      "loss": 0.2086,
      "num_input_tokens_seen": 10044688,
      "step": 11590
    },
    {
      "epoch": 5.466760961810467,
      "grad_norm": 0.0016893640859052539,
      "learning_rate": 0.24199190060453535,
      "loss": 0.2387,
      "num_input_tokens_seen": 10048928,
      "step": 11595
    },
    {
      "epoch": 5.469118340405469,
      "grad_norm": 0.00309431622736156,
      "learning_rate": 0.2419453665588158,
      "loss": 0.2223,
      "num_input_tokens_seen": 10053056,
      "step": 11600
    },
    {
      "epoch": 5.469118340405469,
      "eval_loss": 0.22025780379772186,
      "eval_runtime": 21.8714,
      "eval_samples_per_second": 43.116,
      "eval_steps_per_second": 21.581,
      "num_input_tokens_seen": 10053056,
      "step": 11600
    },
    {
      "epoch": 5.4714757190004715,
      "grad_norm": 0.002789038699120283,
      "learning_rate": 0.24189881833396523,
      "loss": 0.2049,
      "num_input_tokens_seen": 10057584,
      "step": 11605
    },
    {
      "epoch": 5.473833097595474,
      "grad_norm": 0.003727812087163329,
      "learning_rate": 0.24185225593716203,
      "loss": 0.2558,
      "num_input_tokens_seen": 10062016,
      "step": 11610
    },
    {
      "epoch": 5.476190476190476,
      "grad_norm": 0.002059090416878462,
      "learning_rate": 0.2418056793755867,
      "loss": 0.2088,
      "num_input_tokens_seen": 10067376,
      "step": 11615
    },
    {
      "epoch": 5.478547854785479,
      "grad_norm": 0.005041768774390221,
      "learning_rate": 0.24175908865642187,
      "loss": 0.2837,
      "num_input_tokens_seen": 10071568,
      "step": 11620
    },
    {
      "epoch": 5.480905233380481,
      "grad_norm": 0.00348680280148983,
      "learning_rate": 0.24171248378685248,
      "loss": 0.2069,
      "num_input_tokens_seen": 10074976,
      "step": 11625
    },
    {
      "epoch": 5.4832626119754835,
      "grad_norm": 0.005859711207449436,
      "learning_rate": 0.24166586477406554,
      "loss": 0.2314,
      "num_input_tokens_seen": 10079776,
      "step": 11630
    },
    {
      "epoch": 5.485619990570486,
      "grad_norm": 0.004862848203629255,
      "learning_rate": 0.24161923162525034,
      "loss": 0.1996,
      "num_input_tokens_seen": 10084416,
      "step": 11635
    },
    {
      "epoch": 5.487977369165488,
      "grad_norm": 0.011332573369145393,
      "learning_rate": 0.2415725843475982,
      "loss": 0.2579,
      "num_input_tokens_seen": 10089536,
      "step": 11640
    },
    {
      "epoch": 5.490334747760491,
      "grad_norm": 0.0020823704544454813,
      "learning_rate": 0.24152592294830286,
      "loss": 0.2346,
      "num_input_tokens_seen": 10094112,
      "step": 11645
    },
    {
      "epoch": 5.492692126355493,
      "grad_norm": 0.001893607317470014,
      "learning_rate": 0.24147924743455995,
      "loss": 0.2308,
      "num_input_tokens_seen": 10098336,
      "step": 11650
    },
    {
      "epoch": 5.4950495049504955,
      "grad_norm": 0.0020629488863050938,
      "learning_rate": 0.24143255781356754,
      "loss": 0.2235,
      "num_input_tokens_seen": 10102144,
      "step": 11655
    },
    {
      "epoch": 5.497406883545497,
      "grad_norm": 0.004417514428496361,
      "learning_rate": 0.24138585409252566,
      "loss": 0.2239,
      "num_input_tokens_seen": 10106240,
      "step": 11660
    },
    {
      "epoch": 5.499764262140499,
      "grad_norm": 0.00349277607165277,
      "learning_rate": 0.24133913627863662,
      "loss": 0.2486,
      "num_input_tokens_seen": 10111296,
      "step": 11665
    },
    {
      "epoch": 5.502121640735502,
      "grad_norm": 0.0022747863549739122,
      "learning_rate": 0.241292404379105,
      "loss": 0.2318,
      "num_input_tokens_seen": 10115280,
      "step": 11670
    },
    {
      "epoch": 5.504479019330504,
      "grad_norm": 0.003932808060199022,
      "learning_rate": 0.24124565840113735,
      "loss": 0.2376,
      "num_input_tokens_seen": 10118896,
      "step": 11675
    },
    {
      "epoch": 5.506836397925507,
      "grad_norm": 0.00474990950897336,
      "learning_rate": 0.2411988983519425,
      "loss": 0.2583,
      "num_input_tokens_seen": 10123360,
      "step": 11680
    },
    {
      "epoch": 5.509193776520509,
      "grad_norm": 0.003137154970318079,
      "learning_rate": 0.24115212423873145,
      "loss": 0.2141,
      "num_input_tokens_seen": 10127584,
      "step": 11685
    },
    {
      "epoch": 5.511551155115511,
      "grad_norm": 0.002722302684560418,
      "learning_rate": 0.24110533606871737,
      "loss": 0.2196,
      "num_input_tokens_seen": 10132480,
      "step": 11690
    },
    {
      "epoch": 5.513908533710514,
      "grad_norm": 0.004700688645243645,
      "learning_rate": 0.24105853384911552,
      "loss": 0.2324,
      "num_input_tokens_seen": 10137040,
      "step": 11695
    },
    {
      "epoch": 5.516265912305516,
      "grad_norm": 0.003957216162234545,
      "learning_rate": 0.24101171758714346,
      "loss": 0.2514,
      "num_input_tokens_seen": 10141984,
      "step": 11700
    },
    {
      "epoch": 5.518623290900519,
      "grad_norm": 0.004483755677938461,
      "learning_rate": 0.24096488729002086,
      "loss": 0.246,
      "num_input_tokens_seen": 10146880,
      "step": 11705
    },
    {
      "epoch": 5.520980669495521,
      "grad_norm": 0.0032473516184836626,
      "learning_rate": 0.24091804296496946,
      "loss": 0.2421,
      "num_input_tokens_seen": 10150880,
      "step": 11710
    },
    {
      "epoch": 5.523338048090523,
      "grad_norm": 0.0028087906539440155,
      "learning_rate": 0.2408711846192133,
      "loss": 0.2552,
      "num_input_tokens_seen": 10155184,
      "step": 11715
    },
    {
      "epoch": 5.525695426685526,
      "grad_norm": 0.003793142270296812,
      "learning_rate": 0.24082431225997855,
      "loss": 0.1796,
      "num_input_tokens_seen": 10159040,
      "step": 11720
    },
    {
      "epoch": 5.528052805280528,
      "grad_norm": 0.002156143309548497,
      "learning_rate": 0.24077742589449344,
      "loss": 0.2584,
      "num_input_tokens_seen": 10162880,
      "step": 11725
    },
    {
      "epoch": 5.530410183875531,
      "grad_norm": 0.002776357578113675,
      "learning_rate": 0.24073052552998844,
      "loss": 0.1528,
      "num_input_tokens_seen": 10167664,
      "step": 11730
    },
    {
      "epoch": 5.532767562470533,
      "grad_norm": 0.0022845561616122723,
      "learning_rate": 0.2406836111736963,
      "loss": 0.239,
      "num_input_tokens_seen": 10171808,
      "step": 11735
    },
    {
      "epoch": 5.535124941065535,
      "grad_norm": 0.004417682997882366,
      "learning_rate": 0.2406366828328517,
      "loss": 0.1832,
      "num_input_tokens_seen": 10176064,
      "step": 11740
    },
    {
      "epoch": 5.537482319660538,
      "grad_norm": 0.003715238533914089,
      "learning_rate": 0.2405897405146915,
      "loss": 0.2835,
      "num_input_tokens_seen": 10181328,
      "step": 11745
    },
    {
      "epoch": 5.539839698255539,
      "grad_norm": 0.0050498549826443195,
      "learning_rate": 0.240542784226455,
      "loss": 0.237,
      "num_input_tokens_seen": 10185088,
      "step": 11750
    },
    {
      "epoch": 5.542197076850542,
      "grad_norm": 0.0045802053064107895,
      "learning_rate": 0.24049581397538328,
      "loss": 0.2427,
      "num_input_tokens_seen": 10189440,
      "step": 11755
    },
    {
      "epoch": 5.544554455445544,
      "grad_norm": 0.003715166822075844,
      "learning_rate": 0.24044882976871984,
      "loss": 0.2302,
      "num_input_tokens_seen": 10194096,
      "step": 11760
    },
    {
      "epoch": 5.5469118340405466,
      "grad_norm": 0.003746785456314683,
      "learning_rate": 0.2404018316137102,
      "loss": 0.2324,
      "num_input_tokens_seen": 10198240,
      "step": 11765
    },
    {
      "epoch": 5.549269212635549,
      "grad_norm": 0.0029834131710231304,
      "learning_rate": 0.24035481951760204,
      "loss": 0.1971,
      "num_input_tokens_seen": 10202640,
      "step": 11770
    },
    {
      "epoch": 5.551626591230551,
      "grad_norm": 0.0029536550864577293,
      "learning_rate": 0.2403077934876452,
      "loss": 0.1963,
      "num_input_tokens_seen": 10206672,
      "step": 11775
    },
    {
      "epoch": 5.553983969825554,
      "grad_norm": 0.0037315948866307735,
      "learning_rate": 0.2402607535310918,
      "loss": 0.209,
      "num_input_tokens_seen": 10210720,
      "step": 11780
    },
    {
      "epoch": 5.556341348420556,
      "grad_norm": 0.002326107118278742,
      "learning_rate": 0.2402136996551959,
      "loss": 0.2134,
      "num_input_tokens_seen": 10215152,
      "step": 11785
    },
    {
      "epoch": 5.558698727015559,
      "grad_norm": 0.0046619838103652,
      "learning_rate": 0.24016663186721376,
      "loss": 0.2405,
      "num_input_tokens_seen": 10219536,
      "step": 11790
    },
    {
      "epoch": 5.561056105610561,
      "grad_norm": 0.0037690310273319483,
      "learning_rate": 0.24011955017440395,
      "loss": 0.1996,
      "num_input_tokens_seen": 10225104,
      "step": 11795
    },
    {
      "epoch": 5.563413484205563,
      "grad_norm": 0.004508114419877529,
      "learning_rate": 0.24007245458402696,
      "loss": 0.2419,
      "num_input_tokens_seen": 10229152,
      "step": 11800
    },
    {
      "epoch": 5.563413484205563,
      "eval_loss": 0.22546260058879852,
      "eval_runtime": 21.8944,
      "eval_samples_per_second": 43.07,
      "eval_steps_per_second": 21.558,
      "num_input_tokens_seen": 10229152,
      "step": 11800
    },
    {
      "epoch": 5.565770862800566,
      "grad_norm": 0.003787621157243848,
      "learning_rate": 0.2400253451033456,
      "loss": 0.241,
      "num_input_tokens_seen": 10233136,
      "step": 11805
    },
    {
      "epoch": 5.568128241395568,
      "grad_norm": 0.003572129411622882,
      "learning_rate": 0.23997822173962463,
      "loss": 0.2237,
      "num_input_tokens_seen": 10237696,
      "step": 11810
    },
    {
      "epoch": 5.570485619990571,
      "grad_norm": 0.0032877796329557896,
      "learning_rate": 0.23993108450013118,
      "loss": 0.2348,
      "num_input_tokens_seen": 10241504,
      "step": 11815
    },
    {
      "epoch": 5.572842998585573,
      "grad_norm": 0.0030637597665190697,
      "learning_rate": 0.2398839333921343,
      "loss": 0.2539,
      "num_input_tokens_seen": 10246064,
      "step": 11820
    },
    {
      "epoch": 5.575200377180575,
      "grad_norm": 0.0040597496554255486,
      "learning_rate": 0.23983676842290536,
      "loss": 0.2322,
      "num_input_tokens_seen": 10250736,
      "step": 11825
    },
    {
      "epoch": 5.577557755775578,
      "grad_norm": 0.005836512427777052,
      "learning_rate": 0.2397895895997178,
      "loss": 0.2508,
      "num_input_tokens_seen": 10255424,
      "step": 11830
    },
    {
      "epoch": 5.57991513437058,
      "grad_norm": 0.003832130925729871,
      "learning_rate": 0.23974239692984714,
      "loss": 0.23,
      "num_input_tokens_seen": 10259696,
      "step": 11835
    },
    {
      "epoch": 5.582272512965583,
      "grad_norm": 0.002438921947032213,
      "learning_rate": 0.2396951904205711,
      "loss": 0.2204,
      "num_input_tokens_seen": 10263712,
      "step": 11840
    },
    {
      "epoch": 5.584629891560585,
      "grad_norm": 0.004780299961566925,
      "learning_rate": 0.23964797007916952,
      "loss": 0.2383,
      "num_input_tokens_seen": 10269376,
      "step": 11845
    },
    {
      "epoch": 5.586987270155587,
      "grad_norm": 0.005251883529126644,
      "learning_rate": 0.23960073591292436,
      "loss": 0.2254,
      "num_input_tokens_seen": 10273488,
      "step": 11850
    },
    {
      "epoch": 5.58934464875059,
      "grad_norm": 0.004347655922174454,
      "learning_rate": 0.2395534879291197,
      "loss": 0.2675,
      "num_input_tokens_seen": 10277248,
      "step": 11855
    },
    {
      "epoch": 5.591702027345592,
      "grad_norm": 0.006654328666627407,
      "learning_rate": 0.23950622613504186,
      "loss": 0.2357,
      "num_input_tokens_seen": 10282784,
      "step": 11860
    },
    {
      "epoch": 5.594059405940594,
      "grad_norm": 0.00681394012644887,
      "learning_rate": 0.2394589505379791,
      "loss": 0.2429,
      "num_input_tokens_seen": 10287856,
      "step": 11865
    },
    {
      "epoch": 5.596416784535596,
      "grad_norm": 0.003026197897270322,
      "learning_rate": 0.23941166114522197,
      "loss": 0.2344,
      "num_input_tokens_seen": 10292128,
      "step": 11870
    },
    {
      "epoch": 5.5987741631305985,
      "grad_norm": 0.0019126959377899766,
      "learning_rate": 0.23936435796406308,
      "loss": 0.2296,
      "num_input_tokens_seen": 10296912,
      "step": 11875
    },
    {
      "epoch": 5.601131541725601,
      "grad_norm": 0.0026859338395297527,
      "learning_rate": 0.23931704100179715,
      "loss": 0.1882,
      "num_input_tokens_seen": 10302112,
      "step": 11880
    },
    {
      "epoch": 5.603488920320603,
      "grad_norm": 0.0031750337220728397,
      "learning_rate": 0.2392697102657211,
      "loss": 0.2381,
      "num_input_tokens_seen": 10306352,
      "step": 11885
    },
    {
      "epoch": 5.605846298915606,
      "grad_norm": 0.004438304342329502,
      "learning_rate": 0.23922236576313388,
      "loss": 0.2782,
      "num_input_tokens_seen": 10310464,
      "step": 11890
    },
    {
      "epoch": 5.608203677510608,
      "grad_norm": 0.0038150011096149683,
      "learning_rate": 0.2391750075013366,
      "loss": 0.2299,
      "num_input_tokens_seen": 10315120,
      "step": 11895
    },
    {
      "epoch": 5.6105610561056105,
      "grad_norm": 0.006884389091283083,
      "learning_rate": 0.2391276354876326,
      "loss": 0.2324,
      "num_input_tokens_seen": 10319344,
      "step": 11900
    },
    {
      "epoch": 5.612918434700613,
      "grad_norm": 0.002862598979845643,
      "learning_rate": 0.23908024972932707,
      "loss": 0.2308,
      "num_input_tokens_seen": 10323776,
      "step": 11905
    },
    {
      "epoch": 5.615275813295615,
      "grad_norm": 0.0035583414137363434,
      "learning_rate": 0.2390328502337276,
      "loss": 0.2251,
      "num_input_tokens_seen": 10327552,
      "step": 11910
    },
    {
      "epoch": 5.617633191890618,
      "grad_norm": 0.0027046960312873125,
      "learning_rate": 0.23898543700814376,
      "loss": 0.2342,
      "num_input_tokens_seen": 10331648,
      "step": 11915
    },
    {
      "epoch": 5.61999057048562,
      "grad_norm": 0.0032091052271425724,
      "learning_rate": 0.2389380100598873,
      "loss": 0.229,
      "num_input_tokens_seen": 10336112,
      "step": 11920
    },
    {
      "epoch": 5.6223479490806225,
      "grad_norm": 0.005163058638572693,
      "learning_rate": 0.23889056939627207,
      "loss": 0.2332,
      "num_input_tokens_seen": 10340640,
      "step": 11925
    },
    {
      "epoch": 5.624705327675625,
      "grad_norm": 0.003621079493314028,
      "learning_rate": 0.23884311502461386,
      "loss": 0.2067,
      "num_input_tokens_seen": 10344896,
      "step": 11930
    },
    {
      "epoch": 5.627062706270627,
      "grad_norm": 0.002058083424344659,
      "learning_rate": 0.23879564695223088,
      "loss": 0.1926,
      "num_input_tokens_seen": 10349136,
      "step": 11935
    },
    {
      "epoch": 5.62942008486563,
      "grad_norm": 0.005798717495054007,
      "learning_rate": 0.23874816518644332,
      "loss": 0.2479,
      "num_input_tokens_seen": 10353360,
      "step": 11940
    },
    {
      "epoch": 5.631777463460632,
      "grad_norm": 0.005166816525161266,
      "learning_rate": 0.23870066973457335,
      "loss": 0.2594,
      "num_input_tokens_seen": 10358000,
      "step": 11945
    },
    {
      "epoch": 5.634134842055634,
      "grad_norm": 0.010093100368976593,
      "learning_rate": 0.23865316060394545,
      "loss": 0.2614,
      "num_input_tokens_seen": 10361872,
      "step": 11950
    },
    {
      "epoch": 5.636492220650636,
      "grad_norm": 0.010243802331387997,
      "learning_rate": 0.2386056378018861,
      "loss": 0.2383,
      "num_input_tokens_seen": 10366416,
      "step": 11955
    },
    {
      "epoch": 5.6388495992456384,
      "grad_norm": 0.004981286823749542,
      "learning_rate": 0.2385581013357239,
      "loss": 0.2217,
      "num_input_tokens_seen": 10370672,
      "step": 11960
    },
    {
      "epoch": 5.641206977840641,
      "grad_norm": 0.007122271228581667,
      "learning_rate": 0.23851055121278958,
      "loss": 0.2333,
      "num_input_tokens_seen": 10375088,
      "step": 11965
    },
    {
      "epoch": 5.643564356435643,
      "grad_norm": 0.00337813189253211,
      "learning_rate": 0.23846298744041594,
      "loss": 0.2159,
      "num_input_tokens_seen": 10379712,
      "step": 11970
    },
    {
      "epoch": 5.645921735030646,
      "grad_norm": 0.002766592428088188,
      "learning_rate": 0.23841541002593802,
      "loss": 0.2368,
      "num_input_tokens_seen": 10383392,
      "step": 11975
    },
    {
      "epoch": 5.648279113625648,
      "grad_norm": 0.002178257331252098,
      "learning_rate": 0.23836781897669276,
      "loss": 0.1975,
      "num_input_tokens_seen": 10387328,
      "step": 11980
    },
    {
      "epoch": 5.6506364922206505,
      "grad_norm": 0.002414044924080372,
      "learning_rate": 0.23832021430001926,
      "loss": 0.2219,
      "num_input_tokens_seen": 10391760,
      "step": 11985
    },
    {
      "epoch": 5.652993870815653,
      "grad_norm": 0.004736360628157854,
      "learning_rate": 0.2382725960032588,
      "loss": 0.2652,
      "num_input_tokens_seen": 10396480,
      "step": 11990
    },
    {
      "epoch": 5.655351249410655,
      "grad_norm": 0.0019061478087678552,
      "learning_rate": 0.23822496409375482,
      "loss": 0.2628,
      "num_input_tokens_seen": 10400448,
      "step": 11995
    },
    {
      "epoch": 5.657708628005658,
      "grad_norm": 0.002871745266020298,
      "learning_rate": 0.2381773185788526,
      "loss": 0.2415,
      "num_input_tokens_seen": 10404384,
      "step": 12000
    },
    {
      "epoch": 5.657708628005658,
      "eval_loss": 0.22231729328632355,
      "eval_runtime": 21.9168,
      "eval_samples_per_second": 43.026,
      "eval_steps_per_second": 21.536,
      "num_input_tokens_seen": 10404384,
      "step": 12000
    },
    {
      "epoch": 5.66006600660066,
      "grad_norm": 0.0013145023258402944,
      "learning_rate": 0.2381296594658998,
      "loss": 0.2141,
      "num_input_tokens_seen": 10408064,
      "step": 12005
    },
    {
      "epoch": 5.6624233851956625,
      "grad_norm": 0.002741508651524782,
      "learning_rate": 0.238081986762246,
      "loss": 0.2301,
      "num_input_tokens_seen": 10412800,
      "step": 12010
    },
    {
      "epoch": 5.664780763790665,
      "grad_norm": 0.002253303537145257,
      "learning_rate": 0.23803430047524293,
      "loss": 0.2489,
      "num_input_tokens_seen": 10416640,
      "step": 12015
    },
    {
      "epoch": 5.667138142385667,
      "grad_norm": 0.0020363039802759886,
      "learning_rate": 0.23798660061224441,
      "loss": 0.2351,
      "num_input_tokens_seen": 10420736,
      "step": 12020
    },
    {
      "epoch": 5.66949552098067,
      "grad_norm": 0.0017612709198147058,
      "learning_rate": 0.23793888718060632,
      "loss": 0.2304,
      "num_input_tokens_seen": 10424976,
      "step": 12025
    },
    {
      "epoch": 5.671852899575672,
      "grad_norm": 0.0014409400755539536,
      "learning_rate": 0.23789116018768675,
      "loss": 0.2134,
      "num_input_tokens_seen": 10428768,
      "step": 12030
    },
    {
      "epoch": 5.6742102781706745,
      "grad_norm": 0.0018772990442812443,
      "learning_rate": 0.2378434196408458,
      "loss": 0.2209,
      "num_input_tokens_seen": 10432928,
      "step": 12035
    },
    {
      "epoch": 5.676567656765677,
      "grad_norm": 0.0033909196499735117,
      "learning_rate": 0.23779566554744563,
      "loss": 0.251,
      "num_input_tokens_seen": 10437168,
      "step": 12040
    },
    {
      "epoch": 5.678925035360679,
      "grad_norm": 0.0013322409940883517,
      "learning_rate": 0.23774789791485051,
      "loss": 0.2322,
      "num_input_tokens_seen": 10441632,
      "step": 12045
    },
    {
      "epoch": 5.681282413955682,
      "grad_norm": 0.00291747716255486,
      "learning_rate": 0.2377001167504268,
      "loss": 0.2165,
      "num_input_tokens_seen": 10445216,
      "step": 12050
    },
    {
      "epoch": 5.683639792550684,
      "grad_norm": 0.0015680603682994843,
      "learning_rate": 0.23765232206154302,
      "loss": 0.2202,
      "num_input_tokens_seen": 10450032,
      "step": 12055
    },
    {
      "epoch": 5.6859971711456865,
      "grad_norm": 0.0031468949746340513,
      "learning_rate": 0.23760451385556966,
      "loss": 0.2188,
      "num_input_tokens_seen": 10453856,
      "step": 12060
    },
    {
      "epoch": 5.688354549740688,
      "grad_norm": 0.0019345019245520234,
      "learning_rate": 0.23755669213987932,
      "loss": 0.2391,
      "num_input_tokens_seen": 10458832,
      "step": 12065
    },
    {
      "epoch": 5.69071192833569,
      "grad_norm": 0.0028907679952681065,
      "learning_rate": 0.23750885692184676,
      "loss": 0.2027,
      "num_input_tokens_seen": 10462768,
      "step": 12070
    },
    {
      "epoch": 5.693069306930693,
      "grad_norm": 0.0021425168961286545,
      "learning_rate": 0.23746100820884875,
      "loss": 0.2507,
      "num_input_tokens_seen": 10467360,
      "step": 12075
    },
    {
      "epoch": 5.695426685525695,
      "grad_norm": 0.003289925865828991,
      "learning_rate": 0.23741314600826421,
      "loss": 0.2362,
      "num_input_tokens_seen": 10471920,
      "step": 12080
    },
    {
      "epoch": 5.697784064120698,
      "grad_norm": 0.002098443917930126,
      "learning_rate": 0.23736527032747406,
      "loss": 0.2016,
      "num_input_tokens_seen": 10476496,
      "step": 12085
    },
    {
      "epoch": 5.7001414427157,
      "grad_norm": 0.0027826421428471804,
      "learning_rate": 0.23731738117386128,
      "loss": 0.2337,
      "num_input_tokens_seen": 10480528,
      "step": 12090
    },
    {
      "epoch": 5.702498821310702,
      "grad_norm": 0.002512238686904311,
      "learning_rate": 0.237269478554811,
      "loss": 0.2327,
      "num_input_tokens_seen": 10485536,
      "step": 12095
    },
    {
      "epoch": 5.704856199905705,
      "grad_norm": 0.002275969833135605,
      "learning_rate": 0.23722156247771053,
      "loss": 0.2163,
      "num_input_tokens_seen": 10490032,
      "step": 12100
    },
    {
      "epoch": 5.707213578500707,
      "grad_norm": 0.0034169454593211412,
      "learning_rate": 0.23717363294994895,
      "loss": 0.2066,
      "num_input_tokens_seen": 10493920,
      "step": 12105
    },
    {
      "epoch": 5.70957095709571,
      "grad_norm": 0.004329087678343058,
      "learning_rate": 0.2371256899789177,
      "loss": 0.2581,
      "num_input_tokens_seen": 10497408,
      "step": 12110
    },
    {
      "epoch": 5.711928335690712,
      "grad_norm": 0.0017217560671269894,
      "learning_rate": 0.23707773357201017,
      "loss": 0.2446,
      "num_input_tokens_seen": 10501152,
      "step": 12115
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.004449272062629461,
      "learning_rate": 0.2370297637366218,
      "loss": 0.2367,
      "num_input_tokens_seen": 10505360,
      "step": 12120
    },
    {
      "epoch": 5.716643092880717,
      "grad_norm": 0.002114554401487112,
      "learning_rate": 0.23698178048015026,
      "loss": 0.219,
      "num_input_tokens_seen": 10509600,
      "step": 12125
    },
    {
      "epoch": 5.719000471475719,
      "grad_norm": 0.0022752247750759125,
      "learning_rate": 0.236933783809995,
      "loss": 0.1937,
      "num_input_tokens_seen": 10513616,
      "step": 12130
    },
    {
      "epoch": 5.721357850070722,
      "grad_norm": 0.0014830430736765265,
      "learning_rate": 0.23688577373355785,
      "loss": 0.2348,
      "num_input_tokens_seen": 10517728,
      "step": 12135
    },
    {
      "epoch": 5.723715228665724,
      "grad_norm": 0.011997158639132977,
      "learning_rate": 0.23683775025824247,
      "loss": 0.2596,
      "num_input_tokens_seen": 10522352,
      "step": 12140
    },
    {
      "epoch": 5.726072607260726,
      "grad_norm": 0.002105731749907136,
      "learning_rate": 0.2367897133914548,
      "loss": 0.1941,
      "num_input_tokens_seen": 10526368,
      "step": 12145
    },
    {
      "epoch": 5.728429985855728,
      "grad_norm": 0.0030811415053904057,
      "learning_rate": 0.2367416631406026,
      "loss": 0.2477,
      "num_input_tokens_seen": 10531104,
      "step": 12150
    },
    {
      "epoch": 5.73078736445073,
      "grad_norm": 0.002786726225167513,
      "learning_rate": 0.23669359951309588,
      "loss": 0.2327,
      "num_input_tokens_seen": 10534800,
      "step": 12155
    },
    {
      "epoch": 5.733144743045733,
      "grad_norm": 0.003152052639052272,
      "learning_rate": 0.23664552251634666,
      "loss": 0.2348,
      "num_input_tokens_seen": 10539056,
      "step": 12160
    },
    {
      "epoch": 5.735502121640735,
      "grad_norm": 0.002401920733973384,
      "learning_rate": 0.23659743215776907,
      "loss": 0.2309,
      "num_input_tokens_seen": 10542704,
      "step": 12165
    },
    {
      "epoch": 5.7378595002357375,
      "grad_norm": 0.0027268915437161922,
      "learning_rate": 0.23654932844477908,
      "loss": 0.2289,
      "num_input_tokens_seen": 10546832,
      "step": 12170
    },
    {
      "epoch": 5.74021687883074,
      "grad_norm": 0.004867551848292351,
      "learning_rate": 0.23650121138479507,
      "loss": 0.2331,
      "num_input_tokens_seen": 10551328,
      "step": 12175
    },
    {
      "epoch": 5.742574257425742,
      "grad_norm": 0.0012184649240225554,
      "learning_rate": 0.23645308098523724,
      "loss": 0.2315,
      "num_input_tokens_seen": 10555056,
      "step": 12180
    },
    {
      "epoch": 5.744931636020745,
      "grad_norm": 0.002394026145339012,
      "learning_rate": 0.23640493725352785,
      "loss": 0.2078,
      "num_input_tokens_seen": 10559680,
      "step": 12185
    },
    {
      "epoch": 5.747289014615747,
      "grad_norm": 0.0033655078150331974,
      "learning_rate": 0.2363567801970913,
      "loss": 0.2248,
      "num_input_tokens_seen": 10564480,
      "step": 12190
    },
    {
      "epoch": 5.7496463932107496,
      "grad_norm": 0.0030171151738613844,
      "learning_rate": 0.236308609823354,
      "loss": 0.2416,
      "num_input_tokens_seen": 10568672,
      "step": 12195
    },
    {
      "epoch": 5.752003771805752,
      "grad_norm": 0.002261737361550331,
      "learning_rate": 0.23626042613974452,
      "loss": 0.2475,
      "num_input_tokens_seen": 10573872,
      "step": 12200
    },
    {
      "epoch": 5.752003771805752,
      "eval_loss": 0.22059963643550873,
      "eval_runtime": 21.9231,
      "eval_samples_per_second": 43.014,
      "eval_steps_per_second": 21.53,
      "num_input_tokens_seen": 10573872,
      "step": 12200
    },
    {
      "epoch": 5.754361150400754,
      "grad_norm": 0.001770464819855988,
      "learning_rate": 0.23621222915369325,
      "loss": 0.2076,
      "num_input_tokens_seen": 10578272,
      "step": 12205
    },
    {
      "epoch": 5.756718528995757,
      "grad_norm": 0.0026349718682467937,
      "learning_rate": 0.23616401887263283,
      "loss": 0.2184,
      "num_input_tokens_seen": 10582000,
      "step": 12210
    },
    {
      "epoch": 5.759075907590759,
      "grad_norm": 0.0030838397797197104,
      "learning_rate": 0.23611579530399793,
      "loss": 0.2469,
      "num_input_tokens_seen": 10586720,
      "step": 12215
    },
    {
      "epoch": 5.761433286185762,
      "grad_norm": 0.0024073028471320868,
      "learning_rate": 0.23606755845522517,
      "loss": 0.2418,
      "num_input_tokens_seen": 10590960,
      "step": 12220
    },
    {
      "epoch": 5.763790664780764,
      "grad_norm": 0.0019175256602466106,
      "learning_rate": 0.23601930833375329,
      "loss": 0.2355,
      "num_input_tokens_seen": 10595568,
      "step": 12225
    },
    {
      "epoch": 5.766148043375766,
      "grad_norm": 0.0023678780999034643,
      "learning_rate": 0.23597104494702312,
      "loss": 0.2143,
      "num_input_tokens_seen": 10599392,
      "step": 12230
    },
    {
      "epoch": 5.768505421970769,
      "grad_norm": 0.003325186436995864,
      "learning_rate": 0.23592276830247744,
      "loss": 0.2857,
      "num_input_tokens_seen": 10603824,
      "step": 12235
    },
    {
      "epoch": 5.770862800565771,
      "grad_norm": 0.0029391483403742313,
      "learning_rate": 0.2358744784075611,
      "loss": 0.2459,
      "num_input_tokens_seen": 10608064,
      "step": 12240
    },
    {
      "epoch": 5.773220179160774,
      "grad_norm": 0.0034665463026612997,
      "learning_rate": 0.235826175269721,
      "loss": 0.2294,
      "num_input_tokens_seen": 10613536,
      "step": 12245
    },
    {
      "epoch": 5.775577557755776,
      "grad_norm": 0.001762371277436614,
      "learning_rate": 0.23577785889640612,
      "loss": 0.2214,
      "num_input_tokens_seen": 10618176,
      "step": 12250
    },
    {
      "epoch": 5.777934936350778,
      "grad_norm": 0.0037857822608202696,
      "learning_rate": 0.23572952929506744,
      "loss": 0.2062,
      "num_input_tokens_seen": 10623536,
      "step": 12255
    },
    {
      "epoch": 5.780292314945781,
      "grad_norm": 0.001770454808138311,
      "learning_rate": 0.23568118647315803,
      "loss": 0.2368,
      "num_input_tokens_seen": 10628336,
      "step": 12260
    },
    {
      "epoch": 5.782649693540782,
      "grad_norm": 0.0031121803913265467,
      "learning_rate": 0.23563283043813296,
      "loss": 0.2299,
      "num_input_tokens_seen": 10632336,
      "step": 12265
    },
    {
      "epoch": 5.785007072135785,
      "grad_norm": 0.002708704676479101,
      "learning_rate": 0.23558446119744922,
      "loss": 0.1791,
      "num_input_tokens_seen": 10636960,
      "step": 12270
    },
    {
      "epoch": 5.787364450730787,
      "grad_norm": 0.0024953093379735947,
      "learning_rate": 0.23553607875856608,
      "loss": 0.2755,
      "num_input_tokens_seen": 10640960,
      "step": 12275
    },
    {
      "epoch": 5.7897218293257895,
      "grad_norm": 0.0020556319504976273,
      "learning_rate": 0.2354876831289447,
      "loss": 0.2446,
      "num_input_tokens_seen": 10644544,
      "step": 12280
    },
    {
      "epoch": 5.792079207920792,
      "grad_norm": 0.003997668623924255,
      "learning_rate": 0.23543927431604827,
      "loss": 0.2359,
      "num_input_tokens_seen": 10648272,
      "step": 12285
    },
    {
      "epoch": 5.794436586515794,
      "grad_norm": 0.0042749145068228245,
      "learning_rate": 0.23539085232734203,
      "loss": 0.2499,
      "num_input_tokens_seen": 10651920,
      "step": 12290
    },
    {
      "epoch": 5.796793965110797,
      "grad_norm": 0.0029182955622673035,
      "learning_rate": 0.2353424171702933,
      "loss": 0.2514,
      "num_input_tokens_seen": 10656288,
      "step": 12295
    },
    {
      "epoch": 5.799151343705799,
      "grad_norm": 0.0016234625363722444,
      "learning_rate": 0.23529396885237133,
      "loss": 0.2362,
      "num_input_tokens_seen": 10660304,
      "step": 12300
    },
    {
      "epoch": 5.8015087223008015,
      "grad_norm": 0.002593948505818844,
      "learning_rate": 0.2352455073810475,
      "loss": 0.2207,
      "num_input_tokens_seen": 10664608,
      "step": 12305
    },
    {
      "epoch": 5.803866100895804,
      "grad_norm": 0.002961423946544528,
      "learning_rate": 0.23519703276379517,
      "loss": 0.2047,
      "num_input_tokens_seen": 10670064,
      "step": 12310
    },
    {
      "epoch": 5.806223479490806,
      "grad_norm": 0.001962879905477166,
      "learning_rate": 0.2351485450080897,
      "loss": 0.2564,
      "num_input_tokens_seen": 10674928,
      "step": 12315
    },
    {
      "epoch": 5.808580858085809,
      "grad_norm": 0.0020464598201215267,
      "learning_rate": 0.2351000441214086,
      "loss": 0.228,
      "num_input_tokens_seen": 10679264,
      "step": 12320
    },
    {
      "epoch": 5.810938236680811,
      "grad_norm": 0.002720191376283765,
      "learning_rate": 0.23505153011123125,
      "loss": 0.2269,
      "num_input_tokens_seen": 10683376,
      "step": 12325
    },
    {
      "epoch": 5.8132956152758135,
      "grad_norm": 0.0031840079464018345,
      "learning_rate": 0.23500300298503912,
      "loss": 0.1967,
      "num_input_tokens_seen": 10686928,
      "step": 12330
    },
    {
      "epoch": 5.815652993870816,
      "grad_norm": 0.0030439544934779406,
      "learning_rate": 0.23495446275031576,
      "loss": 0.2141,
      "num_input_tokens_seen": 10691632,
      "step": 12335
    },
    {
      "epoch": 5.818010372465818,
      "grad_norm": 0.0018445487366989255,
      "learning_rate": 0.2349059094145466,
      "loss": 0.2451,
      "num_input_tokens_seen": 10696096,
      "step": 12340
    },
    {
      "epoch": 5.820367751060821,
      "grad_norm": 0.003526841988787055,
      "learning_rate": 0.2348573429852192,
      "loss": 0.2664,
      "num_input_tokens_seen": 10700544,
      "step": 12345
    },
    {
      "epoch": 5.822725129655822,
      "grad_norm": 0.001826312392950058,
      "learning_rate": 0.23480876346982313,
      "loss": 0.2402,
      "num_input_tokens_seen": 10704944,
      "step": 12350
    },
    {
      "epoch": 5.825082508250825,
      "grad_norm": 0.00144754140637815,
      "learning_rate": 0.23476017087585,
      "loss": 0.2293,
      "num_input_tokens_seen": 10709632,
      "step": 12355
    },
    {
      "epoch": 5.827439886845827,
      "grad_norm": 0.002339455299079418,
      "learning_rate": 0.23471156521079334,
      "loss": 0.226,
      "num_input_tokens_seen": 10714336,
      "step": 12360
    },
    {
      "epoch": 5.829797265440829,
      "grad_norm": 0.002207937417551875,
      "learning_rate": 0.23466294648214875,
      "loss": 0.1906,
      "num_input_tokens_seen": 10719056,
      "step": 12365
    },
    {
      "epoch": 5.832154644035832,
      "grad_norm": 0.0034571720752865076,
      "learning_rate": 0.2346143146974139,
      "loss": 0.2498,
      "num_input_tokens_seen": 10722736,
      "step": 12370
    },
    {
      "epoch": 5.834512022630834,
      "grad_norm": 0.0023815277963876724,
      "learning_rate": 0.23456566986408836,
      "loss": 0.2589,
      "num_input_tokens_seen": 10727104,
      "step": 12375
    },
    {
      "epoch": 5.836869401225837,
      "grad_norm": 0.001980734756216407,
      "learning_rate": 0.23451701198967384,
      "loss": 0.2139,
      "num_input_tokens_seen": 10731504,
      "step": 12380
    },
    {
      "epoch": 5.839226779820839,
      "grad_norm": 0.001704265596345067,
      "learning_rate": 0.23446834108167397,
      "loss": 0.2064,
      "num_input_tokens_seen": 10735312,
      "step": 12385
    },
    {
      "epoch": 5.841584158415841,
      "grad_norm": 0.002763498807325959,
      "learning_rate": 0.23441965714759438,
      "loss": 0.2384,
      "num_input_tokens_seen": 10738928,
      "step": 12390
    },
    {
      "epoch": 5.843941537010844,
      "grad_norm": 0.0023877169005572796,
      "learning_rate": 0.23437096019494277,
      "loss": 0.2521,
      "num_input_tokens_seen": 10743568,
      "step": 12395
    },
    {
      "epoch": 5.846298915605846,
      "grad_norm": 0.0027713722083717585,
      "learning_rate": 0.23432225023122885,
      "loss": 0.2396,
      "num_input_tokens_seen": 10748304,
      "step": 12400
    },
    {
      "epoch": 5.846298915605846,
      "eval_loss": 0.2352769374847412,
      "eval_runtime": 21.8727,
      "eval_samples_per_second": 43.113,
      "eval_steps_per_second": 21.579,
      "num_input_tokens_seen": 10748304,
      "step": 12400
    },
    {
      "epoch": 5.848656294200849,
      "grad_norm": 0.0020673151593655348,
      "learning_rate": 0.23427352726396428,
      "loss": 0.2414,
      "num_input_tokens_seen": 10752736,
      "step": 12405
    },
    {
      "epoch": 5.851013672795851,
      "grad_norm": 0.0024210019037127495,
      "learning_rate": 0.2342247913006628,
      "loss": 0.2222,
      "num_input_tokens_seen": 10757216,
      "step": 12410
    },
    {
      "epoch": 5.8533710513908535,
      "grad_norm": 0.001864079968072474,
      "learning_rate": 0.23417604234883999,
      "loss": 0.2299,
      "num_input_tokens_seen": 10761056,
      "step": 12415
    },
    {
      "epoch": 5.855728429985856,
      "grad_norm": 0.0025835155975073576,
      "learning_rate": 0.23412728041601363,
      "loss": 0.2279,
      "num_input_tokens_seen": 10764784,
      "step": 12420
    },
    {
      "epoch": 5.858085808580858,
      "grad_norm": 0.0016529257409274578,
      "learning_rate": 0.23407850550970347,
      "loss": 0.2741,
      "num_input_tokens_seen": 10769296,
      "step": 12425
    },
    {
      "epoch": 5.860443187175861,
      "grad_norm": 0.0013685290468856692,
      "learning_rate": 0.23402971763743116,
      "loss": 0.2394,
      "num_input_tokens_seen": 10773200,
      "step": 12430
    },
    {
      "epoch": 5.862800565770863,
      "grad_norm": 0.001540546421892941,
      "learning_rate": 0.23398091680672037,
      "loss": 0.2498,
      "num_input_tokens_seen": 10777568,
      "step": 12435
    },
    {
      "epoch": 5.8651579443658655,
      "grad_norm": 0.0034385027829557657,
      "learning_rate": 0.23393210302509687,
      "loss": 0.24,
      "num_input_tokens_seen": 10782112,
      "step": 12440
    },
    {
      "epoch": 5.867515322960868,
      "grad_norm": 0.0028305246960371733,
      "learning_rate": 0.23388327630008832,
      "loss": 0.2464,
      "num_input_tokens_seen": 10786176,
      "step": 12445
    },
    {
      "epoch": 5.86987270155587,
      "grad_norm": 0.0017376882024109364,
      "learning_rate": 0.23383443663922443,
      "loss": 0.234,
      "num_input_tokens_seen": 10790592,
      "step": 12450
    },
    {
      "epoch": 5.872230080150873,
      "grad_norm": 0.0033119251020252705,
      "learning_rate": 0.23378558405003685,
      "loss": 0.2365,
      "num_input_tokens_seen": 10794480,
      "step": 12455
    },
    {
      "epoch": 5.874587458745875,
      "grad_norm": 0.0028617975767701864,
      "learning_rate": 0.2337367185400593,
      "loss": 0.2187,
      "num_input_tokens_seen": 10799120,
      "step": 12460
    },
    {
      "epoch": 5.876944837340877,
      "grad_norm": 0.00145920820068568,
      "learning_rate": 0.23368784011682747,
      "loss": 0.224,
      "num_input_tokens_seen": 10803936,
      "step": 12465
    },
    {
      "epoch": 5.879302215935879,
      "grad_norm": 0.0030155812855809927,
      "learning_rate": 0.23363894878787902,
      "loss": 0.2395,
      "num_input_tokens_seen": 10807936,
      "step": 12470
    },
    {
      "epoch": 5.881659594530881,
      "grad_norm": 0.0026623005978763103,
      "learning_rate": 0.23359004456075352,
      "loss": 0.2415,
      "num_input_tokens_seen": 10812032,
      "step": 12475
    },
    {
      "epoch": 5.884016973125884,
      "grad_norm": 0.0015894895186647773,
      "learning_rate": 0.23354112744299277,
      "loss": 0.2272,
      "num_input_tokens_seen": 10816032,
      "step": 12480
    },
    {
      "epoch": 5.886374351720886,
      "grad_norm": 0.001163095817901194,
      "learning_rate": 0.2334921974421403,
      "loss": 0.2324,
      "num_input_tokens_seen": 10819840,
      "step": 12485
    },
    {
      "epoch": 5.888731730315889,
      "grad_norm": 0.0015502591850236058,
      "learning_rate": 0.23344325456574178,
      "loss": 0.2084,
      "num_input_tokens_seen": 10823920,
      "step": 12490
    },
    {
      "epoch": 5.891089108910891,
      "grad_norm": 0.0013511625584214926,
      "learning_rate": 0.23339429882134477,
      "loss": 0.1981,
      "num_input_tokens_seen": 10827776,
      "step": 12495
    },
    {
      "epoch": 5.893446487505893,
      "grad_norm": 0.0033754869364202023,
      "learning_rate": 0.23334533021649884,
      "loss": 0.2348,
      "num_input_tokens_seen": 10833952,
      "step": 12500
    },
    {
      "epoch": 5.895803866100896,
      "grad_norm": 0.0011947082821279764,
      "learning_rate": 0.23329634875875566,
      "loss": 0.2233,
      "num_input_tokens_seen": 10837664,
      "step": 12505
    },
    {
      "epoch": 5.898161244695898,
      "grad_norm": 0.0013754245592281222,
      "learning_rate": 0.23324735445566874,
      "loss": 0.2927,
      "num_input_tokens_seen": 10841616,
      "step": 12510
    },
    {
      "epoch": 5.900518623290901,
      "grad_norm": 0.0012503743637353182,
      "learning_rate": 0.2331983473147936,
      "loss": 0.2141,
      "num_input_tokens_seen": 10845760,
      "step": 12515
    },
    {
      "epoch": 5.902876001885903,
      "grad_norm": 0.0026672605890780687,
      "learning_rate": 0.23314932734368776,
      "loss": 0.2235,
      "num_input_tokens_seen": 10850128,
      "step": 12520
    },
    {
      "epoch": 5.905233380480905,
      "grad_norm": 0.003437110222876072,
      "learning_rate": 0.2331002945499107,
      "loss": 0.2335,
      "num_input_tokens_seen": 10854448,
      "step": 12525
    },
    {
      "epoch": 5.907590759075908,
      "grad_norm": 0.0038402690552175045,
      "learning_rate": 0.23305124894102397,
      "loss": 0.2118,
      "num_input_tokens_seen": 10858656,
      "step": 12530
    },
    {
      "epoch": 5.90994813767091,
      "grad_norm": 0.004162314813584089,
      "learning_rate": 0.23300219052459092,
      "loss": 0.2584,
      "num_input_tokens_seen": 10863216,
      "step": 12535
    },
    {
      "epoch": 5.912305516265913,
      "grad_norm": 0.0019859131425619125,
      "learning_rate": 0.23295311930817708,
      "loss": 0.2433,
      "num_input_tokens_seen": 10867168,
      "step": 12540
    },
    {
      "epoch": 5.914662894860915,
      "grad_norm": 0.0019307526526972651,
      "learning_rate": 0.23290403529934972,
      "loss": 0.2285,
      "num_input_tokens_seen": 10871312,
      "step": 12545
    },
    {
      "epoch": 5.9170202734559165,
      "grad_norm": 0.0045627751387655735,
      "learning_rate": 0.23285493850567832,
      "loss": 0.2366,
      "num_input_tokens_seen": 10876528,
      "step": 12550
    },
    {
      "epoch": 5.919377652050919,
      "grad_norm": 0.0032950923778116703,
      "learning_rate": 0.23280582893473414,
      "loss": 0.2286,
      "num_input_tokens_seen": 10881120,
      "step": 12555
    },
    {
      "epoch": 5.921735030645921,
      "grad_norm": 0.003994523081928492,
      "learning_rate": 0.2327567065940906,
      "loss": 0.2379,
      "num_input_tokens_seen": 10884832,
      "step": 12560
    },
    {
      "epoch": 5.924092409240924,
      "grad_norm": 0.005103159695863724,
      "learning_rate": 0.23270757149132285,
      "loss": 0.2384,
      "num_input_tokens_seen": 10888816,
      "step": 12565
    },
    {
      "epoch": 5.926449787835926,
      "grad_norm": 0.00353094725869596,
      "learning_rate": 0.23265842363400827,
      "loss": 0.1811,
      "num_input_tokens_seen": 10892448,
      "step": 12570
    },
    {
      "epoch": 5.9288071664309285,
      "grad_norm": 0.005643037613481283,
      "learning_rate": 0.23260926302972595,
      "loss": 0.2172,
      "num_input_tokens_seen": 10896688,
      "step": 12575
    },
    {
      "epoch": 5.931164545025931,
      "grad_norm": 0.003823287785053253,
      "learning_rate": 0.2325600896860572,
      "loss": 0.2654,
      "num_input_tokens_seen": 10900736,
      "step": 12580
    },
    {
      "epoch": 5.933521923620933,
      "grad_norm": 0.002154023153707385,
      "learning_rate": 0.23251090361058505,
      "loss": 0.2544,
      "num_input_tokens_seen": 10905344,
      "step": 12585
    },
    {
      "epoch": 5.935879302215936,
      "grad_norm": 0.0033004574943333864,
      "learning_rate": 0.23246170481089476,
      "loss": 0.2598,
      "num_input_tokens_seen": 10909616,
      "step": 12590
    },
    {
      "epoch": 5.938236680810938,
      "grad_norm": 0.001611672225408256,
      "learning_rate": 0.23241249329457317,
      "loss": 0.2256,
      "num_input_tokens_seen": 10913840,
      "step": 12595
    },
    {
      "epoch": 5.9405940594059405,
      "grad_norm": 0.00158850965090096,
      "learning_rate": 0.23236326906920957,
      "loss": 0.1792,
      "num_input_tokens_seen": 10917920,
      "step": 12600
    },
    {
      "epoch": 5.9405940594059405,
      "eval_loss": 0.24834607541561127,
      "eval_runtime": 21.88,
      "eval_samples_per_second": 43.099,
      "eval_steps_per_second": 21.572,
      "num_input_tokens_seen": 10917920,
      "step": 12600
    },
    {
      "epoch": 5.942951438000943,
      "grad_norm": 0.005649749655276537,
      "learning_rate": 0.2323140321423948,
      "loss": 0.3518,
      "num_input_tokens_seen": 10922128,
      "step": 12605
    },
    {
      "epoch": 5.945308816595945,
      "grad_norm": 0.0037630286533385515,
      "learning_rate": 0.23226478252172184,
      "loss": 0.2146,
      "num_input_tokens_seen": 10926528,
      "step": 12610
    },
    {
      "epoch": 5.947666195190948,
      "grad_norm": 0.00246005249209702,
      "learning_rate": 0.23221552021478561,
      "loss": 0.1904,
      "num_input_tokens_seen": 10930816,
      "step": 12615
    },
    {
      "epoch": 5.95002357378595,
      "grad_norm": 0.0024156710132956505,
      "learning_rate": 0.232166245229183,
      "loss": 0.2159,
      "num_input_tokens_seen": 10935488,
      "step": 12620
    },
    {
      "epoch": 5.9523809523809526,
      "grad_norm": 0.0016405563801527023,
      "learning_rate": 0.2321169575725128,
      "loss": 0.2412,
      "num_input_tokens_seen": 10939616,
      "step": 12625
    },
    {
      "epoch": 5.954738330975955,
      "grad_norm": 0.0015273899771273136,
      "learning_rate": 0.23206765725237577,
      "loss": 0.2173,
      "num_input_tokens_seen": 10943616,
      "step": 12630
    },
    {
      "epoch": 5.957095709570957,
      "grad_norm": 0.0014597258996218443,
      "learning_rate": 0.2320183442763747,
      "loss": 0.2216,
      "num_input_tokens_seen": 10947456,
      "step": 12635
    },
    {
      "epoch": 5.95945308816596,
      "grad_norm": 0.003078373847529292,
      "learning_rate": 0.23196901865211422,
      "loss": 0.2405,
      "num_input_tokens_seen": 10951744,
      "step": 12640
    },
    {
      "epoch": 5.961810466760962,
      "grad_norm": 0.002240431262180209,
      "learning_rate": 0.231919680387201,
      "loss": 0.1852,
      "num_input_tokens_seen": 10956416,
      "step": 12645
    },
    {
      "epoch": 5.964167845355965,
      "grad_norm": 0.0017702371114864945,
      "learning_rate": 0.23187032948924358,
      "loss": 0.2046,
      "num_input_tokens_seen": 10960672,
      "step": 12650
    },
    {
      "epoch": 5.966525223950967,
      "grad_norm": 0.001617515692487359,
      "learning_rate": 0.23182096596585247,
      "loss": 0.2364,
      "num_input_tokens_seen": 10964896,
      "step": 12655
    },
    {
      "epoch": 5.968882602545969,
      "grad_norm": 0.0013805010821670294,
      "learning_rate": 0.23177158982464025,
      "loss": 0.2665,
      "num_input_tokens_seen": 10969568,
      "step": 12660
    },
    {
      "epoch": 5.971239981140971,
      "grad_norm": 0.0028485977090895176,
      "learning_rate": 0.23172220107322122,
      "loss": 0.2178,
      "num_input_tokens_seen": 10973712,
      "step": 12665
    },
    {
      "epoch": 5.973597359735973,
      "grad_norm": 0.002199440263211727,
      "learning_rate": 0.23167279971921184,
      "loss": 0.2298,
      "num_input_tokens_seen": 10978464,
      "step": 12670
    },
    {
      "epoch": 5.975954738330976,
      "grad_norm": 0.005057504400610924,
      "learning_rate": 0.23162338577023034,
      "loss": 0.2162,
      "num_input_tokens_seen": 10983392,
      "step": 12675
    },
    {
      "epoch": 5.978312116925978,
      "grad_norm": 0.001169491559267044,
      "learning_rate": 0.23157395923389704,
      "loss": 0.1782,
      "num_input_tokens_seen": 10986928,
      "step": 12680
    },
    {
      "epoch": 5.9806694955209805,
      "grad_norm": 0.0016375831328332424,
      "learning_rate": 0.2315245201178341,
      "loss": 0.29,
      "num_input_tokens_seen": 10990448,
      "step": 12685
    },
    {
      "epoch": 5.983026874115983,
      "grad_norm": 0.0029446629341691732,
      "learning_rate": 0.23147506842966564,
      "loss": 0.2921,
      "num_input_tokens_seen": 10995056,
      "step": 12690
    },
    {
      "epoch": 5.985384252710985,
      "grad_norm": 0.0011914746137335896,
      "learning_rate": 0.23142560417701774,
      "loss": 0.1977,
      "num_input_tokens_seen": 10999568,
      "step": 12695
    },
    {
      "epoch": 5.987741631305988,
      "grad_norm": 0.0010897465981543064,
      "learning_rate": 0.23137612736751845,
      "loss": 0.2362,
      "num_input_tokens_seen": 11003648,
      "step": 12700
    },
    {
      "epoch": 5.99009900990099,
      "grad_norm": 0.002389785135164857,
      "learning_rate": 0.23132663800879766,
      "loss": 0.2203,
      "num_input_tokens_seen": 11009488,
      "step": 12705
    },
    {
      "epoch": 5.9924563884959925,
      "grad_norm": 0.0009399185655638576,
      "learning_rate": 0.2312771361084873,
      "loss": 0.2415,
      "num_input_tokens_seen": 11015024,
      "step": 12710
    },
    {
      "epoch": 5.994813767090995,
      "grad_norm": 0.0021879307460039854,
      "learning_rate": 0.23122762167422112,
      "loss": 0.219,
      "num_input_tokens_seen": 11018768,
      "step": 12715
    },
    {
      "epoch": 5.997171145685997,
      "grad_norm": 0.0021859491243958473,
      "learning_rate": 0.23117809471363493,
      "loss": 0.2844,
      "num_input_tokens_seen": 11023872,
      "step": 12720
    },
    {
      "epoch": 5.999528524281,
      "grad_norm": 0.0009013023227453232,
      "learning_rate": 0.23112855523436637,
      "loss": 0.2157,
      "num_input_tokens_seen": 11027488,
      "step": 12725
    },
    {
      "epoch": 6.001885902876002,
      "grad_norm": 0.002629445167258382,
      "learning_rate": 0.23107900324405511,
      "loss": 0.2291,
      "num_input_tokens_seen": 11031792,
      "step": 12730
    },
    {
      "epoch": 6.0042432814710045,
      "grad_norm": 0.0011355848982930183,
      "learning_rate": 0.2310294387503426,
      "loss": 0.2081,
      "num_input_tokens_seen": 11036432,
      "step": 12735
    },
    {
      "epoch": 6.006600660066007,
      "grad_norm": 0.0005956445238552988,
      "learning_rate": 0.23097986176087237,
      "loss": 0.1925,
      "num_input_tokens_seen": 11040832,
      "step": 12740
    },
    {
      "epoch": 6.008958038661009,
      "grad_norm": 0.0006888190982863307,
      "learning_rate": 0.23093027228328986,
      "loss": 0.2421,
      "num_input_tokens_seen": 11045216,
      "step": 12745
    },
    {
      "epoch": 6.011315417256012,
      "grad_norm": 0.0016853951383382082,
      "learning_rate": 0.23088067032524226,
      "loss": 0.2138,
      "num_input_tokens_seen": 11048560,
      "step": 12750
    },
    {
      "epoch": 6.013672795851014,
      "grad_norm": 0.0016277659451588988,
      "learning_rate": 0.23083105589437888,
      "loss": 0.2091,
      "num_input_tokens_seen": 11053232,
      "step": 12755
    },
    {
      "epoch": 6.016030174446016,
      "grad_norm": 0.0018908953061327338,
      "learning_rate": 0.23078142899835094,
      "loss": 0.2641,
      "num_input_tokens_seen": 11057840,
      "step": 12760
    },
    {
      "epoch": 6.018387553041018,
      "grad_norm": 0.0008780216448940337,
      "learning_rate": 0.23073178964481147,
      "loss": 0.2416,
      "num_input_tokens_seen": 11061616,
      "step": 12765
    },
    {
      "epoch": 6.02074493163602,
      "grad_norm": 0.001263982616364956,
      "learning_rate": 0.2306821378414155,
      "loss": 0.2523,
      "num_input_tokens_seen": 11065584,
      "step": 12770
    },
    {
      "epoch": 6.023102310231023,
      "grad_norm": 0.001900622621178627,
      "learning_rate": 0.2306324735958199,
      "loss": 0.2302,
      "num_input_tokens_seen": 11070272,
      "step": 12775
    },
    {
      "epoch": 6.025459688826025,
      "grad_norm": 0.0019365698099136353,
      "learning_rate": 0.23058279691568362,
      "loss": 0.236,
      "num_input_tokens_seen": 11073904,
      "step": 12780
    },
    {
      "epoch": 6.027817067421028,
      "grad_norm": 0.001042835647240281,
      "learning_rate": 0.23053310780866745,
      "loss": 0.259,
      "num_input_tokens_seen": 11078400,
      "step": 12785
    },
    {
      "epoch": 6.03017444601603,
      "grad_norm": 0.00531243858858943,
      "learning_rate": 0.23048340628243397,
      "loss": 0.2262,
      "num_input_tokens_seen": 11083488,
      "step": 12790
    },
    {
      "epoch": 6.032531824611032,
      "grad_norm": 0.0017946711741387844,
      "learning_rate": 0.23043369234464783,
      "loss": 0.2231,
      "num_input_tokens_seen": 11088656,
      "step": 12795
    },
    {
      "epoch": 6.034889203206035,
      "grad_norm": 0.001121607143431902,
      "learning_rate": 0.2303839660029755,
      "loss": 0.2638,
      "num_input_tokens_seen": 11092736,
      "step": 12800
    },
    {
      "epoch": 6.034889203206035,
      "eval_loss": 0.22787941992282867,
      "eval_runtime": 21.9517,
      "eval_samples_per_second": 42.958,
      "eval_steps_per_second": 21.502,
      "num_input_tokens_seen": 11092736,
      "step": 12800
    },
    {
      "epoch": 6.037246581801037,
      "grad_norm": 0.0012658574851229787,
      "learning_rate": 0.23033422726508548,
      "loss": 0.2116,
      "num_input_tokens_seen": 11097248,
      "step": 12805
    },
    {
      "epoch": 6.03960396039604,
      "grad_norm": 0.002086346037685871,
      "learning_rate": 0.23028447613864808,
      "loss": 0.2496,
      "num_input_tokens_seen": 11102208,
      "step": 12810
    },
    {
      "epoch": 6.041961338991042,
      "grad_norm": 0.001454114099033177,
      "learning_rate": 0.2302347126313355,
      "loss": 0.1992,
      "num_input_tokens_seen": 11106880,
      "step": 12815
    },
    {
      "epoch": 6.044318717586044,
      "grad_norm": 0.0008547223405912519,
      "learning_rate": 0.23018493675082197,
      "loss": 0.2158,
      "num_input_tokens_seen": 11110560,
      "step": 12820
    },
    {
      "epoch": 6.046676096181047,
      "grad_norm": 0.0010011035483330488,
      "learning_rate": 0.2301351485047835,
      "loss": 0.2184,
      "num_input_tokens_seen": 11114496,
      "step": 12825
    },
    {
      "epoch": 6.049033474776049,
      "grad_norm": 0.0009463587775826454,
      "learning_rate": 0.23008534790089813,
      "loss": 0.2302,
      "num_input_tokens_seen": 11118640,
      "step": 12830
    },
    {
      "epoch": 6.051390853371052,
      "grad_norm": 0.0009053719695657492,
      "learning_rate": 0.2300355349468457,
      "loss": 0.2139,
      "num_input_tokens_seen": 11122720,
      "step": 12835
    },
    {
      "epoch": 6.053748231966054,
      "grad_norm": 0.0011583255836740136,
      "learning_rate": 0.22998570965030793,
      "loss": 0.2329,
      "num_input_tokens_seen": 11127840,
      "step": 12840
    },
    {
      "epoch": 6.0561056105610565,
      "grad_norm": 0.0011554175289347768,
      "learning_rate": 0.22993587201896862,
      "loss": 0.1967,
      "num_input_tokens_seen": 11131712,
      "step": 12845
    },
    {
      "epoch": 6.058462989156059,
      "grad_norm": 0.0008630368392914534,
      "learning_rate": 0.2298860220605133,
      "loss": 0.2065,
      "num_input_tokens_seen": 11135808,
      "step": 12850
    },
    {
      "epoch": 6.060820367751061,
      "grad_norm": 0.0015693021705374122,
      "learning_rate": 0.22983615978262942,
      "loss": 0.2602,
      "num_input_tokens_seen": 11140400,
      "step": 12855
    },
    {
      "epoch": 6.063177746346063,
      "grad_norm": 0.002148232888430357,
      "learning_rate": 0.22978628519300648,
      "loss": 0.2469,
      "num_input_tokens_seen": 11144880,
      "step": 12860
    },
    {
      "epoch": 6.065535124941065,
      "grad_norm": 0.0012318813242018223,
      "learning_rate": 0.22973639829933568,
      "loss": 0.2139,
      "num_input_tokens_seen": 11149232,
      "step": 12865
    },
    {
      "epoch": 6.067892503536068,
      "grad_norm": 0.0017584586748853326,
      "learning_rate": 0.22968649910931027,
      "loss": 0.2125,
      "num_input_tokens_seen": 11153792,
      "step": 12870
    },
    {
      "epoch": 6.07024988213107,
      "grad_norm": 0.0011191613739356399,
      "learning_rate": 0.22963658763062528,
      "loss": 0.231,
      "num_input_tokens_seen": 11158320,
      "step": 12875
    },
    {
      "epoch": 6.072607260726072,
      "grad_norm": 0.002522860886529088,
      "learning_rate": 0.22958666387097765,
      "loss": 0.2199,
      "num_input_tokens_seen": 11163424,
      "step": 12880
    },
    {
      "epoch": 6.074964639321075,
      "grad_norm": 0.0012958089355379343,
      "learning_rate": 0.22953672783806633,
      "loss": 0.2461,
      "num_input_tokens_seen": 11168016,
      "step": 12885
    },
    {
      "epoch": 6.077322017916077,
      "grad_norm": 0.0021013442892581224,
      "learning_rate": 0.22948677953959207,
      "loss": 0.1998,
      "num_input_tokens_seen": 11172288,
      "step": 12890
    },
    {
      "epoch": 6.07967939651108,
      "grad_norm": 0.0009940849849954247,
      "learning_rate": 0.2294368189832575,
      "loss": 0.1915,
      "num_input_tokens_seen": 11176112,
      "step": 12895
    },
    {
      "epoch": 6.082036775106082,
      "grad_norm": 0.0015855350065976381,
      "learning_rate": 0.2293868461767672,
      "loss": 0.1963,
      "num_input_tokens_seen": 11180656,
      "step": 12900
    },
    {
      "epoch": 6.084394153701084,
      "grad_norm": 0.0027663009241223335,
      "learning_rate": 0.22933686112782758,
      "loss": 0.2354,
      "num_input_tokens_seen": 11184400,
      "step": 12905
    },
    {
      "epoch": 6.086751532296087,
      "grad_norm": 0.0029524292331188917,
      "learning_rate": 0.22928686384414698,
      "loss": 0.2505,
      "num_input_tokens_seen": 11188704,
      "step": 12910
    },
    {
      "epoch": 6.089108910891089,
      "grad_norm": 0.002262555994093418,
      "learning_rate": 0.22923685433343552,
      "loss": 0.2112,
      "num_input_tokens_seen": 11192848,
      "step": 12915
    },
    {
      "epoch": 6.091466289486092,
      "grad_norm": 0.0012576004955917597,
      "learning_rate": 0.22918683260340542,
      "loss": 0.2634,
      "num_input_tokens_seen": 11197536,
      "step": 12920
    },
    {
      "epoch": 6.093823668081094,
      "grad_norm": 0.0012853280641138554,
      "learning_rate": 0.2291367986617706,
      "loss": 0.234,
      "num_input_tokens_seen": 11201984,
      "step": 12925
    },
    {
      "epoch": 6.096181046676096,
      "grad_norm": 0.002993642818182707,
      "learning_rate": 0.22908675251624697,
      "loss": 0.2416,
      "num_input_tokens_seen": 11206736,
      "step": 12930
    },
    {
      "epoch": 6.098538425271099,
      "grad_norm": 0.0017695165006443858,
      "learning_rate": 0.22903669417455216,
      "loss": 0.2287,
      "num_input_tokens_seen": 11211072,
      "step": 12935
    },
    {
      "epoch": 6.100895803866101,
      "grad_norm": 0.0020101468544453382,
      "learning_rate": 0.22898662364440592,
      "loss": 0.2351,
      "num_input_tokens_seen": 11215392,
      "step": 12940
    },
    {
      "epoch": 6.103253182461104,
      "grad_norm": 0.002004379639402032,
      "learning_rate": 0.2289365409335297,
      "loss": 0.2045,
      "num_input_tokens_seen": 11220000,
      "step": 12945
    },
    {
      "epoch": 6.105610561056106,
      "grad_norm": 0.0008654922130517662,
      "learning_rate": 0.2288864460496469,
      "loss": 0.2127,
      "num_input_tokens_seen": 11224128,
      "step": 12950
    },
    {
      "epoch": 6.107967939651108,
      "grad_norm": 0.0011194651015102863,
      "learning_rate": 0.22883633900048272,
      "loss": 0.2251,
      "num_input_tokens_seen": 11228000,
      "step": 12955
    },
    {
      "epoch": 6.11032531824611,
      "grad_norm": 0.0014099939726293087,
      "learning_rate": 0.2287862197937644,
      "loss": 0.2536,
      "num_input_tokens_seen": 11232304,
      "step": 12960
    },
    {
      "epoch": 6.112682696841112,
      "grad_norm": 0.005159828811883926,
      "learning_rate": 0.2287360884372209,
      "loss": 0.2671,
      "num_input_tokens_seen": 11237584,
      "step": 12965
    },
    {
      "epoch": 6.115040075436115,
      "grad_norm": 0.0009257703786715865,
      "learning_rate": 0.22868594493858307,
      "loss": 0.237,
      "num_input_tokens_seen": 11240992,
      "step": 12970
    },
    {
      "epoch": 6.117397454031117,
      "grad_norm": 0.0016210287576541305,
      "learning_rate": 0.2286357893055837,
      "loss": 0.2427,
      "num_input_tokens_seen": 11245568,
      "step": 12975
    },
    {
      "epoch": 6.1197548326261195,
      "grad_norm": 0.002432390348985791,
      "learning_rate": 0.22858562154595746,
      "loss": 0.2318,
      "num_input_tokens_seen": 11249840,
      "step": 12980
    },
    {
      "epoch": 6.122112211221122,
      "grad_norm": 0.0015136764850467443,
      "learning_rate": 0.22853544166744078,
      "loss": 0.2383,
      "num_input_tokens_seen": 11255968,
      "step": 12985
    },
    {
      "epoch": 6.124469589816124,
      "grad_norm": 0.0009550959803164005,
      "learning_rate": 0.22848524967777206,
      "loss": 0.219,
      "num_input_tokens_seen": 11260752,
      "step": 12990
    },
    {
      "epoch": 6.126826968411127,
      "grad_norm": 0.00235031358897686,
      "learning_rate": 0.22843504558469152,
      "loss": 0.2636,
      "num_input_tokens_seen": 11264720,
      "step": 12995
    },
    {
      "epoch": 6.129184347006129,
      "grad_norm": 0.0013155497144907713,
      "learning_rate": 0.2283848293959413,
      "loss": 0.2217,
      "num_input_tokens_seen": 11269264,
      "step": 13000
    },
    {
      "epoch": 6.129184347006129,
      "eval_loss": 0.2193126231431961,
      "eval_runtime": 21.9105,
      "eval_samples_per_second": 43.039,
      "eval_steps_per_second": 21.542,
      "num_input_tokens_seen": 11269264,
      "step": 13000
    },
    {
      "epoch": 6.1315417256011315,
      "grad_norm": 0.0008669832022860646,
      "learning_rate": 0.22833460111926532,
      "loss": 0.2345,
      "num_input_tokens_seen": 11273904,
      "step": 13005
    },
    {
      "epoch": 6.133899104196134,
      "grad_norm": 0.0024079419672489166,
      "learning_rate": 0.22828436076240946,
      "loss": 0.2313,
      "num_input_tokens_seen": 11279424,
      "step": 13010
    },
    {
      "epoch": 6.136256482791136,
      "grad_norm": 0.0014385926770046353,
      "learning_rate": 0.22823410833312135,
      "loss": 0.2334,
      "num_input_tokens_seen": 11284192,
      "step": 13015
    },
    {
      "epoch": 6.138613861386139,
      "grad_norm": 0.00253147236071527,
      "learning_rate": 0.2281838438391506,
      "loss": 0.2328,
      "num_input_tokens_seen": 11288992,
      "step": 13020
    },
    {
      "epoch": 6.140971239981141,
      "grad_norm": 0.0008541835704818368,
      "learning_rate": 0.22813356728824863,
      "loss": 0.2278,
      "num_input_tokens_seen": 11293456,
      "step": 13025
    },
    {
      "epoch": 6.1433286185761435,
      "grad_norm": 0.001154087483882904,
      "learning_rate": 0.2280832786881687,
      "loss": 0.2291,
      "num_input_tokens_seen": 11297488,
      "step": 13030
    },
    {
      "epoch": 6.145685997171146,
      "grad_norm": 0.0010083775268867612,
      "learning_rate": 0.22803297804666592,
      "loss": 0.233,
      "num_input_tokens_seen": 11302096,
      "step": 13035
    },
    {
      "epoch": 6.148043375766148,
      "grad_norm": 0.002139996038749814,
      "learning_rate": 0.22798266537149728,
      "loss": 0.2196,
      "num_input_tokens_seen": 11305856,
      "step": 13040
    },
    {
      "epoch": 6.150400754361151,
      "grad_norm": 0.0009965640492737293,
      "learning_rate": 0.22793234067042167,
      "loss": 0.1907,
      "num_input_tokens_seen": 11310768,
      "step": 13045
    },
    {
      "epoch": 6.152758132956153,
      "grad_norm": 0.0031632347963750362,
      "learning_rate": 0.22788200395119979,
      "loss": 0.2647,
      "num_input_tokens_seen": 11315152,
      "step": 13050
    },
    {
      "epoch": 6.1551155115511555,
      "grad_norm": 0.0016818954609334469,
      "learning_rate": 0.2278316552215942,
      "loss": 0.2589,
      "num_input_tokens_seen": 11319648,
      "step": 13055
    },
    {
      "epoch": 6.157472890146157,
      "grad_norm": 0.001628947095014155,
      "learning_rate": 0.22778129448936918,
      "loss": 0.232,
      "num_input_tokens_seen": 11323664,
      "step": 13060
    },
    {
      "epoch": 6.1598302687411595,
      "grad_norm": 0.001700666849501431,
      "learning_rate": 0.22773092176229118,
      "loss": 0.2648,
      "num_input_tokens_seen": 11328384,
      "step": 13065
    },
    {
      "epoch": 6.162187647336162,
      "grad_norm": 0.0012237941846251488,
      "learning_rate": 0.22768053704812816,
      "loss": 0.2266,
      "num_input_tokens_seen": 11332608,
      "step": 13070
    },
    {
      "epoch": 6.164545025931164,
      "grad_norm": 0.0009233771124854684,
      "learning_rate": 0.22763014035465018,
      "loss": 0.2099,
      "num_input_tokens_seen": 11337088,
      "step": 13075
    },
    {
      "epoch": 6.166902404526167,
      "grad_norm": 0.0012322289403527975,
      "learning_rate": 0.22757973168962892,
      "loss": 0.2533,
      "num_input_tokens_seen": 11340864,
      "step": 13080
    },
    {
      "epoch": 6.169259783121169,
      "grad_norm": 0.003005904844030738,
      "learning_rate": 0.22752931106083818,
      "loss": 0.269,
      "num_input_tokens_seen": 11345712,
      "step": 13085
    },
    {
      "epoch": 6.1716171617161715,
      "grad_norm": 0.0021580446045845747,
      "learning_rate": 0.22747887847605341,
      "loss": 0.2255,
      "num_input_tokens_seen": 11349296,
      "step": 13090
    },
    {
      "epoch": 6.173974540311174,
      "grad_norm": 0.0011636470444500446,
      "learning_rate": 0.22742843394305184,
      "loss": 0.2373,
      "num_input_tokens_seen": 11353200,
      "step": 13095
    },
    {
      "epoch": 6.176331918906176,
      "grad_norm": 0.0015956066781654954,
      "learning_rate": 0.22737797746961272,
      "loss": 0.2311,
      "num_input_tokens_seen": 11357184,
      "step": 13100
    },
    {
      "epoch": 6.178689297501179,
      "grad_norm": 0.0019788972567766905,
      "learning_rate": 0.22732750906351712,
      "loss": 0.2529,
      "num_input_tokens_seen": 11360736,
      "step": 13105
    },
    {
      "epoch": 6.181046676096181,
      "grad_norm": 0.0042763277888298035,
      "learning_rate": 0.22727702873254785,
      "loss": 0.2388,
      "num_input_tokens_seen": 11365888,
      "step": 13110
    },
    {
      "epoch": 6.1834040546911835,
      "grad_norm": 0.0010787597857415676,
      "learning_rate": 0.22722653648448968,
      "loss": 0.2305,
      "num_input_tokens_seen": 11370256,
      "step": 13115
    },
    {
      "epoch": 6.185761433286186,
      "grad_norm": 0.0011703427881002426,
      "learning_rate": 0.22717603232712902,
      "loss": 0.2286,
      "num_input_tokens_seen": 11375248,
      "step": 13120
    },
    {
      "epoch": 6.188118811881188,
      "grad_norm": 0.002150709507986903,
      "learning_rate": 0.22712551626825436,
      "loss": 0.2132,
      "num_input_tokens_seen": 11380064,
      "step": 13125
    },
    {
      "epoch": 6.190476190476191,
      "grad_norm": 0.0018260151846334338,
      "learning_rate": 0.2270749883156559,
      "loss": 0.228,
      "num_input_tokens_seen": 11384784,
      "step": 13130
    },
    {
      "epoch": 6.192833569071193,
      "grad_norm": 0.0008629740332253277,
      "learning_rate": 0.22702444847712563,
      "loss": 0.2064,
      "num_input_tokens_seen": 11389104,
      "step": 13135
    },
    {
      "epoch": 6.1951909476661955,
      "grad_norm": 0.0013775562401860952,
      "learning_rate": 0.22697389676045743,
      "loss": 0.2048,
      "num_input_tokens_seen": 11393200,
      "step": 13140
    },
    {
      "epoch": 6.197548326261198,
      "grad_norm": 0.0007523690583184361,
      "learning_rate": 0.22692333317344704,
      "loss": 0.2466,
      "num_input_tokens_seen": 11397200,
      "step": 13145
    },
    {
      "epoch": 6.1999057048562,
      "grad_norm": 0.0022523042280226946,
      "learning_rate": 0.22687275772389198,
      "loss": 0.218,
      "num_input_tokens_seen": 11400816,
      "step": 13150
    },
    {
      "epoch": 6.202263083451202,
      "grad_norm": 0.0016035367734730244,
      "learning_rate": 0.22682217041959168,
      "loss": 0.1835,
      "num_input_tokens_seen": 11404432,
      "step": 13155
    },
    {
      "epoch": 6.204620462046204,
      "grad_norm": 0.0015332121402025223,
      "learning_rate": 0.2267715712683473,
      "loss": 0.1668,
      "num_input_tokens_seen": 11408528,
      "step": 13160
    },
    {
      "epoch": 6.206977840641207,
      "grad_norm": 0.0016260864213109016,
      "learning_rate": 0.22672096027796182,
      "loss": 0.2475,
      "num_input_tokens_seen": 11413424,
      "step": 13165
    },
    {
      "epoch": 6.209335219236209,
      "grad_norm": 0.002393875503912568,
      "learning_rate": 0.22667033745624016,
      "loss": 0.266,
      "num_input_tokens_seen": 11416944,
      "step": 13170
    },
    {
      "epoch": 6.211692597831211,
      "grad_norm": 0.0012190823908895254,
      "learning_rate": 0.22661970281098895,
      "loss": 0.2568,
      "num_input_tokens_seen": 11420848,
      "step": 13175
    },
    {
      "epoch": 6.214049976426214,
      "grad_norm": 0.0014554086374118924,
      "learning_rate": 0.22656905635001667,
      "loss": 0.2303,
      "num_input_tokens_seen": 11424976,
      "step": 13180
    },
    {
      "epoch": 6.216407355021216,
      "grad_norm": 0.0010306069161742926,
      "learning_rate": 0.2265183980811337,
      "loss": 0.2329,
      "num_input_tokens_seen": 11429008,
      "step": 13185
    },
    {
      "epoch": 6.218764733616219,
      "grad_norm": 0.0010611936450004578,
      "learning_rate": 0.22646772801215218,
      "loss": 0.2367,
      "num_input_tokens_seen": 11432928,
      "step": 13190
    },
    {
      "epoch": 6.221122112211221,
      "grad_norm": 0.0019025321817025542,
      "learning_rate": 0.22641704615088598,
      "loss": 0.2392,
      "num_input_tokens_seen": 11437120,
      "step": 13195
    },
    {
      "epoch": 6.223479490806223,
      "grad_norm": 0.002721330849453807,
      "learning_rate": 0.22636635250515103,
      "loss": 0.2475,
      "num_input_tokens_seen": 11441120,
      "step": 13200
    },
    {
      "epoch": 6.223479490806223,
      "eval_loss": 0.22213885188102722,
      "eval_runtime": 21.906,
      "eval_samples_per_second": 43.048,
      "eval_steps_per_second": 21.547,
      "num_input_tokens_seen": 11441120,
      "step": 13200
    },
    {
      "epoch": 6.225836869401226,
      "grad_norm": 0.00204481091350317,
      "learning_rate": 0.2263156470827648,
      "loss": 0.22,
      "num_input_tokens_seen": 11445408,
      "step": 13205
    },
    {
      "epoch": 6.228194247996228,
      "grad_norm": 0.0019294159719720483,
      "learning_rate": 0.22626492989154678,
      "loss": 0.2118,
      "num_input_tokens_seen": 11449024,
      "step": 13210
    },
    {
      "epoch": 6.230551626591231,
      "grad_norm": 0.001992699457332492,
      "learning_rate": 0.22621420093931813,
      "loss": 0.2376,
      "num_input_tokens_seen": 11453728,
      "step": 13215
    },
    {
      "epoch": 6.232909005186233,
      "grad_norm": 0.0011381824733689427,
      "learning_rate": 0.22616346023390194,
      "loss": 0.2272,
      "num_input_tokens_seen": 11457600,
      "step": 13220
    },
    {
      "epoch": 6.235266383781235,
      "grad_norm": 0.0010387592483311892,
      "learning_rate": 0.22611270778312306,
      "loss": 0.2055,
      "num_input_tokens_seen": 11461696,
      "step": 13225
    },
    {
      "epoch": 6.237623762376238,
      "grad_norm": 0.0011006842833012342,
      "learning_rate": 0.2260619435948081,
      "loss": 0.259,
      "num_input_tokens_seen": 11465520,
      "step": 13230
    },
    {
      "epoch": 6.23998114097124,
      "grad_norm": 0.001763340551406145,
      "learning_rate": 0.22601116767678567,
      "loss": 0.243,
      "num_input_tokens_seen": 11470688,
      "step": 13235
    },
    {
      "epoch": 6.242338519566243,
      "grad_norm": 0.0009628318948671222,
      "learning_rate": 0.2259603800368859,
      "loss": 0.2254,
      "num_input_tokens_seen": 11475664,
      "step": 13240
    },
    {
      "epoch": 6.244695898161245,
      "grad_norm": 0.001001010532490909,
      "learning_rate": 0.22590958068294098,
      "loss": 0.224,
      "num_input_tokens_seen": 11479984,
      "step": 13245
    },
    {
      "epoch": 6.247053276756247,
      "grad_norm": 0.0011504285503178835,
      "learning_rate": 0.22585876962278478,
      "loss": 0.2374,
      "num_input_tokens_seen": 11484928,
      "step": 13250
    },
    {
      "epoch": 6.24941065535125,
      "grad_norm": 0.001338228932581842,
      "learning_rate": 0.22580794686425298,
      "loss": 0.2343,
      "num_input_tokens_seen": 11489760,
      "step": 13255
    },
    {
      "epoch": 6.251768033946251,
      "grad_norm": 0.0008723827195353806,
      "learning_rate": 0.22575711241518312,
      "loss": 0.2368,
      "num_input_tokens_seen": 11493472,
      "step": 13260
    },
    {
      "epoch": 6.254125412541254,
      "grad_norm": 0.0009143627248704433,
      "learning_rate": 0.22570626628341453,
      "loss": 0.2278,
      "num_input_tokens_seen": 11497152,
      "step": 13265
    },
    {
      "epoch": 6.256482791136256,
      "grad_norm": 0.0012281552189961076,
      "learning_rate": 0.22565540847678828,
      "loss": 0.2107,
      "num_input_tokens_seen": 11501568,
      "step": 13270
    },
    {
      "epoch": 6.258840169731259,
      "grad_norm": 0.001585561316460371,
      "learning_rate": 0.2256045390031473,
      "loss": 0.2503,
      "num_input_tokens_seen": 11505456,
      "step": 13275
    },
    {
      "epoch": 6.261197548326261,
      "grad_norm": 0.0009572579292580485,
      "learning_rate": 0.22555365787033627,
      "loss": 0.2253,
      "num_input_tokens_seen": 11509680,
      "step": 13280
    },
    {
      "epoch": 6.263554926921263,
      "grad_norm": 0.0012867635814473033,
      "learning_rate": 0.22550276508620173,
      "loss": 0.2301,
      "num_input_tokens_seen": 11514208,
      "step": 13285
    },
    {
      "epoch": 6.265912305516266,
      "grad_norm": 0.0019787573255598545,
      "learning_rate": 0.22545186065859202,
      "loss": 0.2019,
      "num_input_tokens_seen": 11518128,
      "step": 13290
    },
    {
      "epoch": 6.268269684111268,
      "grad_norm": 0.0018884100718423724,
      "learning_rate": 0.2254009445953572,
      "loss": 0.2139,
      "num_input_tokens_seen": 11522384,
      "step": 13295
    },
    {
      "epoch": 6.270627062706271,
      "grad_norm": 0.003030792810022831,
      "learning_rate": 0.22535001690434917,
      "loss": 0.246,
      "num_input_tokens_seen": 11526848,
      "step": 13300
    },
    {
      "epoch": 6.272984441301273,
      "grad_norm": 0.0011117053218185902,
      "learning_rate": 0.22529907759342163,
      "loss": 0.2186,
      "num_input_tokens_seen": 11531360,
      "step": 13305
    },
    {
      "epoch": 6.275341819896275,
      "grad_norm": 0.0012730286689475179,
      "learning_rate": 0.22524812667043007,
      "loss": 0.2403,
      "num_input_tokens_seen": 11536464,
      "step": 13310
    },
    {
      "epoch": 6.277699198491278,
      "grad_norm": 0.002316867234185338,
      "learning_rate": 0.22519716414323177,
      "loss": 0.2148,
      "num_input_tokens_seen": 11540816,
      "step": 13315
    },
    {
      "epoch": 6.28005657708628,
      "grad_norm": 0.003442081855610013,
      "learning_rate": 0.22514619001968567,
      "loss": 0.221,
      "num_input_tokens_seen": 11547056,
      "step": 13320
    },
    {
      "epoch": 6.282413955681283,
      "grad_norm": 0.0009833906078711152,
      "learning_rate": 0.2250952043076528,
      "loss": 0.2232,
      "num_input_tokens_seen": 11550496,
      "step": 13325
    },
    {
      "epoch": 6.284771334276285,
      "grad_norm": 0.0012660115025937557,
      "learning_rate": 0.2250442070149957,
      "loss": 0.2265,
      "num_input_tokens_seen": 11554240,
      "step": 13330
    },
    {
      "epoch": 6.287128712871287,
      "grad_norm": 0.0010071959113702178,
      "learning_rate": 0.22499319814957885,
      "loss": 0.2139,
      "num_input_tokens_seen": 11558064,
      "step": 13335
    },
    {
      "epoch": 6.28948609146629,
      "grad_norm": 0.0008060761028900743,
      "learning_rate": 0.2249421777192684,
      "loss": 0.1877,
      "num_input_tokens_seen": 11562368,
      "step": 13340
    },
    {
      "epoch": 6.291843470061292,
      "grad_norm": 0.0008081582491286099,
      "learning_rate": 0.22489114573193236,
      "loss": 0.2333,
      "num_input_tokens_seen": 11566704,
      "step": 13345
    },
    {
      "epoch": 6.294200848656295,
      "grad_norm": 0.0007041542557999492,
      "learning_rate": 0.2248401021954405,
      "loss": 0.2567,
      "num_input_tokens_seen": 11571408,
      "step": 13350
    },
    {
      "epoch": 6.296558227251296,
      "grad_norm": 0.0010830189567059278,
      "learning_rate": 0.22478904711766443,
      "loss": 0.2572,
      "num_input_tokens_seen": 11575408,
      "step": 13355
    },
    {
      "epoch": 6.2989156058462985,
      "grad_norm": 0.0015778144588693976,
      "learning_rate": 0.22473798050647734,
      "loss": 0.2438,
      "num_input_tokens_seen": 11580160,
      "step": 13360
    },
    {
      "epoch": 6.301272984441301,
      "grad_norm": 0.001337361754849553,
      "learning_rate": 0.22468690236975453,
      "loss": 0.2291,
      "num_input_tokens_seen": 11584144,
      "step": 13365
    },
    {
      "epoch": 6.303630363036303,
      "grad_norm": 0.0007003672071732581,
      "learning_rate": 0.22463581271537272,
      "loss": 0.2409,
      "num_input_tokens_seen": 11588048,
      "step": 13370
    },
    {
      "epoch": 6.305987741631306,
      "grad_norm": 0.0019723675213754177,
      "learning_rate": 0.22458471155121076,
      "loss": 0.2189,
      "num_input_tokens_seen": 11592144,
      "step": 13375
    },
    {
      "epoch": 6.308345120226308,
      "grad_norm": 0.001178054721094668,
      "learning_rate": 0.2245335988851489,
      "loss": 0.2193,
      "num_input_tokens_seen": 11596080,
      "step": 13380
    },
    {
      "epoch": 6.3107024988213105,
      "grad_norm": 0.003723704721778631,
      "learning_rate": 0.2244824747250695,
      "loss": 0.2514,
      "num_input_tokens_seen": 11601232,
      "step": 13385
    },
    {
      "epoch": 6.313059877416313,
      "grad_norm": 0.0010304072638973594,
      "learning_rate": 0.22443133907885646,
      "loss": 0.2189,
      "num_input_tokens_seen": 11605472,
      "step": 13390
    },
    {
      "epoch": 6.315417256011315,
      "grad_norm": 0.0007270880159921944,
      "learning_rate": 0.22438019195439557,
      "loss": 0.2224,
      "num_input_tokens_seen": 11610256,
      "step": 13395
    },
    {
      "epoch": 6.317774634606318,
      "grad_norm": 0.0010840314207598567,
      "learning_rate": 0.22432903335957435,
      "loss": 0.2348,
      "num_input_tokens_seen": 11614176,
      "step": 13400
    },
    {
      "epoch": 6.317774634606318,
      "eval_loss": 0.21879000961780548,
      "eval_runtime": 21.9065,
      "eval_samples_per_second": 43.047,
      "eval_steps_per_second": 21.546,
      "num_input_tokens_seen": 11614176,
      "step": 13400
    },
    {
      "epoch": 6.32013201320132,
      "grad_norm": 0.0018415770027786493,
      "learning_rate": 0.22427786330228214,
      "loss": 0.2209,
      "num_input_tokens_seen": 11617808,
      "step": 13405
    },
    {
      "epoch": 6.3224893917963225,
      "grad_norm": 0.0014010494342073798,
      "learning_rate": 0.22422668179040997,
      "loss": 0.2352,
      "num_input_tokens_seen": 11622544,
      "step": 13410
    },
    {
      "epoch": 6.324846770391325,
      "grad_norm": 0.0009197145118378103,
      "learning_rate": 0.2241754888318507,
      "loss": 0.206,
      "num_input_tokens_seen": 11627168,
      "step": 13415
    },
    {
      "epoch": 6.327204148986327,
      "grad_norm": 0.0009297403157688677,
      "learning_rate": 0.22412428443449886,
      "loss": 0.2084,
      "num_input_tokens_seen": 11631296,
      "step": 13420
    },
    {
      "epoch": 6.32956152758133,
      "grad_norm": 0.0006599495536647737,
      "learning_rate": 0.22407306860625087,
      "loss": 0.1864,
      "num_input_tokens_seen": 11636192,
      "step": 13425
    },
    {
      "epoch": 6.331918906176332,
      "grad_norm": 0.0016978821950033307,
      "learning_rate": 0.22402184135500483,
      "loss": 0.207,
      "num_input_tokens_seen": 11640592,
      "step": 13430
    },
    {
      "epoch": 6.3342762847713345,
      "grad_norm": 0.0014252585824579,
      "learning_rate": 0.22397060268866067,
      "loss": 0.2803,
      "num_input_tokens_seen": 11645024,
      "step": 13435
    },
    {
      "epoch": 6.336633663366337,
      "grad_norm": 0.0016570795560255647,
      "learning_rate": 0.22391935261511994,
      "loss": 0.2446,
      "num_input_tokens_seen": 11649184,
      "step": 13440
    },
    {
      "epoch": 6.338991041961339,
      "grad_norm": 0.0012716721976175904,
      "learning_rate": 0.22386809114228615,
      "loss": 0.2211,
      "num_input_tokens_seen": 11653584,
      "step": 13445
    },
    {
      "epoch": 6.341348420556342,
      "grad_norm": 0.0011080008698627353,
      "learning_rate": 0.22381681827806446,
      "loss": 0.2187,
      "num_input_tokens_seen": 11658176,
      "step": 13450
    },
    {
      "epoch": 6.343705799151344,
      "grad_norm": 0.0018002965953201056,
      "learning_rate": 0.22376553403036173,
      "loss": 0.2306,
      "num_input_tokens_seen": 11663008,
      "step": 13455
    },
    {
      "epoch": 6.346063177746346,
      "grad_norm": 0.002703767968341708,
      "learning_rate": 0.22371423840708662,
      "loss": 0.2335,
      "num_input_tokens_seen": 11668144,
      "step": 13460
    },
    {
      "epoch": 6.348420556341348,
      "grad_norm": 0.0010058965999633074,
      "learning_rate": 0.22366293141614962,
      "loss": 0.2282,
      "num_input_tokens_seen": 11672048,
      "step": 13465
    },
    {
      "epoch": 6.3507779349363505,
      "grad_norm": 0.0017756945453584194,
      "learning_rate": 0.22361161306546287,
      "loss": 0.2002,
      "num_input_tokens_seen": 11676000,
      "step": 13470
    },
    {
      "epoch": 6.353135313531353,
      "grad_norm": 0.001078912173397839,
      "learning_rate": 0.22356028336294037,
      "loss": 0.2383,
      "num_input_tokens_seen": 11680224,
      "step": 13475
    },
    {
      "epoch": 6.355492692126355,
      "grad_norm": 0.0011349486885592341,
      "learning_rate": 0.2235089423164977,
      "loss": 0.2243,
      "num_input_tokens_seen": 11684240,
      "step": 13480
    },
    {
      "epoch": 6.357850070721358,
      "grad_norm": 0.003357166424393654,
      "learning_rate": 0.22345758993405243,
      "loss": 0.2018,
      "num_input_tokens_seen": 11689632,
      "step": 13485
    },
    {
      "epoch": 6.36020744931636,
      "grad_norm": 0.0011042780242860317,
      "learning_rate": 0.2234062262235236,
      "loss": 0.23,
      "num_input_tokens_seen": 11693328,
      "step": 13490
    },
    {
      "epoch": 6.3625648279113625,
      "grad_norm": 0.0011562854051589966,
      "learning_rate": 0.22335485119283222,
      "loss": 0.2074,
      "num_input_tokens_seen": 11697984,
      "step": 13495
    },
    {
      "epoch": 6.364922206506365,
      "grad_norm": 0.0009417433757334948,
      "learning_rate": 0.22330346484990093,
      "loss": 0.2326,
      "num_input_tokens_seen": 11702480,
      "step": 13500
    },
    {
      "epoch": 6.367279585101367,
      "grad_norm": 0.0018067003693431616,
      "learning_rate": 0.22325206720265425,
      "loss": 0.2316,
      "num_input_tokens_seen": 11706128,
      "step": 13505
    },
    {
      "epoch": 6.36963696369637,
      "grad_norm": 0.0008320966153405607,
      "learning_rate": 0.2232006582590182,
      "loss": 0.2264,
      "num_input_tokens_seen": 11709680,
      "step": 13510
    },
    {
      "epoch": 6.371994342291372,
      "grad_norm": 0.0012325753923505545,
      "learning_rate": 0.22314923802692077,
      "loss": 0.2043,
      "num_input_tokens_seen": 11714976,
      "step": 13515
    },
    {
      "epoch": 6.3743517208863745,
      "grad_norm": 0.0010789618827402592,
      "learning_rate": 0.22309780651429156,
      "loss": 0.198,
      "num_input_tokens_seen": 11718704,
      "step": 13520
    },
    {
      "epoch": 6.376709099481377,
      "grad_norm": 0.0006451330264098942,
      "learning_rate": 0.22304636372906203,
      "loss": 0.1883,
      "num_input_tokens_seen": 11722768,
      "step": 13525
    },
    {
      "epoch": 6.379066478076379,
      "grad_norm": 0.001135817845351994,
      "learning_rate": 0.22299490967916522,
      "loss": 0.2047,
      "num_input_tokens_seen": 11727136,
      "step": 13530
    },
    {
      "epoch": 6.381423856671382,
      "grad_norm": 0.0009520763996988535,
      "learning_rate": 0.22294344437253602,
      "loss": 0.2229,
      "num_input_tokens_seen": 11730400,
      "step": 13535
    },
    {
      "epoch": 6.383781235266384,
      "grad_norm": 0.0007598455995321274,
      "learning_rate": 0.22289196781711101,
      "loss": 0.1864,
      "num_input_tokens_seen": 11734736,
      "step": 13540
    },
    {
      "epoch": 6.3861386138613865,
      "grad_norm": 0.0027822055853903294,
      "learning_rate": 0.2228404800208286,
      "loss": 0.2665,
      "num_input_tokens_seen": 11739744,
      "step": 13545
    },
    {
      "epoch": 6.388495992456389,
      "grad_norm": 0.0009701235685497522,
      "learning_rate": 0.22278898099162875,
      "loss": 0.2207,
      "num_input_tokens_seen": 11743712,
      "step": 13550
    },
    {
      "epoch": 6.39085337105139,
      "grad_norm": 0.0015511472010985017,
      "learning_rate": 0.22273747073745337,
      "loss": 0.2183,
      "num_input_tokens_seen": 11748384,
      "step": 13555
    },
    {
      "epoch": 6.393210749646393,
      "grad_norm": 0.0012719410005956888,
      "learning_rate": 0.22268594926624588,
      "loss": 0.24,
      "num_input_tokens_seen": 11753120,
      "step": 13560
    },
    {
      "epoch": 6.395568128241395,
      "grad_norm": 0.0011915857903659344,
      "learning_rate": 0.22263441658595162,
      "loss": 0.223,
      "num_input_tokens_seen": 11757744,
      "step": 13565
    },
    {
      "epoch": 6.397925506836398,
      "grad_norm": 0.0010721537983044982,
      "learning_rate": 0.2225828727045175,
      "loss": 0.2127,
      "num_input_tokens_seen": 11761600,
      "step": 13570
    },
    {
      "epoch": 6.4002828854314,
      "grad_norm": 0.000847398245241493,
      "learning_rate": 0.22253131762989228,
      "loss": 0.2256,
      "num_input_tokens_seen": 11765472,
      "step": 13575
    },
    {
      "epoch": 6.402640264026402,
      "grad_norm": 0.0011851497692987323,
      "learning_rate": 0.2224797513700264,
      "loss": 0.2423,
      "num_input_tokens_seen": 11769824,
      "step": 13580
    },
    {
      "epoch": 6.404997642621405,
      "grad_norm": 0.0015840467531234026,
      "learning_rate": 0.22242817393287204,
      "loss": 0.1854,
      "num_input_tokens_seen": 11773648,
      "step": 13585
    },
    {
      "epoch": 6.407355021216407,
      "grad_norm": 0.0005339236231520772,
      "learning_rate": 0.22237658532638305,
      "loss": 0.1915,
      "num_input_tokens_seen": 11777456,
      "step": 13590
    },
    {
      "epoch": 6.40971239981141,
      "grad_norm": 0.001420942833647132,
      "learning_rate": 0.22232498555851513,
      "loss": 0.229,
      "num_input_tokens_seen": 11781792,
      "step": 13595
    },
    {
      "epoch": 6.412069778406412,
      "grad_norm": 0.0008033700869418681,
      "learning_rate": 0.22227337463722546,
      "loss": 0.1659,
      "num_input_tokens_seen": 11785424,
      "step": 13600
    },
    {
      "epoch": 6.412069778406412,
      "eval_loss": 0.22639155387878418,
      "eval_runtime": 21.9253,
      "eval_samples_per_second": 43.01,
      "eval_steps_per_second": 21.528,
      "num_input_tokens_seen": 11785424,
      "step": 13600
    },
    {
      "epoch": 6.414427157001414,
      "grad_norm": 0.0008431847090832889,
      "learning_rate": 0.2222217525704732,
      "loss": 0.2197,
      "num_input_tokens_seen": 11789584,
      "step": 13605
    },
    {
      "epoch": 6.416784535596417,
      "grad_norm": 0.0013008592650294304,
      "learning_rate": 0.22217011936621908,
      "loss": 0.1686,
      "num_input_tokens_seen": 11793728,
      "step": 13610
    },
    {
      "epoch": 6.419141914191419,
      "grad_norm": 0.0013016933808103204,
      "learning_rate": 0.22211847503242566,
      "loss": 0.2165,
      "num_input_tokens_seen": 11798288,
      "step": 13615
    },
    {
      "epoch": 6.421499292786422,
      "grad_norm": 0.001297068316489458,
      "learning_rate": 0.22206681957705704,
      "loss": 0.2326,
      "num_input_tokens_seen": 11802368,
      "step": 13620
    },
    {
      "epoch": 6.423856671381424,
      "grad_norm": 0.001070246915332973,
      "learning_rate": 0.2220151530080792,
      "loss": 0.256,
      "num_input_tokens_seen": 11806656,
      "step": 13625
    },
    {
      "epoch": 6.426214049976426,
      "grad_norm": 0.0008172846864908934,
      "learning_rate": 0.2219634753334598,
      "loss": 0.23,
      "num_input_tokens_seen": 11810912,
      "step": 13630
    },
    {
      "epoch": 6.428571428571429,
      "grad_norm": 0.0012299135560169816,
      "learning_rate": 0.22191178656116817,
      "loss": 0.2221,
      "num_input_tokens_seen": 11815056,
      "step": 13635
    },
    {
      "epoch": 6.430928807166431,
      "grad_norm": 0.0009284638217650354,
      "learning_rate": 0.2218600866991753,
      "loss": 0.2203,
      "num_input_tokens_seen": 11819824,
      "step": 13640
    },
    {
      "epoch": 6.433286185761434,
      "grad_norm": 0.0011982939904555678,
      "learning_rate": 0.221808375755454,
      "loss": 0.1999,
      "num_input_tokens_seen": 11823888,
      "step": 13645
    },
    {
      "epoch": 6.435643564356436,
      "grad_norm": 0.0013545859837904572,
      "learning_rate": 0.22175665373797881,
      "loss": 0.1694,
      "num_input_tokens_seen": 11828256,
      "step": 13650
    },
    {
      "epoch": 6.438000942951438,
      "grad_norm": 0.0008710703114047647,
      "learning_rate": 0.22170492065472583,
      "loss": 0.2263,
      "num_input_tokens_seen": 11832848,
      "step": 13655
    },
    {
      "epoch": 6.44035832154644,
      "grad_norm": 0.0010835877619683743,
      "learning_rate": 0.221653176513673,
      "loss": 0.2121,
      "num_input_tokens_seen": 11837616,
      "step": 13660
    },
    {
      "epoch": 6.442715700141442,
      "grad_norm": 0.000717986433301121,
      "learning_rate": 0.2216014213227999,
      "loss": 0.1887,
      "num_input_tokens_seen": 11842144,
      "step": 13665
    },
    {
      "epoch": 6.445073078736445,
      "grad_norm": 0.0017253223340958357,
      "learning_rate": 0.22154965509008784,
      "loss": 0.1897,
      "num_input_tokens_seen": 11846560,
      "step": 13670
    },
    {
      "epoch": 6.447430457331447,
      "grad_norm": 0.0029545503202825785,
      "learning_rate": 0.2214978778235198,
      "loss": 0.2797,
      "num_input_tokens_seen": 11850208,
      "step": 13675
    },
    {
      "epoch": 6.4497878359264496,
      "grad_norm": 0.000699088501278311,
      "learning_rate": 0.2214460895310805,
      "loss": 0.1575,
      "num_input_tokens_seen": 11854624,
      "step": 13680
    },
    {
      "epoch": 6.452145214521452,
      "grad_norm": 0.002054176526144147,
      "learning_rate": 0.22139429022075635,
      "loss": 0.2904,
      "num_input_tokens_seen": 11859680,
      "step": 13685
    },
    {
      "epoch": 6.454502593116454,
      "grad_norm": 0.0012557499576359987,
      "learning_rate": 0.22134247990053546,
      "loss": 0.1981,
      "num_input_tokens_seen": 11863888,
      "step": 13690
    },
    {
      "epoch": 6.456859971711457,
      "grad_norm": 0.0009751489269547164,
      "learning_rate": 0.2212906585784076,
      "loss": 0.2085,
      "num_input_tokens_seen": 11868640,
      "step": 13695
    },
    {
      "epoch": 6.459217350306459,
      "grad_norm": 0.0026942628901451826,
      "learning_rate": 0.22123882626236432,
      "loss": 0.2429,
      "num_input_tokens_seen": 11872816,
      "step": 13700
    },
    {
      "epoch": 6.461574728901462,
      "grad_norm": 0.0016532782465219498,
      "learning_rate": 0.2211869829603988,
      "loss": 0.218,
      "num_input_tokens_seen": 11876288,
      "step": 13705
    },
    {
      "epoch": 6.463932107496464,
      "grad_norm": 0.0011220305459573865,
      "learning_rate": 0.22113512868050592,
      "loss": 0.2131,
      "num_input_tokens_seen": 11880528,
      "step": 13710
    },
    {
      "epoch": 6.466289486091466,
      "grad_norm": 0.0019360133446753025,
      "learning_rate": 0.2210832634306822,
      "loss": 0.2156,
      "num_input_tokens_seen": 11884512,
      "step": 13715
    },
    {
      "epoch": 6.468646864686469,
      "grad_norm": 0.0020713992416858673,
      "learning_rate": 0.22103138721892598,
      "loss": 0.2303,
      "num_input_tokens_seen": 11889184,
      "step": 13720
    },
    {
      "epoch": 6.471004243281471,
      "grad_norm": 0.0010106242261826992,
      "learning_rate": 0.22097950005323724,
      "loss": 0.2152,
      "num_input_tokens_seen": 11894064,
      "step": 13725
    },
    {
      "epoch": 6.473361621876474,
      "grad_norm": 0.0021346251014620066,
      "learning_rate": 0.22092760194161762,
      "loss": 0.2397,
      "num_input_tokens_seen": 11898992,
      "step": 13730
    },
    {
      "epoch": 6.475719000471476,
      "grad_norm": 0.0012591255363076925,
      "learning_rate": 0.2208756928920704,
      "loss": 0.2453,
      "num_input_tokens_seen": 11903840,
      "step": 13735
    },
    {
      "epoch": 6.478076379066478,
      "grad_norm": 0.001246446743607521,
      "learning_rate": 0.22082377291260072,
      "loss": 0.2233,
      "num_input_tokens_seen": 11907792,
      "step": 13740
    },
    {
      "epoch": 6.480433757661481,
      "grad_norm": 0.001191683579236269,
      "learning_rate": 0.2207718420112152,
      "loss": 0.2179,
      "num_input_tokens_seen": 11911536,
      "step": 13745
    },
    {
      "epoch": 6.482791136256483,
      "grad_norm": 0.0011459790403023362,
      "learning_rate": 0.22071990019592228,
      "loss": 0.2098,
      "num_input_tokens_seen": 11915920,
      "step": 13750
    },
    {
      "epoch": 6.485148514851485,
      "grad_norm": 0.0010541322408244014,
      "learning_rate": 0.22066794747473198,
      "loss": 0.2049,
      "num_input_tokens_seen": 11920528,
      "step": 13755
    },
    {
      "epoch": 6.487505893446487,
      "grad_norm": 0.0008160487632267177,
      "learning_rate": 0.2206159838556562,
      "loss": 0.1813,
      "num_input_tokens_seen": 11924784,
      "step": 13760
    },
    {
      "epoch": 6.4898632720414895,
      "grad_norm": 0.0019650927279144526,
      "learning_rate": 0.2205640093467082,
      "loss": 0.1918,
      "num_input_tokens_seen": 11928976,
      "step": 13765
    },
    {
      "epoch": 6.492220650636492,
      "grad_norm": 0.001733272336423397,
      "learning_rate": 0.22051202395590322,
      "loss": 0.2427,
      "num_input_tokens_seen": 11933984,
      "step": 13770
    },
    {
      "epoch": 6.494578029231494,
      "grad_norm": 0.0015685093821957707,
      "learning_rate": 0.22046002769125808,
      "loss": 0.2583,
      "num_input_tokens_seen": 11938400,
      "step": 13775
    },
    {
      "epoch": 6.496935407826497,
      "grad_norm": 0.0021439415868371725,
      "learning_rate": 0.2204080205607912,
      "loss": 0.2374,
      "num_input_tokens_seen": 11942928,
      "step": 13780
    },
    {
      "epoch": 6.499292786421499,
      "grad_norm": 0.0013968285638839006,
      "learning_rate": 0.22035600257252272,
      "loss": 0.2286,
      "num_input_tokens_seen": 11947536,
      "step": 13785
    },
    {
      "epoch": 6.5016501650165015,
      "grad_norm": 0.0019134548492729664,
      "learning_rate": 0.2203039737344745,
      "loss": 0.2182,
      "num_input_tokens_seen": 11951392,
      "step": 13790
    },
    {
      "epoch": 6.504007543611504,
      "grad_norm": 0.001454550656490028,
      "learning_rate": 0.22025193405467003,
      "loss": 0.2238,
      "num_input_tokens_seen": 11955776,
      "step": 13795
    },
    {
      "epoch": 6.506364922206506,
      "grad_norm": 0.001808667671866715,
      "learning_rate": 0.2201998835411345,
      "loss": 0.1838,
      "num_input_tokens_seen": 11960752,
      "step": 13800
    },
    {
      "epoch": 6.506364922206506,
      "eval_loss": 0.21857322752475739,
      "eval_runtime": 21.8866,
      "eval_samples_per_second": 43.086,
      "eval_steps_per_second": 21.566,
      "num_input_tokens_seen": 11960752,
      "step": 13800
    },
    {
      "epoch": 6.508722300801509,
      "grad_norm": 0.0036930737551301718,
      "learning_rate": 0.22014782220189474,
      "loss": 0.2434,
      "num_input_tokens_seen": 11965200,
      "step": 13805
    },
    {
      "epoch": 6.511079679396511,
      "grad_norm": 0.002268847543746233,
      "learning_rate": 0.2200957500449793,
      "loss": 0.2343,
      "num_input_tokens_seen": 11969120,
      "step": 13810
    },
    {
      "epoch": 6.5134370579915135,
      "grad_norm": 0.002026001224294305,
      "learning_rate": 0.22004366707841827,
      "loss": 0.2153,
      "num_input_tokens_seen": 11972944,
      "step": 13815
    },
    {
      "epoch": 6.515794436586516,
      "grad_norm": 0.0013748891651630402,
      "learning_rate": 0.21999157331024358,
      "loss": 0.2174,
      "num_input_tokens_seen": 11976752,
      "step": 13820
    },
    {
      "epoch": 6.518151815181518,
      "grad_norm": 0.0009646362741477787,
      "learning_rate": 0.21993946874848871,
      "loss": 0.2029,
      "num_input_tokens_seen": 11980896,
      "step": 13825
    },
    {
      "epoch": 6.520509193776521,
      "grad_norm": 0.0010318633867427707,
      "learning_rate": 0.2198873534011888,
      "loss": 0.1945,
      "num_input_tokens_seen": 11984960,
      "step": 13830
    },
    {
      "epoch": 6.522866572371523,
      "grad_norm": 0.001288180355913937,
      "learning_rate": 0.2198352272763808,
      "loss": 0.2305,
      "num_input_tokens_seen": 11989552,
      "step": 13835
    },
    {
      "epoch": 6.5252239509665255,
      "grad_norm": 0.0008407334098592401,
      "learning_rate": 0.2197830903821031,
      "loss": 0.1662,
      "num_input_tokens_seen": 11993584,
      "step": 13840
    },
    {
      "epoch": 6.527581329561528,
      "grad_norm": 0.001323329284787178,
      "learning_rate": 0.21973094272639598,
      "loss": 0.1991,
      "num_input_tokens_seen": 11997936,
      "step": 13845
    },
    {
      "epoch": 6.52993870815653,
      "grad_norm": 0.0010704640299081802,
      "learning_rate": 0.21967878431730117,
      "loss": 0.163,
      "num_input_tokens_seen": 12002848,
      "step": 13850
    },
    {
      "epoch": 6.532296086751533,
      "grad_norm": 0.0009565091459080577,
      "learning_rate": 0.21962661516286217,
      "loss": 0.2335,
      "num_input_tokens_seen": 12006912,
      "step": 13855
    },
    {
      "epoch": 6.534653465346535,
      "grad_norm": 0.0009982381016016006,
      "learning_rate": 0.21957443527112414,
      "loss": 0.2455,
      "num_input_tokens_seen": 12010976,
      "step": 13860
    },
    {
      "epoch": 6.537010843941537,
      "grad_norm": 0.0016621286049485207,
      "learning_rate": 0.21952224465013384,
      "loss": 0.2518,
      "num_input_tokens_seen": 12016208,
      "step": 13865
    },
    {
      "epoch": 6.539368222536539,
      "grad_norm": 0.0015303674153983593,
      "learning_rate": 0.21947004330793976,
      "loss": 0.2285,
      "num_input_tokens_seen": 12019856,
      "step": 13870
    },
    {
      "epoch": 6.5417256011315414,
      "grad_norm": 0.001197399222292006,
      "learning_rate": 0.21941783125259198,
      "loss": 0.2203,
      "num_input_tokens_seen": 12024480,
      "step": 13875
    },
    {
      "epoch": 6.544082979726544,
      "grad_norm": 0.0017466588178649545,
      "learning_rate": 0.21936560849214226,
      "loss": 0.2075,
      "num_input_tokens_seen": 12029072,
      "step": 13880
    },
    {
      "epoch": 6.546440358321546,
      "grad_norm": 0.0017335258889943361,
      "learning_rate": 0.21931337503464404,
      "loss": 0.2232,
      "num_input_tokens_seen": 12032688,
      "step": 13885
    },
    {
      "epoch": 6.548797736916549,
      "grad_norm": 0.001155829755589366,
      "learning_rate": 0.21926113088815233,
      "loss": 0.2096,
      "num_input_tokens_seen": 12036752,
      "step": 13890
    },
    {
      "epoch": 6.551155115511551,
      "grad_norm": 0.0014500155812129378,
      "learning_rate": 0.2192088760607238,
      "loss": 0.1907,
      "num_input_tokens_seen": 12040816,
      "step": 13895
    },
    {
      "epoch": 6.5535124941065535,
      "grad_norm": 0.001542670768685639,
      "learning_rate": 0.2191566105604169,
      "loss": 0.2291,
      "num_input_tokens_seen": 12046176,
      "step": 13900
    },
    {
      "epoch": 6.555869872701556,
      "grad_norm": 0.0013711112551391125,
      "learning_rate": 0.21910433439529153,
      "loss": 0.2466,
      "num_input_tokens_seen": 12050112,
      "step": 13905
    },
    {
      "epoch": 6.558227251296558,
      "grad_norm": 0.0011703994823619723,
      "learning_rate": 0.2190520475734094,
      "loss": 0.2462,
      "num_input_tokens_seen": 12054656,
      "step": 13910
    },
    {
      "epoch": 6.560584629891561,
      "grad_norm": 0.0010362978791818023,
      "learning_rate": 0.2189997501028338,
      "loss": 0.2049,
      "num_input_tokens_seen": 12058640,
      "step": 13915
    },
    {
      "epoch": 6.562942008486563,
      "grad_norm": 0.0031943386420607567,
      "learning_rate": 0.2189474419916296,
      "loss": 0.2066,
      "num_input_tokens_seen": 12063888,
      "step": 13920
    },
    {
      "epoch": 6.5652993870815655,
      "grad_norm": 0.0010580772068351507,
      "learning_rate": 0.21889512324786342,
      "loss": 0.2357,
      "num_input_tokens_seen": 12068064,
      "step": 13925
    },
    {
      "epoch": 6.567656765676568,
      "grad_norm": 0.003131187055259943,
      "learning_rate": 0.21884279387960345,
      "loss": 0.2366,
      "num_input_tokens_seen": 12072704,
      "step": 13930
    },
    {
      "epoch": 6.57001414427157,
      "grad_norm": 0.0009805160807445645,
      "learning_rate": 0.2187904538949195,
      "loss": 0.2007,
      "num_input_tokens_seen": 12076976,
      "step": 13935
    },
    {
      "epoch": 6.572371522866573,
      "grad_norm": 0.0008368046255782247,
      "learning_rate": 0.2187381033018831,
      "loss": 0.1932,
      "num_input_tokens_seen": 12080960,
      "step": 13940
    },
    {
      "epoch": 6.574728901461575,
      "grad_norm": 0.0008214358240365982,
      "learning_rate": 0.2186857421085673,
      "loss": 0.2011,
      "num_input_tokens_seen": 12085664,
      "step": 13945
    },
    {
      "epoch": 6.5770862800565775,
      "grad_norm": 0.0014826244441792369,
      "learning_rate": 0.21863337032304697,
      "loss": 0.2242,
      "num_input_tokens_seen": 12090448,
      "step": 13950
    },
    {
      "epoch": 6.579443658651579,
      "grad_norm": 0.0007693592342548072,
      "learning_rate": 0.21858098795339845,
      "loss": 0.2367,
      "num_input_tokens_seen": 12094384,
      "step": 13955
    },
    {
      "epoch": 6.581801037246581,
      "grad_norm": 0.0011449640151113272,
      "learning_rate": 0.21852859500769975,
      "loss": 0.2009,
      "num_input_tokens_seen": 12098720,
      "step": 13960
    },
    {
      "epoch": 6.584158415841584,
      "grad_norm": 0.0011215161066502333,
      "learning_rate": 0.21847619149403044,
      "loss": 0.2424,
      "num_input_tokens_seen": 12103072,
      "step": 13965
    },
    {
      "epoch": 6.586515794436586,
      "grad_norm": 0.0010358479339629412,
      "learning_rate": 0.21842377742047195,
      "loss": 0.2258,
      "num_input_tokens_seen": 12107840,
      "step": 13970
    },
    {
      "epoch": 6.588873173031589,
      "grad_norm": 0.0015076696872711182,
      "learning_rate": 0.21837135279510705,
      "loss": 0.2318,
      "num_input_tokens_seen": 12111984,
      "step": 13975
    },
    {
      "epoch": 6.591230551626591,
      "grad_norm": 0.0018098392756655812,
      "learning_rate": 0.21831891762602038,
      "loss": 0.2324,
      "num_input_tokens_seen": 12116800,
      "step": 13980
    },
    {
      "epoch": 6.593587930221593,
      "grad_norm": 0.0009791397023946047,
      "learning_rate": 0.21826647192129806,
      "loss": 0.2481,
      "num_input_tokens_seen": 12120752,
      "step": 13985
    },
    {
      "epoch": 6.595945308816596,
      "grad_norm": 0.0014681629836559296,
      "learning_rate": 0.21821401568902787,
      "loss": 0.2392,
      "num_input_tokens_seen": 12124848,
      "step": 13990
    },
    {
      "epoch": 6.598302687411598,
      "grad_norm": 0.0008760616183280945,
      "learning_rate": 0.21816154893729925,
      "loss": 0.2329,
      "num_input_tokens_seen": 12128784,
      "step": 13995
    },
    {
      "epoch": 6.600660066006601,
      "grad_norm": 0.0011897670337930322,
      "learning_rate": 0.2181090716742032,
      "loss": 0.222,
      "num_input_tokens_seen": 12132672,
      "step": 14000
    },
    {
      "epoch": 6.600660066006601,
      "eval_loss": 0.21828874945640564,
      "eval_runtime": 21.8717,
      "eval_samples_per_second": 43.115,
      "eval_steps_per_second": 21.58,
      "num_input_tokens_seen": 12132672,
      "step": 14000
    },
    {
      "epoch": 6.603017444601603,
      "grad_norm": 0.0012794890208169818,
      "learning_rate": 0.21805658390783236,
      "loss": 0.1993,
      "num_input_tokens_seen": 12136768,
      "step": 14005
    },
    {
      "epoch": 6.605374823196605,
      "grad_norm": 0.000785662850830704,
      "learning_rate": 0.21800408564628107,
      "loss": 0.2398,
      "num_input_tokens_seen": 12141456,
      "step": 14010
    },
    {
      "epoch": 6.607732201791608,
      "grad_norm": 0.0013250084593892097,
      "learning_rate": 0.21795157689764516,
      "loss": 0.1657,
      "num_input_tokens_seen": 12145920,
      "step": 14015
    },
    {
      "epoch": 6.61008958038661,
      "grad_norm": 0.0008349485578946769,
      "learning_rate": 0.21789905767002216,
      "loss": 0.2691,
      "num_input_tokens_seen": 12149760,
      "step": 14020
    },
    {
      "epoch": 6.612446958981613,
      "grad_norm": 0.0014035304775461555,
      "learning_rate": 0.2178465279715112,
      "loss": 0.1522,
      "num_input_tokens_seen": 12153088,
      "step": 14025
    },
    {
      "epoch": 6.614804337576615,
      "grad_norm": 0.0023050843738019466,
      "learning_rate": 0.21779398781021303,
      "loss": 0.2429,
      "num_input_tokens_seen": 12158048,
      "step": 14030
    },
    {
      "epoch": 6.617161716171617,
      "grad_norm": 0.001078520668670535,
      "learning_rate": 0.21774143719422998,
      "loss": 0.2879,
      "num_input_tokens_seen": 12162864,
      "step": 14035
    },
    {
      "epoch": 6.61951909476662,
      "grad_norm": 0.001807692926377058,
      "learning_rate": 0.21768887613166601,
      "loss": 0.2009,
      "num_input_tokens_seen": 12168352,
      "step": 14040
    },
    {
      "epoch": 6.621876473361622,
      "grad_norm": 0.0025743923615664244,
      "learning_rate": 0.2176363046306267,
      "loss": 0.2327,
      "num_input_tokens_seen": 12172768,
      "step": 14045
    },
    {
      "epoch": 6.624233851956625,
      "grad_norm": 0.0013855824945494533,
      "learning_rate": 0.21758372269921925,
      "loss": 0.2286,
      "num_input_tokens_seen": 12177296,
      "step": 14050
    },
    {
      "epoch": 6.626591230551627,
      "grad_norm": 0.0009394611115567386,
      "learning_rate": 0.21753113034555244,
      "loss": 0.2383,
      "num_input_tokens_seen": 12180576,
      "step": 14055
    },
    {
      "epoch": 6.628948609146629,
      "grad_norm": 0.0020144490990787745,
      "learning_rate": 0.2174785275777367,
      "loss": 0.2314,
      "num_input_tokens_seen": 12185184,
      "step": 14060
    },
    {
      "epoch": 6.631305987741631,
      "grad_norm": 0.001557944342494011,
      "learning_rate": 0.21742591440388404,
      "loss": 0.2271,
      "num_input_tokens_seen": 12189376,
      "step": 14065
    },
    {
      "epoch": 6.633663366336633,
      "grad_norm": 0.002066993387416005,
      "learning_rate": 0.21737329083210802,
      "loss": 0.2295,
      "num_input_tokens_seen": 12192992,
      "step": 14070
    },
    {
      "epoch": 6.636020744931636,
      "grad_norm": 0.001033844193443656,
      "learning_rate": 0.2173206568705239,
      "loss": 0.2175,
      "num_input_tokens_seen": 12197520,
      "step": 14075
    },
    {
      "epoch": 6.638378123526638,
      "grad_norm": 0.0018377258675172925,
      "learning_rate": 0.2172680125272485,
      "loss": 0.2237,
      "num_input_tokens_seen": 12201728,
      "step": 14080
    },
    {
      "epoch": 6.6407355021216405,
      "grad_norm": 0.0008944787550717592,
      "learning_rate": 0.2172153578104002,
      "loss": 0.2191,
      "num_input_tokens_seen": 12205648,
      "step": 14085
    },
    {
      "epoch": 6.643092880716643,
      "grad_norm": 0.0010799543233588338,
      "learning_rate": 0.21716269272809902,
      "loss": 0.2033,
      "num_input_tokens_seen": 12209312,
      "step": 14090
    },
    {
      "epoch": 6.645450259311645,
      "grad_norm": 0.0018022455042228103,
      "learning_rate": 0.21711001728846666,
      "loss": 0.1952,
      "num_input_tokens_seen": 12213728,
      "step": 14095
    },
    {
      "epoch": 6.647807637906648,
      "grad_norm": 0.002291918033733964,
      "learning_rate": 0.21705733149962628,
      "loss": 0.2702,
      "num_input_tokens_seen": 12218208,
      "step": 14100
    },
    {
      "epoch": 6.65016501650165,
      "grad_norm": 0.0008098890539258718,
      "learning_rate": 0.21700463536970263,
      "loss": 0.1928,
      "num_input_tokens_seen": 12222608,
      "step": 14105
    },
    {
      "epoch": 6.6525223950966526,
      "grad_norm": 0.0014783412916585803,
      "learning_rate": 0.21695192890682222,
      "loss": 0.212,
      "num_input_tokens_seen": 12226224,
      "step": 14110
    },
    {
      "epoch": 6.654879773691655,
      "grad_norm": 0.001821559970267117,
      "learning_rate": 0.21689921211911298,
      "loss": 0.2149,
      "num_input_tokens_seen": 12230192,
      "step": 14115
    },
    {
      "epoch": 6.657237152286657,
      "grad_norm": 0.0020028608851134777,
      "learning_rate": 0.21684648501470452,
      "loss": 0.2207,
      "num_input_tokens_seen": 12234768,
      "step": 14120
    },
    {
      "epoch": 6.65959453088166,
      "grad_norm": 0.0007979890797287226,
      "learning_rate": 0.216793747601728,
      "loss": 0.2001,
      "num_input_tokens_seen": 12238960,
      "step": 14125
    },
    {
      "epoch": 6.661951909476662,
      "grad_norm": 0.0014030808815732598,
      "learning_rate": 0.21674099988831627,
      "loss": 0.2027,
      "num_input_tokens_seen": 12243616,
      "step": 14130
    },
    {
      "epoch": 6.664309288071665,
      "grad_norm": 0.0007040109485387802,
      "learning_rate": 0.21668824188260363,
      "loss": 0.2349,
      "num_input_tokens_seen": 12248496,
      "step": 14135
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.0029200236313045025,
      "learning_rate": 0.21663547359272606,
      "loss": 0.2395,
      "num_input_tokens_seen": 12253184,
      "step": 14140
    },
    {
      "epoch": 6.669024045261669,
      "grad_norm": 0.0009673293679952621,
      "learning_rate": 0.216582695026821,
      "loss": 0.2233,
      "num_input_tokens_seen": 12256848,
      "step": 14145
    },
    {
      "epoch": 6.671381423856672,
      "grad_norm": 0.003125124843791127,
      "learning_rate": 0.21652990619302767,
      "loss": 0.2249,
      "num_input_tokens_seen": 12261184,
      "step": 14150
    },
    {
      "epoch": 6.673738802451673,
      "grad_norm": 0.0009346173610538244,
      "learning_rate": 0.21647710709948673,
      "loss": 0.1769,
      "num_input_tokens_seen": 12265504,
      "step": 14155
    },
    {
      "epoch": 6.676096181046676,
      "grad_norm": 0.0015139473835006356,
      "learning_rate": 0.2164242977543405,
      "loss": 0.2006,
      "num_input_tokens_seen": 12269248,
      "step": 14160
    },
    {
      "epoch": 6.678453559641678,
      "grad_norm": 0.0007926283869892359,
      "learning_rate": 0.21637147816573277,
      "loss": 0.2125,
      "num_input_tokens_seen": 12273072,
      "step": 14165
    },
    {
      "epoch": 6.6808109382366805,
      "grad_norm": 0.001193562988191843,
      "learning_rate": 0.21631864834180908,
      "loss": 0.226,
      "num_input_tokens_seen": 12278112,
      "step": 14170
    },
    {
      "epoch": 6.683168316831683,
      "grad_norm": 0.0011515958467498422,
      "learning_rate": 0.21626580829071637,
      "loss": 0.2133,
      "num_input_tokens_seen": 12282320,
      "step": 14175
    },
    {
      "epoch": 6.685525695426685,
      "grad_norm": 0.000817668333183974,
      "learning_rate": 0.21621295802060328,
      "loss": 0.2283,
      "num_input_tokens_seen": 12286320,
      "step": 14180
    },
    {
      "epoch": 6.687883074021688,
      "grad_norm": 0.002206181176006794,
      "learning_rate": 0.21616009753961996,
      "loss": 0.2097,
      "num_input_tokens_seen": 12290448,
      "step": 14185
    },
    {
      "epoch": 6.69024045261669,
      "grad_norm": 0.0017704832134768367,
      "learning_rate": 0.2161072268559182,
      "loss": 0.2088,
      "num_input_tokens_seen": 12294864,
      "step": 14190
    },
    {
      "epoch": 6.6925978312116925,
      "grad_norm": 0.0014467319706454873,
      "learning_rate": 0.21605434597765133,
      "loss": 0.2303,
      "num_input_tokens_seen": 12299360,
      "step": 14195
    },
    {
      "epoch": 6.694955209806695,
      "grad_norm": 0.0021095958072692156,
      "learning_rate": 0.21600145491297418,
      "loss": 0.2841,
      "num_input_tokens_seen": 12303424,
      "step": 14200
    },
    {
      "epoch": 6.694955209806695,
      "eval_loss": 0.2159731090068817,
      "eval_runtime": 21.9216,
      "eval_samples_per_second": 43.017,
      "eval_steps_per_second": 21.531,
      "num_input_tokens_seen": 12303424,
      "step": 14200
    },
    {
      "epoch": 6.697312588401697,
      "grad_norm": 0.0008643734036013484,
      "learning_rate": 0.21594855367004326,
      "loss": 0.1991,
      "num_input_tokens_seen": 12307952,
      "step": 14205
    },
    {
      "epoch": 6.6996699669967,
      "grad_norm": 0.00099911168217659,
      "learning_rate": 0.21589564225701663,
      "loss": 0.1988,
      "num_input_tokens_seen": 12311552,
      "step": 14210
    },
    {
      "epoch": 6.702027345591702,
      "grad_norm": 0.000859889667481184,
      "learning_rate": 0.21584272068205385,
      "loss": 0.1859,
      "num_input_tokens_seen": 12314848,
      "step": 14215
    },
    {
      "epoch": 6.7043847241867045,
      "grad_norm": 0.0019989567808806896,
      "learning_rate": 0.2157897889533161,
      "loss": 0.2642,
      "num_input_tokens_seen": 12318800,
      "step": 14220
    },
    {
      "epoch": 6.706742102781707,
      "grad_norm": 0.001182570238597691,
      "learning_rate": 0.21573684707896612,
      "loss": 0.2393,
      "num_input_tokens_seen": 12323520,
      "step": 14225
    },
    {
      "epoch": 6.709099481376709,
      "grad_norm": 0.0009859168203547597,
      "learning_rate": 0.21568389506716826,
      "loss": 0.2222,
      "num_input_tokens_seen": 12328048,
      "step": 14230
    },
    {
      "epoch": 6.711456859971712,
      "grad_norm": 0.0027129182126373053,
      "learning_rate": 0.21563093292608831,
      "loss": 0.2617,
      "num_input_tokens_seen": 12332224,
      "step": 14235
    },
    {
      "epoch": 6.713814238566714,
      "grad_norm": 0.0008775991154834628,
      "learning_rate": 0.21557796066389376,
      "loss": 0.2275,
      "num_input_tokens_seen": 12336368,
      "step": 14240
    },
    {
      "epoch": 6.7161716171617165,
      "grad_norm": 0.0010439966572448611,
      "learning_rate": 0.21552497828875353,
      "loss": 0.2241,
      "num_input_tokens_seen": 12340544,
      "step": 14245
    },
    {
      "epoch": 6.718528995756719,
      "grad_norm": 0.000740572577342391,
      "learning_rate": 0.21547198580883828,
      "loss": 0.224,
      "num_input_tokens_seen": 12345072,
      "step": 14250
    },
    {
      "epoch": 6.720886374351721,
      "grad_norm": 0.0015781527617946267,
      "learning_rate": 0.21541898323232,
      "loss": 0.1997,
      "num_input_tokens_seen": 12351008,
      "step": 14255
    },
    {
      "epoch": 6.723243752946724,
      "grad_norm": 0.0012488086940720677,
      "learning_rate": 0.2153659705673724,
      "loss": 0.2146,
      "num_input_tokens_seen": 12355584,
      "step": 14260
    },
    {
      "epoch": 6.725601131541725,
      "grad_norm": 0.0024416050873696804,
      "learning_rate": 0.2153129478221707,
      "loss": 0.2455,
      "num_input_tokens_seen": 12359680,
      "step": 14265
    },
    {
      "epoch": 6.727958510136728,
      "grad_norm": 0.0008311145356856287,
      "learning_rate": 0.21525991500489164,
      "loss": 0.2439,
      "num_input_tokens_seen": 12363184,
      "step": 14270
    },
    {
      "epoch": 6.73031588873173,
      "grad_norm": 0.0007930436404421926,
      "learning_rate": 0.21520687212371362,
      "loss": 0.241,
      "num_input_tokens_seen": 12367200,
      "step": 14275
    },
    {
      "epoch": 6.732673267326732,
      "grad_norm": 0.00171434145886451,
      "learning_rate": 0.21515381918681648,
      "loss": 0.2016,
      "num_input_tokens_seen": 12371808,
      "step": 14280
    },
    {
      "epoch": 6.735030645921735,
      "grad_norm": 0.0007788133807480335,
      "learning_rate": 0.21510075620238167,
      "loss": 0.2511,
      "num_input_tokens_seen": 12375872,
      "step": 14285
    },
    {
      "epoch": 6.737388024516737,
      "grad_norm": 0.0016605884302407503,
      "learning_rate": 0.21504768317859208,
      "loss": 0.1864,
      "num_input_tokens_seen": 12379664,
      "step": 14290
    },
    {
      "epoch": 6.73974540311174,
      "grad_norm": 0.0015451674116775393,
      "learning_rate": 0.2149946001236323,
      "loss": 0.2635,
      "num_input_tokens_seen": 12383792,
      "step": 14295
    },
    {
      "epoch": 6.742102781706742,
      "grad_norm": 0.0018625977681949735,
      "learning_rate": 0.21494150704568848,
      "loss": 0.2212,
      "num_input_tokens_seen": 12389120,
      "step": 14300
    },
    {
      "epoch": 6.7444601603017444,
      "grad_norm": 0.0005674086860381067,
      "learning_rate": 0.21488840395294811,
      "loss": 0.2385,
      "num_input_tokens_seen": 12392992,
      "step": 14305
    },
    {
      "epoch": 6.746817538896747,
      "grad_norm": 0.0023185175377875566,
      "learning_rate": 0.21483529085360042,
      "loss": 0.2444,
      "num_input_tokens_seen": 12397616,
      "step": 14310
    },
    {
      "epoch": 6.749174917491749,
      "grad_norm": 0.001108133583329618,
      "learning_rate": 0.2147821677558361,
      "loss": 0.2367,
      "num_input_tokens_seen": 12402096,
      "step": 14315
    },
    {
      "epoch": 6.751532296086752,
      "grad_norm": 0.0007715950487181544,
      "learning_rate": 0.2147290346678475,
      "loss": 0.2103,
      "num_input_tokens_seen": 12406528,
      "step": 14320
    },
    {
      "epoch": 6.753889674681754,
      "grad_norm": 0.0015502843307331204,
      "learning_rate": 0.21467589159782827,
      "loss": 0.2176,
      "num_input_tokens_seen": 12411648,
      "step": 14325
    },
    {
      "epoch": 6.7562470532767565,
      "grad_norm": 0.0017109279287979007,
      "learning_rate": 0.21462273855397374,
      "loss": 0.2424,
      "num_input_tokens_seen": 12416800,
      "step": 14330
    },
    {
      "epoch": 6.758604431871759,
      "grad_norm": 0.0011203422909602523,
      "learning_rate": 0.21456957554448083,
      "loss": 0.2519,
      "num_input_tokens_seen": 12420608,
      "step": 14335
    },
    {
      "epoch": 6.760961810466761,
      "grad_norm": 0.0007673196960240602,
      "learning_rate": 0.21451640257754795,
      "loss": 0.2212,
      "num_input_tokens_seen": 12424832,
      "step": 14340
    },
    {
      "epoch": 6.763319189061764,
      "grad_norm": 0.0007216534577310085,
      "learning_rate": 0.21446321966137508,
      "loss": 0.2131,
      "num_input_tokens_seen": 12429024,
      "step": 14345
    },
    {
      "epoch": 6.765676567656766,
      "grad_norm": 0.0020378034096211195,
      "learning_rate": 0.21441002680416354,
      "loss": 0.2425,
      "num_input_tokens_seen": 12433040,
      "step": 14350
    },
    {
      "epoch": 6.768033946251768,
      "grad_norm": 0.0010856781154870987,
      "learning_rate": 0.21435682401411654,
      "loss": 0.2335,
      "num_input_tokens_seen": 12436544,
      "step": 14355
    },
    {
      "epoch": 6.77039132484677,
      "grad_norm": 0.000835278129670769,
      "learning_rate": 0.2143036112994385,
      "loss": 0.2226,
      "num_input_tokens_seen": 12440912,
      "step": 14360
    },
    {
      "epoch": 6.772748703441772,
      "grad_norm": 0.0015102807665243745,
      "learning_rate": 0.21425038866833548,
      "loss": 0.233,
      "num_input_tokens_seen": 12444672,
      "step": 14365
    },
    {
      "epoch": 6.775106082036775,
      "grad_norm": 0.0010432589333504438,
      "learning_rate": 0.21419715612901508,
      "loss": 0.2343,
      "num_input_tokens_seen": 12449344,
      "step": 14370
    },
    {
      "epoch": 6.777463460631777,
      "grad_norm": 0.0008407129789702594,
      "learning_rate": 0.21414391368968652,
      "loss": 0.2192,
      "num_input_tokens_seen": 12453120,
      "step": 14375
    },
    {
      "epoch": 6.77982083922678,
      "grad_norm": 0.0009093019179999828,
      "learning_rate": 0.21409066135856034,
      "loss": 0.1922,
      "num_input_tokens_seen": 12457120,
      "step": 14380
    },
    {
      "epoch": 6.782178217821782,
      "grad_norm": 0.0008792545995675027,
      "learning_rate": 0.21403739914384878,
      "loss": 0.2313,
      "num_input_tokens_seen": 12461392,
      "step": 14385
    },
    {
      "epoch": 6.784535596416784,
      "grad_norm": 0.0007856666925363243,
      "learning_rate": 0.21398412705376554,
      "loss": 0.2415,
      "num_input_tokens_seen": 12465408,
      "step": 14390
    },
    {
      "epoch": 6.786892975011787,
      "grad_norm": 0.0007584239356219769,
      "learning_rate": 0.2139308450965258,
      "loss": 0.2373,
      "num_input_tokens_seen": 12470256,
      "step": 14395
    },
    {
      "epoch": 6.789250353606789,
      "grad_norm": 0.001454116660170257,
      "learning_rate": 0.21387755328034638,
      "loss": 0.2265,
      "num_input_tokens_seen": 12474592,
      "step": 14400
    },
    {
      "epoch": 6.789250353606789,
      "eval_loss": 0.22272275388240814,
      "eval_runtime": 21.946,
      "eval_samples_per_second": 42.969,
      "eval_steps_per_second": 21.507,
      "num_input_tokens_seen": 12474592,
      "step": 14400
    },
    {
      "epoch": 6.791607732201792,
      "grad_norm": 0.0012333409395068884,
      "learning_rate": 0.2138242516134455,
      "loss": 0.2356,
      "num_input_tokens_seen": 12479136,
      "step": 14405
    },
    {
      "epoch": 6.793965110796794,
      "grad_norm": 0.0014046004507690668,
      "learning_rate": 0.2137709401040429,
      "loss": 0.2303,
      "num_input_tokens_seen": 12484368,
      "step": 14410
    },
    {
      "epoch": 6.796322489391796,
      "grad_norm": 0.001160518848337233,
      "learning_rate": 0.21371761876036,
      "loss": 0.2253,
      "num_input_tokens_seen": 12488832,
      "step": 14415
    },
    {
      "epoch": 6.798679867986799,
      "grad_norm": 0.0019423667108640075,
      "learning_rate": 0.21366428759061956,
      "loss": 0.2159,
      "num_input_tokens_seen": 12493440,
      "step": 14420
    },
    {
      "epoch": 6.801037246581801,
      "grad_norm": 0.0010687402682378888,
      "learning_rate": 0.2136109466030459,
      "loss": 0.2383,
      "num_input_tokens_seen": 12497264,
      "step": 14425
    },
    {
      "epoch": 6.803394625176804,
      "grad_norm": 0.0008811026928015053,
      "learning_rate": 0.2135575958058649,
      "loss": 0.2155,
      "num_input_tokens_seen": 12501872,
      "step": 14430
    },
    {
      "epoch": 6.805752003771806,
      "grad_norm": 0.0009930875385180116,
      "learning_rate": 0.2135042352073039,
      "loss": 0.2547,
      "num_input_tokens_seen": 12506080,
      "step": 14435
    },
    {
      "epoch": 6.808109382366808,
      "grad_norm": 0.0017960426630452275,
      "learning_rate": 0.2134508648155918,
      "loss": 0.2227,
      "num_input_tokens_seen": 12509536,
      "step": 14440
    },
    {
      "epoch": 6.810466760961811,
      "grad_norm": 0.003220172831788659,
      "learning_rate": 0.213397484638959,
      "loss": 0.2275,
      "num_input_tokens_seen": 12514624,
      "step": 14445
    },
    {
      "epoch": 6.812824139556813,
      "grad_norm": 0.0011971027124673128,
      "learning_rate": 0.21334409468563728,
      "loss": 0.2352,
      "num_input_tokens_seen": 12518320,
      "step": 14450
    },
    {
      "epoch": 6.815181518151816,
      "grad_norm": 0.0009361574775539339,
      "learning_rate": 0.2132906949638602,
      "loss": 0.2128,
      "num_input_tokens_seen": 12522800,
      "step": 14455
    },
    {
      "epoch": 6.817538896746818,
      "grad_norm": 0.0012258393689990044,
      "learning_rate": 0.21323728548186255,
      "loss": 0.2171,
      "num_input_tokens_seen": 12526992,
      "step": 14460
    },
    {
      "epoch": 6.8198962753418195,
      "grad_norm": 0.0015397552633658051,
      "learning_rate": 0.21318386624788088,
      "loss": 0.209,
      "num_input_tokens_seen": 12530784,
      "step": 14465
    },
    {
      "epoch": 6.822253653936822,
      "grad_norm": 0.003101245267316699,
      "learning_rate": 0.21313043727015288,
      "loss": 0.2666,
      "num_input_tokens_seen": 12535248,
      "step": 14470
    },
    {
      "epoch": 6.824611032531824,
      "grad_norm": 0.001084741554223001,
      "learning_rate": 0.2130769985569182,
      "loss": 0.1733,
      "num_input_tokens_seen": 12540160,
      "step": 14475
    },
    {
      "epoch": 6.826968411126827,
      "grad_norm": 0.003937000408768654,
      "learning_rate": 0.21302355011641766,
      "loss": 0.1941,
      "num_input_tokens_seen": 12544608,
      "step": 14480
    },
    {
      "epoch": 6.829325789721829,
      "grad_norm": 0.000859442341607064,
      "learning_rate": 0.21297009195689365,
      "loss": 0.1886,
      "num_input_tokens_seen": 12549920,
      "step": 14485
    },
    {
      "epoch": 6.8316831683168315,
      "grad_norm": 0.0006839822162874043,
      "learning_rate": 0.21291662408659015,
      "loss": 0.2425,
      "num_input_tokens_seen": 12554320,
      "step": 14490
    },
    {
      "epoch": 6.834040546911834,
      "grad_norm": 0.0012058347929269075,
      "learning_rate": 0.21286314651375254,
      "loss": 0.2444,
      "num_input_tokens_seen": 12558304,
      "step": 14495
    },
    {
      "epoch": 6.836397925506836,
      "grad_norm": 0.0018578370800241828,
      "learning_rate": 0.2128096592466278,
      "loss": 0.2431,
      "num_input_tokens_seen": 12562336,
      "step": 14500
    },
    {
      "epoch": 6.838755304101839,
      "grad_norm": 0.0011263216147199273,
      "learning_rate": 0.21275616229346428,
      "loss": 0.2368,
      "num_input_tokens_seen": 12565728,
      "step": 14505
    },
    {
      "epoch": 6.841112682696841,
      "grad_norm": 0.001029518898576498,
      "learning_rate": 0.21270265566251184,
      "loss": 0.2378,
      "num_input_tokens_seen": 12569440,
      "step": 14510
    },
    {
      "epoch": 6.8434700612918435,
      "grad_norm": 0.0009583405335433781,
      "learning_rate": 0.21264913936202193,
      "loss": 0.2369,
      "num_input_tokens_seen": 12573920,
      "step": 14515
    },
    {
      "epoch": 6.845827439886846,
      "grad_norm": 0.002080637263134122,
      "learning_rate": 0.2125956134002475,
      "loss": 0.2312,
      "num_input_tokens_seen": 12578464,
      "step": 14520
    },
    {
      "epoch": 6.848184818481848,
      "grad_norm": 0.000705792976077646,
      "learning_rate": 0.2125420777854428,
      "loss": 0.2279,
      "num_input_tokens_seen": 12582736,
      "step": 14525
    },
    {
      "epoch": 6.850542197076851,
      "grad_norm": 0.0007312246016226709,
      "learning_rate": 0.21248853252586372,
      "loss": 0.2287,
      "num_input_tokens_seen": 12587152,
      "step": 14530
    },
    {
      "epoch": 6.852899575671853,
      "grad_norm": 0.0019782166928052902,
      "learning_rate": 0.21243497762976774,
      "loss": 0.2218,
      "num_input_tokens_seen": 12591760,
      "step": 14535
    },
    {
      "epoch": 6.8552569542668556,
      "grad_norm": 0.0017789436969906092,
      "learning_rate": 0.21238141310541356,
      "loss": 0.2322,
      "num_input_tokens_seen": 12596448,
      "step": 14540
    },
    {
      "epoch": 6.857614332861858,
      "grad_norm": 0.0009594902512617409,
      "learning_rate": 0.21232783896106153,
      "loss": 0.2196,
      "num_input_tokens_seen": 12600496,
      "step": 14545
    },
    {
      "epoch": 6.85997171145686,
      "grad_norm": 0.0013743522576987743,
      "learning_rate": 0.21227425520497345,
      "loss": 0.2114,
      "num_input_tokens_seen": 12604448,
      "step": 14550
    },
    {
      "epoch": 6.862329090051862,
      "grad_norm": 0.00253375805914402,
      "learning_rate": 0.2122206618454127,
      "loss": 0.2769,
      "num_input_tokens_seen": 12608496,
      "step": 14555
    },
    {
      "epoch": 6.864686468646864,
      "grad_norm": 0.0019436494912952185,
      "learning_rate": 0.2121670588906439,
      "loss": 0.2222,
      "num_input_tokens_seen": 12612960,
      "step": 14560
    },
    {
      "epoch": 6.867043847241867,
      "grad_norm": 0.0023845082614570856,
      "learning_rate": 0.21211344634893345,
      "loss": 0.2205,
      "num_input_tokens_seen": 12617456,
      "step": 14565
    },
    {
      "epoch": 6.869401225836869,
      "grad_norm": 0.0019055824959650636,
      "learning_rate": 0.21205982422854897,
      "loss": 0.1972,
      "num_input_tokens_seen": 12621584,
      "step": 14570
    },
    {
      "epoch": 6.8717586044318715,
      "grad_norm": 0.001649558311328292,
      "learning_rate": 0.21200619253775974,
      "loss": 0.2459,
      "num_input_tokens_seen": 12626752,
      "step": 14575
    },
    {
      "epoch": 6.874115983026874,
      "grad_norm": 0.0012768738670274615,
      "learning_rate": 0.21195255128483637,
      "loss": 0.1846,
      "num_input_tokens_seen": 12631136,
      "step": 14580
    },
    {
      "epoch": 6.876473361621876,
      "grad_norm": 0.0010032659629359841,
      "learning_rate": 0.21189890047805102,
      "loss": 0.1697,
      "num_input_tokens_seen": 12636224,
      "step": 14585
    },
    {
      "epoch": 6.878830740216879,
      "grad_norm": 0.0031002480536699295,
      "learning_rate": 0.21184524012567735,
      "loss": 0.2841,
      "num_input_tokens_seen": 12640592,
      "step": 14590
    },
    {
      "epoch": 6.881188118811881,
      "grad_norm": 0.0016019722679629922,
      "learning_rate": 0.2117915702359905,
      "loss": 0.2598,
      "num_input_tokens_seen": 12644912,
      "step": 14595
    },
    {
      "epoch": 6.8835454974068835,
      "grad_norm": 0.001955955056473613,
      "learning_rate": 0.211737890817267,
      "loss": 0.2287,
      "num_input_tokens_seen": 12649424,
      "step": 14600
    },
    {
      "epoch": 6.8835454974068835,
      "eval_loss": 0.23954415321350098,
      "eval_runtime": 21.8956,
      "eval_samples_per_second": 43.068,
      "eval_steps_per_second": 21.557,
      "num_input_tokens_seen": 12649424,
      "step": 14600
    },
    {
      "epoch": 6.885902876001886,
      "grad_norm": 0.002885231049731374,
      "learning_rate": 0.21168420187778483,
      "loss": 0.2187,
      "num_input_tokens_seen": 12653056,
      "step": 14605
    },
    {
      "epoch": 6.888260254596888,
      "grad_norm": 0.0018564697820693254,
      "learning_rate": 0.21163050342582362,
      "loss": 0.2605,
      "num_input_tokens_seen": 12657536,
      "step": 14610
    },
    {
      "epoch": 6.890617633191891,
      "grad_norm": 0.001031252439133823,
      "learning_rate": 0.21157679546966426,
      "loss": 0.2262,
      "num_input_tokens_seen": 12661760,
      "step": 14615
    },
    {
      "epoch": 6.892975011786893,
      "grad_norm": 0.0008444297127425671,
      "learning_rate": 0.2115230780175892,
      "loss": 0.208,
      "num_input_tokens_seen": 12665888,
      "step": 14620
    },
    {
      "epoch": 6.8953323903818955,
      "grad_norm": 0.0014745905064046383,
      "learning_rate": 0.21146935107788237,
      "loss": 0.2147,
      "num_input_tokens_seen": 12669776,
      "step": 14625
    },
    {
      "epoch": 6.897689768976898,
      "grad_norm": 0.0013950886204838753,
      "learning_rate": 0.21141561465882916,
      "loss": 0.2675,
      "num_input_tokens_seen": 12674576,
      "step": 14630
    },
    {
      "epoch": 6.9000471475719,
      "grad_norm": 0.001465690671466291,
      "learning_rate": 0.21136186876871635,
      "loss": 0.2296,
      "num_input_tokens_seen": 12678592,
      "step": 14635
    },
    {
      "epoch": 6.902404526166903,
      "grad_norm": 0.0007786622154526412,
      "learning_rate": 0.21130811341583225,
      "loss": 0.2495,
      "num_input_tokens_seen": 12683296,
      "step": 14640
    },
    {
      "epoch": 6.904761904761905,
      "grad_norm": 0.0015234191669151187,
      "learning_rate": 0.21125434860846667,
      "loss": 0.2576,
      "num_input_tokens_seen": 12687728,
      "step": 14645
    },
    {
      "epoch": 6.9071192833569075,
      "grad_norm": 0.007685125805437565,
      "learning_rate": 0.2112005743549107,
      "loss": 0.2278,
      "num_input_tokens_seen": 12691616,
      "step": 14650
    },
    {
      "epoch": 6.90947666195191,
      "grad_norm": 0.0021657331380993128,
      "learning_rate": 0.21114679066345707,
      "loss": 0.2113,
      "num_input_tokens_seen": 12695456,
      "step": 14655
    },
    {
      "epoch": 6.911834040546912,
      "grad_norm": 0.0011278849560767412,
      "learning_rate": 0.21109299754239993,
      "loss": 0.1852,
      "num_input_tokens_seen": 12699312,
      "step": 14660
    },
    {
      "epoch": 6.914191419141914,
      "grad_norm": 0.00229400722309947,
      "learning_rate": 0.21103919500003482,
      "loss": 0.3198,
      "num_input_tokens_seen": 12704096,
      "step": 14665
    },
    {
      "epoch": 6.916548797736916,
      "grad_norm": 0.0010657842503860593,
      "learning_rate": 0.21098538304465872,
      "loss": 0.2192,
      "num_input_tokens_seen": 12708160,
      "step": 14670
    },
    {
      "epoch": 6.918906176331919,
      "grad_norm": 0.0022248895838856697,
      "learning_rate": 0.2109315616845702,
      "loss": 0.2231,
      "num_input_tokens_seen": 12712096,
      "step": 14675
    },
    {
      "epoch": 6.921263554926921,
      "grad_norm": 0.0015682694502174854,
      "learning_rate": 0.21087773092806925,
      "loss": 0.2298,
      "num_input_tokens_seen": 12716304,
      "step": 14680
    },
    {
      "epoch": 6.923620933521923,
      "grad_norm": 0.0032894080504775047,
      "learning_rate": 0.21082389078345704,
      "loss": 0.2376,
      "num_input_tokens_seen": 12719920,
      "step": 14685
    },
    {
      "epoch": 6.925978312116926,
      "grad_norm": 0.0013068506959825754,
      "learning_rate": 0.2107700412590365,
      "loss": 0.2202,
      "num_input_tokens_seen": 12724480,
      "step": 14690
    },
    {
      "epoch": 6.928335690711928,
      "grad_norm": 0.001223259954713285,
      "learning_rate": 0.210716182363112,
      "loss": 0.2388,
      "num_input_tokens_seen": 12728992,
      "step": 14695
    },
    {
      "epoch": 6.930693069306931,
      "grad_norm": 0.0009543129126541317,
      "learning_rate": 0.2106623141039891,
      "loss": 0.1979,
      "num_input_tokens_seen": 12733904,
      "step": 14700
    },
    {
      "epoch": 6.933050447901933,
      "grad_norm": 0.0010462512727826834,
      "learning_rate": 0.21060843648997507,
      "loss": 0.2761,
      "num_input_tokens_seen": 12738368,
      "step": 14705
    },
    {
      "epoch": 6.935407826496935,
      "grad_norm": 0.000838327337987721,
      "learning_rate": 0.21055454952937844,
      "loss": 0.21,
      "num_input_tokens_seen": 12742016,
      "step": 14710
    },
    {
      "epoch": 6.937765205091938,
      "grad_norm": 0.005613154731690884,
      "learning_rate": 0.21050065323050937,
      "loss": 0.2231,
      "num_input_tokens_seen": 12746576,
      "step": 14715
    },
    {
      "epoch": 6.94012258368694,
      "grad_norm": 0.001659718924202025,
      "learning_rate": 0.21044674760167928,
      "loss": 0.2214,
      "num_input_tokens_seen": 12750432,
      "step": 14720
    },
    {
      "epoch": 6.942479962281943,
      "grad_norm": 0.0026571345515549183,
      "learning_rate": 0.210392832651201,
      "loss": 0.246,
      "num_input_tokens_seen": 12755312,
      "step": 14725
    },
    {
      "epoch": 6.944837340876945,
      "grad_norm": 0.0043987394310534,
      "learning_rate": 0.210338908387389,
      "loss": 0.2334,
      "num_input_tokens_seen": 12759312,
      "step": 14730
    },
    {
      "epoch": 6.947194719471947,
      "grad_norm": 0.014044170267879963,
      "learning_rate": 0.21028497481855912,
      "loss": 0.2466,
      "num_input_tokens_seen": 12763600,
      "step": 14735
    },
    {
      "epoch": 6.94955209806695,
      "grad_norm": 0.0014765411615371704,
      "learning_rate": 0.21023103195302847,
      "loss": 0.211,
      "num_input_tokens_seen": 12768080,
      "step": 14740
    },
    {
      "epoch": 6.951909476661952,
      "grad_norm": 0.0023165519814938307,
      "learning_rate": 0.21017707979911582,
      "loss": 0.2077,
      "num_input_tokens_seen": 12772352,
      "step": 14745
    },
    {
      "epoch": 6.954266855256955,
      "grad_norm": 0.0016585185658186674,
      "learning_rate": 0.21012311836514122,
      "loss": 0.1738,
      "num_input_tokens_seen": 12776560,
      "step": 14750
    },
    {
      "epoch": 6.956624233851956,
      "grad_norm": 0.0016081148060038686,
      "learning_rate": 0.21006914765942622,
      "loss": 0.1522,
      "num_input_tokens_seen": 12780880,
      "step": 14755
    },
    {
      "epoch": 6.958981612446959,
      "grad_norm": 0.0034324100706726313,
      "learning_rate": 0.2100151676902938,
      "loss": 0.325,
      "num_input_tokens_seen": 12785216,
      "step": 14760
    },
    {
      "epoch": 6.961338991041961,
      "grad_norm": 0.013912621885538101,
      "learning_rate": 0.2099611784660683,
      "loss": 0.2483,
      "num_input_tokens_seen": 12789984,
      "step": 14765
    },
    {
      "epoch": 6.963696369636963,
      "grad_norm": 0.005421780981123447,
      "learning_rate": 0.20990717999507552,
      "loss": 0.2252,
      "num_input_tokens_seen": 12794688,
      "step": 14770
    },
    {
      "epoch": 6.966053748231966,
      "grad_norm": 0.005960391368716955,
      "learning_rate": 0.20985317228564276,
      "loss": 0.2282,
      "num_input_tokens_seen": 12799440,
      "step": 14775
    },
    {
      "epoch": 6.968411126826968,
      "grad_norm": 0.0022668817546218634,
      "learning_rate": 0.20979915534609872,
      "loss": 0.2377,
      "num_input_tokens_seen": 12803856,
      "step": 14780
    },
    {
      "epoch": 6.970768505421971,
      "grad_norm": 0.0021234306041151285,
      "learning_rate": 0.20974512918477342,
      "loss": 0.2199,
      "num_input_tokens_seen": 12808256,
      "step": 14785
    },
    {
      "epoch": 6.973125884016973,
      "grad_norm": 0.0023528514429926872,
      "learning_rate": 0.2096910938099984,
      "loss": 0.2269,
      "num_input_tokens_seen": 12812272,
      "step": 14790
    },
    {
      "epoch": 6.975483262611975,
      "grad_norm": 0.006343354471027851,
      "learning_rate": 0.2096370492301066,
      "loss": 0.2338,
      "num_input_tokens_seen": 12816640,
      "step": 14795
    },
    {
      "epoch": 6.977840641206978,
      "grad_norm": 0.00608060834929347,
      "learning_rate": 0.2095829954534323,
      "loss": 0.2388,
      "num_input_tokens_seen": 12821280,
      "step": 14800
    },
    {
      "epoch": 6.977840641206978,
      "eval_loss": 0.23311565816402435,
      "eval_runtime": 21.9006,
      "eval_samples_per_second": 43.058,
      "eval_steps_per_second": 21.552,
      "num_input_tokens_seen": 12821280,
      "step": 14800
    },
    {
      "epoch": 6.98019801980198,
      "grad_norm": 0.00808115117251873,
      "learning_rate": 0.2095289324883114,
      "loss": 0.2834,
      "num_input_tokens_seen": 12826720,
      "step": 14805
    },
    {
      "epoch": 6.982555398396983,
      "grad_norm": 0.0036127918865531683,
      "learning_rate": 0.20947486034308097,
      "loss": 0.2245,
      "num_input_tokens_seen": 12831536,
      "step": 14810
    },
    {
      "epoch": 6.984912776991985,
      "grad_norm": 0.0026020666118711233,
      "learning_rate": 0.2094207790260797,
      "loss": 0.2303,
      "num_input_tokens_seen": 12835536,
      "step": 14815
    },
    {
      "epoch": 6.987270155586987,
      "grad_norm": 0.005497944075614214,
      "learning_rate": 0.20936668854564758,
      "loss": 0.2206,
      "num_input_tokens_seen": 12839744,
      "step": 14820
    },
    {
      "epoch": 6.98962753418199,
      "grad_norm": 0.004506970755755901,
      "learning_rate": 0.20931258891012602,
      "loss": 0.2436,
      "num_input_tokens_seen": 12843856,
      "step": 14825
    },
    {
      "epoch": 6.991984912776992,
      "grad_norm": 0.005029045511037111,
      "learning_rate": 0.20925848012785792,
      "loss": 0.2386,
      "num_input_tokens_seen": 12848464,
      "step": 14830
    },
    {
      "epoch": 6.994342291371995,
      "grad_norm": 0.0025118188932538033,
      "learning_rate": 0.20920436220718747,
      "loss": 0.2251,
      "num_input_tokens_seen": 12852928,
      "step": 14835
    },
    {
      "epoch": 6.996699669966997,
      "grad_norm": 0.0025161406956613064,
      "learning_rate": 0.20915023515646033,
      "loss": 0.165,
      "num_input_tokens_seen": 12858096,
      "step": 14840
    },
    {
      "epoch": 6.999057048561999,
      "grad_norm": 0.0015647062100470066,
      "learning_rate": 0.20909609898402368,
      "loss": 0.123,
      "num_input_tokens_seen": 12861712,
      "step": 14845
    },
    {
      "epoch": 7.001414427157002,
      "grad_norm": 0.002299858955666423,
      "learning_rate": 0.2090419536982258,
      "loss": 0.2385,
      "num_input_tokens_seen": 12865632,
      "step": 14850
    },
    {
      "epoch": 7.003771805752004,
      "grad_norm": 0.00412387540563941,
      "learning_rate": 0.2089877993074168,
      "loss": 0.2773,
      "num_input_tokens_seen": 12870192,
      "step": 14855
    },
    {
      "epoch": 7.006129184347006,
      "grad_norm": 0.002483497140929103,
      "learning_rate": 0.20893363581994784,
      "loss": 0.2562,
      "num_input_tokens_seen": 12875152,
      "step": 14860
    },
    {
      "epoch": 7.008486562942008,
      "grad_norm": 0.0029971140902489424,
      "learning_rate": 0.2088794632441716,
      "loss": 0.2142,
      "num_input_tokens_seen": 12879376,
      "step": 14865
    },
    {
      "epoch": 7.0108439415370105,
      "grad_norm": 0.0053238715045154095,
      "learning_rate": 0.20882528158844219,
      "loss": 0.2217,
      "num_input_tokens_seen": 12884176,
      "step": 14870
    },
    {
      "epoch": 7.013201320132013,
      "grad_norm": 0.0021073271054774523,
      "learning_rate": 0.20877109086111514,
      "loss": 0.1911,
      "num_input_tokens_seen": 12889568,
      "step": 14875
    },
    {
      "epoch": 7.015558698727015,
      "grad_norm": 0.0026726997457444668,
      "learning_rate": 0.2087168910705473,
      "loss": 0.2268,
      "num_input_tokens_seen": 12893808,
      "step": 14880
    },
    {
      "epoch": 7.017916077322018,
      "grad_norm": 0.0016719794366508722,
      "learning_rate": 0.208662682225097,
      "loss": 0.2516,
      "num_input_tokens_seen": 12899184,
      "step": 14885
    },
    {
      "epoch": 7.02027345591702,
      "grad_norm": 0.0018244393868371844,
      "learning_rate": 0.2086084643331239,
      "loss": 0.2525,
      "num_input_tokens_seen": 12903568,
      "step": 14890
    },
    {
      "epoch": 7.0226308345120225,
      "grad_norm": 0.0029970731120556593,
      "learning_rate": 0.20855423740298906,
      "loss": 0.2298,
      "num_input_tokens_seen": 12907696,
      "step": 14895
    },
    {
      "epoch": 7.024988213107025,
      "grad_norm": 0.002714364090934396,
      "learning_rate": 0.208500001443055,
      "loss": 0.4703,
      "num_input_tokens_seen": 12912304,
      "step": 14900
    },
    {
      "epoch": 7.027345591702027,
      "grad_norm": 0.0015578607562929392,
      "learning_rate": 0.20844575646168553,
      "loss": 0.2211,
      "num_input_tokens_seen": 12916256,
      "step": 14905
    },
    {
      "epoch": 7.02970297029703,
      "grad_norm": 0.002424622653052211,
      "learning_rate": 0.20839150246724594,
      "loss": 0.2732,
      "num_input_tokens_seen": 12920992,
      "step": 14910
    },
    {
      "epoch": 7.032060348892032,
      "grad_norm": 0.0039414758794009686,
      "learning_rate": 0.20833723946810287,
      "loss": 0.2453,
      "num_input_tokens_seen": 12925024,
      "step": 14915
    },
    {
      "epoch": 7.0344177274870345,
      "grad_norm": 0.007846972905099392,
      "learning_rate": 0.20828296747262437,
      "loss": 0.236,
      "num_input_tokens_seen": 12929776,
      "step": 14920
    },
    {
      "epoch": 7.036775106082037,
      "grad_norm": 0.004184902645647526,
      "learning_rate": 0.20822868648917986,
      "loss": 0.2526,
      "num_input_tokens_seen": 12934096,
      "step": 14925
    },
    {
      "epoch": 7.039132484677039,
      "grad_norm": 0.0015606291126459837,
      "learning_rate": 0.20817439652614017,
      "loss": 0.2079,
      "num_input_tokens_seen": 12938080,
      "step": 14930
    },
    {
      "epoch": 7.041489863272042,
      "grad_norm": 0.002824156079441309,
      "learning_rate": 0.20812009759187744,
      "loss": 0.2248,
      "num_input_tokens_seen": 12941616,
      "step": 14935
    },
    {
      "epoch": 7.043847241867044,
      "grad_norm": 0.001625856733880937,
      "learning_rate": 0.2080657896947653,
      "loss": 0.2499,
      "num_input_tokens_seen": 12945440,
      "step": 14940
    },
    {
      "epoch": 7.0462046204620465,
      "grad_norm": 0.0014425426488742232,
      "learning_rate": 0.2080114728431787,
      "loss": 0.2102,
      "num_input_tokens_seen": 12949104,
      "step": 14945
    },
    {
      "epoch": 7.048561999057049,
      "grad_norm": 0.0017395182512700558,
      "learning_rate": 0.20795714704549392,
      "loss": 0.2073,
      "num_input_tokens_seen": 12953056,
      "step": 14950
    },
    {
      "epoch": 7.050919377652051,
      "grad_norm": 0.00194422691129148,
      "learning_rate": 0.20790281231008875,
      "loss": 0.2192,
      "num_input_tokens_seen": 12957632,
      "step": 14955
    },
    {
      "epoch": 7.053276756247053,
      "grad_norm": 0.0013639620738103986,
      "learning_rate": 0.20784846864534226,
      "loss": 0.2307,
      "num_input_tokens_seen": 12961408,
      "step": 14960
    },
    {
      "epoch": 7.055634134842055,
      "grad_norm": 0.00409960001707077,
      "learning_rate": 0.20779411605963496,
      "loss": 0.2073,
      "num_input_tokens_seen": 12966352,
      "step": 14965
    },
    {
      "epoch": 7.057991513437058,
      "grad_norm": 0.0013747570337727666,
      "learning_rate": 0.2077397545613487,
      "loss": 0.2359,
      "num_input_tokens_seen": 12970688,
      "step": 14970
    },
    {
      "epoch": 7.06034889203206,
      "grad_norm": 0.0012778989039361477,
      "learning_rate": 0.20768538415886661,
      "loss": 0.2199,
      "num_input_tokens_seen": 12974736,
      "step": 14975
    },
    {
      "epoch": 7.0627062706270625,
      "grad_norm": 0.0011710308026522398,
      "learning_rate": 0.20763100486057343,
      "loss": 0.2265,
      "num_input_tokens_seen": 12979712,
      "step": 14980
    },
    {
      "epoch": 7.065063649222065,
      "grad_norm": 0.0013039769837632775,
      "learning_rate": 0.20757661667485502,
      "loss": 0.2273,
      "num_input_tokens_seen": 12983488,
      "step": 14985
    },
    {
      "epoch": 7.067421027817067,
      "grad_norm": 0.0015102979959920049,
      "learning_rate": 0.2075222196100988,
      "loss": 0.2378,
      "num_input_tokens_seen": 12987936,
      "step": 14990
    },
    {
      "epoch": 7.06977840641207,
      "grad_norm": 0.0020068527664989233,
      "learning_rate": 0.20746781367469344,
      "loss": 0.2424,
      "num_input_tokens_seen": 12992608,
      "step": 14995
    },
    {
      "epoch": 7.072135785007072,
      "grad_norm": 0.002226723125204444,
      "learning_rate": 0.207413398877029,
      "loss": 0.2421,
      "num_input_tokens_seen": 12996208,
      "step": 15000
    },
    {
      "epoch": 7.072135785007072,
      "eval_loss": 0.2265898436307907,
      "eval_runtime": 21.9566,
      "eval_samples_per_second": 42.948,
      "eval_steps_per_second": 21.497,
      "num_input_tokens_seen": 12996208,
      "step": 15000
    },
    {
      "epoch": 7.0744931636020745,
      "grad_norm": 0.002201343420892954,
      "learning_rate": 0.20735897522549698,
      "loss": 0.2196,
      "num_input_tokens_seen": 13000144,
      "step": 15005
    },
    {
      "epoch": 7.076850542197077,
      "grad_norm": 0.002531293546780944,
      "learning_rate": 0.2073045427284902,
      "loss": 0.2313,
      "num_input_tokens_seen": 13004544,
      "step": 15010
    },
    {
      "epoch": 7.079207920792079,
      "grad_norm": 0.0018848689505830407,
      "learning_rate": 0.2072501013944027,
      "loss": 0.196,
      "num_input_tokens_seen": 13009456,
      "step": 15015
    },
    {
      "epoch": 7.081565299387082,
      "grad_norm": 0.0008991016657091677,
      "learning_rate": 0.20719565123163017,
      "loss": 0.1895,
      "num_input_tokens_seen": 13014432,
      "step": 15020
    },
    {
      "epoch": 7.083922677982084,
      "grad_norm": 0.003975799772888422,
      "learning_rate": 0.20714119224856944,
      "loss": 0.276,
      "num_input_tokens_seen": 13018912,
      "step": 15025
    },
    {
      "epoch": 7.0862800565770865,
      "grad_norm": 0.0010439081816002727,
      "learning_rate": 0.2070867244536188,
      "loss": 0.2686,
      "num_input_tokens_seen": 13024064,
      "step": 15030
    },
    {
      "epoch": 7.088637435172089,
      "grad_norm": 0.0022074102889746428,
      "learning_rate": 0.20703224785517785,
      "loss": 0.229,
      "num_input_tokens_seen": 13028016,
      "step": 15035
    },
    {
      "epoch": 7.090994813767091,
      "grad_norm": 0.0019474757136777043,
      "learning_rate": 0.20697776246164754,
      "loss": 0.2308,
      "num_input_tokens_seen": 13032672,
      "step": 15040
    },
    {
      "epoch": 7.093352192362094,
      "grad_norm": 0.0017951992340385914,
      "learning_rate": 0.2069232682814303,
      "loss": 0.2393,
      "num_input_tokens_seen": 13036544,
      "step": 15045
    },
    {
      "epoch": 7.095709570957096,
      "grad_norm": 0.0015423892764374614,
      "learning_rate": 0.20686876532292972,
      "loss": 0.227,
      "num_input_tokens_seen": 13040464,
      "step": 15050
    },
    {
      "epoch": 7.0980669495520985,
      "grad_norm": 0.001461912994273007,
      "learning_rate": 0.20681425359455083,
      "loss": 0.2193,
      "num_input_tokens_seen": 13044368,
      "step": 15055
    },
    {
      "epoch": 7.1004243281471,
      "grad_norm": 0.0014030665624886751,
      "learning_rate": 0.20675973310470008,
      "loss": 0.258,
      "num_input_tokens_seen": 13048928,
      "step": 15060
    },
    {
      "epoch": 7.102781706742102,
      "grad_norm": 0.0014632578240707517,
      "learning_rate": 0.2067052038617852,
      "loss": 0.2541,
      "num_input_tokens_seen": 13053024,
      "step": 15065
    },
    {
      "epoch": 7.105139085337105,
      "grad_norm": 0.0026318735908716917,
      "learning_rate": 0.2066506658742153,
      "loss": 0.237,
      "num_input_tokens_seen": 13057888,
      "step": 15070
    },
    {
      "epoch": 7.107496463932107,
      "grad_norm": 0.002468305639922619,
      "learning_rate": 0.20659611915040077,
      "loss": 0.2297,
      "num_input_tokens_seen": 13061984,
      "step": 15075
    },
    {
      "epoch": 7.10985384252711,
      "grad_norm": 0.002756178379058838,
      "learning_rate": 0.20654156369875348,
      "loss": 0.2328,
      "num_input_tokens_seen": 13066496,
      "step": 15080
    },
    {
      "epoch": 7.112211221122112,
      "grad_norm": 0.0028739015106111765,
      "learning_rate": 0.20648699952768648,
      "loss": 0.2524,
      "num_input_tokens_seen": 13070288,
      "step": 15085
    },
    {
      "epoch": 7.114568599717114,
      "grad_norm": 0.0043462240137159824,
      "learning_rate": 0.20643242664561437,
      "loss": 0.2334,
      "num_input_tokens_seen": 13074800,
      "step": 15090
    },
    {
      "epoch": 7.116925978312117,
      "grad_norm": 0.001739034429192543,
      "learning_rate": 0.20637784506095277,
      "loss": 0.2069,
      "num_input_tokens_seen": 13079808,
      "step": 15095
    },
    {
      "epoch": 7.119283356907119,
      "grad_norm": 0.001652404898777604,
      "learning_rate": 0.20632325478211908,
      "loss": 0.2215,
      "num_input_tokens_seen": 13083824,
      "step": 15100
    },
    {
      "epoch": 7.121640735502122,
      "grad_norm": 0.0013313082745298743,
      "learning_rate": 0.20626865581753165,
      "loss": 0.2037,
      "num_input_tokens_seen": 13088368,
      "step": 15105
    },
    {
      "epoch": 7.123998114097124,
      "grad_norm": 0.0029345087241381407,
      "learning_rate": 0.2062140481756104,
      "loss": 0.2222,
      "num_input_tokens_seen": 13092624,
      "step": 15110
    },
    {
      "epoch": 7.126355492692126,
      "grad_norm": 0.0025416959542781115,
      "learning_rate": 0.20615943186477648,
      "loss": 0.254,
      "num_input_tokens_seen": 13096400,
      "step": 15115
    },
    {
      "epoch": 7.128712871287129,
      "grad_norm": 0.003997765481472015,
      "learning_rate": 0.20610480689345242,
      "loss": 0.2425,
      "num_input_tokens_seen": 13100448,
      "step": 15120
    },
    {
      "epoch": 7.131070249882131,
      "grad_norm": 0.002323502441868186,
      "learning_rate": 0.2060501732700621,
      "loss": 0.2367,
      "num_input_tokens_seen": 13105232,
      "step": 15125
    },
    {
      "epoch": 7.133427628477134,
      "grad_norm": 0.0020868442952632904,
      "learning_rate": 0.20599553100303067,
      "loss": 0.2295,
      "num_input_tokens_seen": 13110208,
      "step": 15130
    },
    {
      "epoch": 7.135785007072136,
      "grad_norm": 0.001666162395849824,
      "learning_rate": 0.20594088010078465,
      "loss": 0.2222,
      "num_input_tokens_seen": 13114800,
      "step": 15135
    },
    {
      "epoch": 7.138142385667138,
      "grad_norm": 0.00163518195040524,
      "learning_rate": 0.20588622057175196,
      "loss": 0.1929,
      "num_input_tokens_seen": 13119792,
      "step": 15140
    },
    {
      "epoch": 7.140499764262141,
      "grad_norm": 0.0010642927372828126,
      "learning_rate": 0.20583155242436177,
      "loss": 0.2236,
      "num_input_tokens_seen": 13123840,
      "step": 15145
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 0.0034376857802271843,
      "learning_rate": 0.20577687566704453,
      "loss": 0.2836,
      "num_input_tokens_seen": 13127760,
      "step": 15150
    },
    {
      "epoch": 7.145214521452146,
      "grad_norm": 0.0010964744724333286,
      "learning_rate": 0.20572219030823213,
      "loss": 0.2193,
      "num_input_tokens_seen": 13133360,
      "step": 15155
    },
    {
      "epoch": 7.147571900047147,
      "grad_norm": 0.0014615522231906652,
      "learning_rate": 0.20566749635635775,
      "loss": 0.2151,
      "num_input_tokens_seen": 13137392,
      "step": 15160
    },
    {
      "epoch": 7.1499292786421496,
      "grad_norm": 0.001081032445654273,
      "learning_rate": 0.20561279381985587,
      "loss": 0.2387,
      "num_input_tokens_seen": 13142160,
      "step": 15165
    },
    {
      "epoch": 7.152286657237152,
      "grad_norm": 0.0021379422396421432,
      "learning_rate": 0.2055580827071623,
      "loss": 0.2357,
      "num_input_tokens_seen": 13148080,
      "step": 15170
    },
    {
      "epoch": 7.154644035832154,
      "grad_norm": 0.0010404668282717466,
      "learning_rate": 0.20550336302671418,
      "loss": 0.2334,
      "num_input_tokens_seen": 13151632,
      "step": 15175
    },
    {
      "epoch": 7.157001414427157,
      "grad_norm": 0.0011815376346930861,
      "learning_rate": 0.20544863478695,
      "loss": 0.2374,
      "num_input_tokens_seen": 13155424,
      "step": 15180
    },
    {
      "epoch": 7.159358793022159,
      "grad_norm": 0.0014085306320339441,
      "learning_rate": 0.20539389799630953,
      "loss": 0.2236,
      "num_input_tokens_seen": 13160064,
      "step": 15185
    },
    {
      "epoch": 7.161716171617162,
      "grad_norm": 0.0038801319897174835,
      "learning_rate": 0.20533915266323388,
      "loss": 0.2464,
      "num_input_tokens_seen": 13164576,
      "step": 15190
    },
    {
      "epoch": 7.164073550212164,
      "grad_norm": 0.0031731300987303257,
      "learning_rate": 0.20528439879616542,
      "loss": 0.2606,
      "num_input_tokens_seen": 13168512,
      "step": 15195
    },
    {
      "epoch": 7.166430928807166,
      "grad_norm": 0.0011300339829176664,
      "learning_rate": 0.20522963640354794,
      "loss": 0.2458,
      "num_input_tokens_seen": 13172592,
      "step": 15200
    },
    {
      "epoch": 7.166430928807166,
      "eval_loss": 0.24147500097751617,
      "eval_runtime": 21.9191,
      "eval_samples_per_second": 43.022,
      "eval_steps_per_second": 21.534,
      "num_input_tokens_seen": 13172592,
      "step": 15200
    },
    {
      "epoch": 7.168788307402169,
      "grad_norm": 0.0028294268995523453,
      "learning_rate": 0.20517486549382644,
      "loss": 0.2612,
      "num_input_tokens_seen": 13177312,
      "step": 15205
    },
    {
      "epoch": 7.171145685997171,
      "grad_norm": 0.0022586060222238302,
      "learning_rate": 0.20512008607544735,
      "loss": 0.232,
      "num_input_tokens_seen": 13181936,
      "step": 15210
    },
    {
      "epoch": 7.173503064592174,
      "grad_norm": 0.0018090792000293732,
      "learning_rate": 0.20506529815685826,
      "loss": 0.2198,
      "num_input_tokens_seen": 13186176,
      "step": 15215
    },
    {
      "epoch": 7.175860443187176,
      "grad_norm": 0.0023372580762952566,
      "learning_rate": 0.2050105017465082,
      "loss": 0.2493,
      "num_input_tokens_seen": 13189872,
      "step": 15220
    },
    {
      "epoch": 7.178217821782178,
      "grad_norm": 0.0014842526288703084,
      "learning_rate": 0.20495569685284754,
      "loss": 0.2412,
      "num_input_tokens_seen": 13194272,
      "step": 15225
    },
    {
      "epoch": 7.180575200377181,
      "grad_norm": 0.0014635034603998065,
      "learning_rate": 0.20490088348432778,
      "loss": 0.2303,
      "num_input_tokens_seen": 13198128,
      "step": 15230
    },
    {
      "epoch": 7.182932578972183,
      "grad_norm": 0.0013976584887132049,
      "learning_rate": 0.2048460616494018,
      "loss": 0.2173,
      "num_input_tokens_seen": 13203216,
      "step": 15235
    },
    {
      "epoch": 7.185289957567186,
      "grad_norm": 0.002899093320593238,
      "learning_rate": 0.2047912313565239,
      "loss": 0.2193,
      "num_input_tokens_seen": 13207744,
      "step": 15240
    },
    {
      "epoch": 7.187647336162188,
      "grad_norm": 0.0010404656641185284,
      "learning_rate": 0.20473639261414958,
      "loss": 0.2126,
      "num_input_tokens_seen": 13211904,
      "step": 15245
    },
    {
      "epoch": 7.19000471475719,
      "grad_norm": 0.0026291217654943466,
      "learning_rate": 0.2046815454307357,
      "loss": 0.2498,
      "num_input_tokens_seen": 13216208,
      "step": 15250
    },
    {
      "epoch": 7.192362093352193,
      "grad_norm": 0.001881828997284174,
      "learning_rate": 0.20462668981474028,
      "loss": 0.2451,
      "num_input_tokens_seen": 13220704,
      "step": 15255
    },
    {
      "epoch": 7.194719471947194,
      "grad_norm": 0.0011119935661554337,
      "learning_rate": 0.20457182577462288,
      "loss": 0.2155,
      "num_input_tokens_seen": 13225136,
      "step": 15260
    },
    {
      "epoch": 7.197076850542197,
      "grad_norm": 0.001778302830643952,
      "learning_rate": 0.2045169533188441,
      "loss": 0.2327,
      "num_input_tokens_seen": 13229264,
      "step": 15265
    },
    {
      "epoch": 7.199434229137199,
      "grad_norm": 0.0009732965263538063,
      "learning_rate": 0.20446207245586603,
      "loss": 0.237,
      "num_input_tokens_seen": 13233040,
      "step": 15270
    },
    {
      "epoch": 7.2017916077322015,
      "grad_norm": 0.0016485468950122595,
      "learning_rate": 0.20440718319415196,
      "loss": 0.233,
      "num_input_tokens_seen": 13237344,
      "step": 15275
    },
    {
      "epoch": 7.204148986327204,
      "grad_norm": 0.0018658199114724994,
      "learning_rate": 0.20435228554216653,
      "loss": 0.2385,
      "num_input_tokens_seen": 13242640,
      "step": 15280
    },
    {
      "epoch": 7.206506364922206,
      "grad_norm": 0.0020858540665358305,
      "learning_rate": 0.20429737950837565,
      "loss": 0.2291,
      "num_input_tokens_seen": 13247008,
      "step": 15285
    },
    {
      "epoch": 7.208863743517209,
      "grad_norm": 0.0015889480710029602,
      "learning_rate": 0.20424246510124647,
      "loss": 0.1571,
      "num_input_tokens_seen": 13251072,
      "step": 15290
    },
    {
      "epoch": 7.211221122112211,
      "grad_norm": 0.0006653518066741526,
      "learning_rate": 0.20418754232924755,
      "loss": 0.24,
      "num_input_tokens_seen": 13254960,
      "step": 15295
    },
    {
      "epoch": 7.2135785007072135,
      "grad_norm": 0.001337501802481711,
      "learning_rate": 0.20413261120084863,
      "loss": 0.2148,
      "num_input_tokens_seen": 13258992,
      "step": 15300
    },
    {
      "epoch": 7.215935879302216,
      "grad_norm": 0.0014349065022543073,
      "learning_rate": 0.2040776717245208,
      "loss": 0.2111,
      "num_input_tokens_seen": 13262880,
      "step": 15305
    },
    {
      "epoch": 7.218293257897218,
      "grad_norm": 0.000719567877240479,
      "learning_rate": 0.2040227239087364,
      "loss": 0.266,
      "num_input_tokens_seen": 13267456,
      "step": 15310
    },
    {
      "epoch": 7.220650636492221,
      "grad_norm": 0.0014087497256696224,
      "learning_rate": 0.20396776776196904,
      "loss": 0.2092,
      "num_input_tokens_seen": 13271776,
      "step": 15315
    },
    {
      "epoch": 7.223008015087223,
      "grad_norm": 0.0010972082382068038,
      "learning_rate": 0.20391280329269373,
      "loss": 0.2005,
      "num_input_tokens_seen": 13275856,
      "step": 15320
    },
    {
      "epoch": 7.2253653936822255,
      "grad_norm": 0.0009513216791674495,
      "learning_rate": 0.20385783050938663,
      "loss": 0.2003,
      "num_input_tokens_seen": 13279872,
      "step": 15325
    },
    {
      "epoch": 7.227722772277228,
      "grad_norm": 0.0010587811702862382,
      "learning_rate": 0.20380284942052526,
      "loss": 0.2362,
      "num_input_tokens_seen": 13283968,
      "step": 15330
    },
    {
      "epoch": 7.23008015087223,
      "grad_norm": 0.0017593813827261329,
      "learning_rate": 0.2037478600345884,
      "loss": 0.2135,
      "num_input_tokens_seen": 13288160,
      "step": 15335
    },
    {
      "epoch": 7.232437529467233,
      "grad_norm": 0.0010429348330944777,
      "learning_rate": 0.20369286236005604,
      "loss": 0.2352,
      "num_input_tokens_seen": 13292944,
      "step": 15340
    },
    {
      "epoch": 7.234794908062235,
      "grad_norm": 0.0012294008629396558,
      "learning_rate": 0.20363785640540957,
      "loss": 0.2075,
      "num_input_tokens_seen": 13296864,
      "step": 15345
    },
    {
      "epoch": 7.2371522866572375,
      "grad_norm": 0.0008983141160570085,
      "learning_rate": 0.2035828421791316,
      "loss": 0.2271,
      "num_input_tokens_seen": 13301744,
      "step": 15350
    },
    {
      "epoch": 7.23950966525224,
      "grad_norm": 0.0015812971396371722,
      "learning_rate": 0.20352781968970599,
      "loss": 0.1815,
      "num_input_tokens_seen": 13305488,
      "step": 15355
    },
    {
      "epoch": 7.2418670438472414,
      "grad_norm": 0.0013623128179460764,
      "learning_rate": 0.2034727889456179,
      "loss": 0.1957,
      "num_input_tokens_seen": 13310352,
      "step": 15360
    },
    {
      "epoch": 7.244224422442244,
      "grad_norm": 0.00042585001210682094,
      "learning_rate": 0.2034177499553538,
      "loss": 0.1887,
      "num_input_tokens_seen": 13314096,
      "step": 15365
    },
    {
      "epoch": 7.246581801037246,
      "grad_norm": 0.0015833373181521893,
      "learning_rate": 0.2033627027274014,
      "loss": 0.1882,
      "num_input_tokens_seen": 13318064,
      "step": 15370
    },
    {
      "epoch": 7.248939179632249,
      "grad_norm": 0.001885665813460946,
      "learning_rate": 0.20330764727024955,
      "loss": 0.2861,
      "num_input_tokens_seen": 13322304,
      "step": 15375
    },
    {
      "epoch": 7.251296558227251,
      "grad_norm": 0.0012640377972275019,
      "learning_rate": 0.20325258359238868,
      "loss": 0.2839,
      "num_input_tokens_seen": 13326688,
      "step": 15380
    },
    {
      "epoch": 7.2536539368222535,
      "grad_norm": 0.001171878888271749,
      "learning_rate": 0.20319751170231018,
      "loss": 0.2491,
      "num_input_tokens_seen": 13330816,
      "step": 15385
    },
    {
      "epoch": 7.256011315417256,
      "grad_norm": 0.0014559588162228465,
      "learning_rate": 0.2031424316085068,
      "loss": 0.2344,
      "num_input_tokens_seen": 13335232,
      "step": 15390
    },
    {
      "epoch": 7.258368694012258,
      "grad_norm": 0.001987034222111106,
      "learning_rate": 0.20308734331947265,
      "loss": 0.2842,
      "num_input_tokens_seen": 13339120,
      "step": 15395
    },
    {
      "epoch": 7.260726072607261,
      "grad_norm": 0.001524219405837357,
      "learning_rate": 0.20303224684370305,
      "loss": 0.2424,
      "num_input_tokens_seen": 13342864,
      "step": 15400
    },
    {
      "epoch": 7.260726072607261,
      "eval_loss": 0.23568521440029144,
      "eval_runtime": 21.8976,
      "eval_samples_per_second": 43.064,
      "eval_steps_per_second": 21.555,
      "num_input_tokens_seen": 13342864,
      "step": 15400
    },
    {
      "epoch": 7.263083451202263,
      "grad_norm": 0.0008001226815395057,
      "learning_rate": 0.20297714218969456,
      "loss": 0.2263,
      "num_input_tokens_seen": 13347184,
      "step": 15405
    },
    {
      "epoch": 7.2654408297972655,
      "grad_norm": 0.001295391470193863,
      "learning_rate": 0.20292202936594497,
      "loss": 0.2194,
      "num_input_tokens_seen": 13352176,
      "step": 15410
    },
    {
      "epoch": 7.267798208392268,
      "grad_norm": 0.00221432582475245,
      "learning_rate": 0.2028669083809534,
      "loss": 0.2495,
      "num_input_tokens_seen": 13356480,
      "step": 15415
    },
    {
      "epoch": 7.27015558698727,
      "grad_norm": 0.0011023107217624784,
      "learning_rate": 0.20281177924322016,
      "loss": 0.2179,
      "num_input_tokens_seen": 13360720,
      "step": 15420
    },
    {
      "epoch": 7.272512965582273,
      "grad_norm": 0.0017586336471140385,
      "learning_rate": 0.2027566419612469,
      "loss": 0.3001,
      "num_input_tokens_seen": 13365552,
      "step": 15425
    },
    {
      "epoch": 7.274870344177275,
      "grad_norm": 0.0010329211363568902,
      "learning_rate": 0.20270149654353647,
      "loss": 0.2358,
      "num_input_tokens_seen": 13369600,
      "step": 15430
    },
    {
      "epoch": 7.2772277227722775,
      "grad_norm": 0.002331542782485485,
      "learning_rate": 0.202646342998593,
      "loss": 0.2265,
      "num_input_tokens_seen": 13374576,
      "step": 15435
    },
    {
      "epoch": 7.27958510136728,
      "grad_norm": 0.0010810852982103825,
      "learning_rate": 0.20259118133492185,
      "loss": 0.2259,
      "num_input_tokens_seen": 13379056,
      "step": 15440
    },
    {
      "epoch": 7.281942479962282,
      "grad_norm": 0.0010518874041736126,
      "learning_rate": 0.20253601156102966,
      "loss": 0.232,
      "num_input_tokens_seen": 13384272,
      "step": 15445
    },
    {
      "epoch": 7.284299858557285,
      "grad_norm": 0.0010362756438553333,
      "learning_rate": 0.20248083368542422,
      "loss": 0.2115,
      "num_input_tokens_seen": 13387632,
      "step": 15450
    },
    {
      "epoch": 7.286657237152287,
      "grad_norm": 0.001184705994091928,
      "learning_rate": 0.2024256477166147,
      "loss": 0.2454,
      "num_input_tokens_seen": 13391408,
      "step": 15455
    },
    {
      "epoch": 7.2890146157472895,
      "grad_norm": 0.0011438924120739102,
      "learning_rate": 0.2023704536631115,
      "loss": 0.2011,
      "num_input_tokens_seen": 13396240,
      "step": 15460
    },
    {
      "epoch": 7.291371994342291,
      "grad_norm": 0.0011968269245699048,
      "learning_rate": 0.20231525153342625,
      "loss": 0.2455,
      "num_input_tokens_seen": 13400336,
      "step": 15465
    },
    {
      "epoch": 7.293729372937293,
      "grad_norm": 0.0008323107031174004,
      "learning_rate": 0.20226004133607173,
      "loss": 0.2113,
      "num_input_tokens_seen": 13404384,
      "step": 15470
    },
    {
      "epoch": 7.296086751532296,
      "grad_norm": 0.0011118457186967134,
      "learning_rate": 0.20220482307956214,
      "loss": 0.2453,
      "num_input_tokens_seen": 13408640,
      "step": 15475
    },
    {
      "epoch": 7.298444130127298,
      "grad_norm": 0.0008762665092945099,
      "learning_rate": 0.20214959677241276,
      "loss": 0.2246,
      "num_input_tokens_seen": 13412528,
      "step": 15480
    },
    {
      "epoch": 7.300801508722301,
      "grad_norm": 0.0017892989562824368,
      "learning_rate": 0.20209436242314022,
      "loss": 0.2393,
      "num_input_tokens_seen": 13416960,
      "step": 15485
    },
    {
      "epoch": 7.303158887317303,
      "grad_norm": 0.0017204215982928872,
      "learning_rate": 0.2020391200402623,
      "loss": 0.2256,
      "num_input_tokens_seen": 13421824,
      "step": 15490
    },
    {
      "epoch": 7.305516265912305,
      "grad_norm": 0.0021427995525300503,
      "learning_rate": 0.2019838696322981,
      "loss": 0.2207,
      "num_input_tokens_seen": 13426240,
      "step": 15495
    },
    {
      "epoch": 7.307873644507308,
      "grad_norm": 0.0010970188304781914,
      "learning_rate": 0.20192861120776798,
      "loss": 0.2153,
      "num_input_tokens_seen": 13430144,
      "step": 15500
    },
    {
      "epoch": 7.31023102310231,
      "grad_norm": 0.001152031240053475,
      "learning_rate": 0.20187334477519345,
      "loss": 0.2277,
      "num_input_tokens_seen": 13434208,
      "step": 15505
    },
    {
      "epoch": 7.312588401697313,
      "grad_norm": 0.0013564972905442119,
      "learning_rate": 0.20181807034309726,
      "loss": 0.2107,
      "num_input_tokens_seen": 13438896,
      "step": 15510
    },
    {
      "epoch": 7.314945780292315,
      "grad_norm": 0.0020762740168720484,
      "learning_rate": 0.2017627879200034,
      "loss": 0.2171,
      "num_input_tokens_seen": 13443152,
      "step": 15515
    },
    {
      "epoch": 7.317303158887317,
      "grad_norm": 0.0009688197751529515,
      "learning_rate": 0.2017074975144372,
      "loss": 0.2026,
      "num_input_tokens_seen": 13447216,
      "step": 15520
    },
    {
      "epoch": 7.31966053748232,
      "grad_norm": 0.0009530129027552903,
      "learning_rate": 0.20165219913492508,
      "loss": 0.2567,
      "num_input_tokens_seen": 13451360,
      "step": 15525
    },
    {
      "epoch": 7.322017916077322,
      "grad_norm": 0.0011105960002169013,
      "learning_rate": 0.20159689278999468,
      "loss": 0.2694,
      "num_input_tokens_seen": 13455584,
      "step": 15530
    },
    {
      "epoch": 7.324375294672325,
      "grad_norm": 0.0020682020112872124,
      "learning_rate": 0.20154157848817508,
      "loss": 0.2153,
      "num_input_tokens_seen": 13460160,
      "step": 15535
    },
    {
      "epoch": 7.326732673267327,
      "grad_norm": 0.00088716228492558,
      "learning_rate": 0.20148625623799632,
      "loss": 0.2431,
      "num_input_tokens_seen": 13463728,
      "step": 15540
    },
    {
      "epoch": 7.329090051862329,
      "grad_norm": 0.0007348372600972652,
      "learning_rate": 0.20143092604798984,
      "loss": 0.2407,
      "num_input_tokens_seen": 13468480,
      "step": 15545
    },
    {
      "epoch": 7.331447430457332,
      "grad_norm": 0.0008155219256877899,
      "learning_rate": 0.2013755879266883,
      "loss": 0.231,
      "num_input_tokens_seen": 13472608,
      "step": 15550
    },
    {
      "epoch": 7.333804809052333,
      "grad_norm": 0.0011456632055342197,
      "learning_rate": 0.20132024188262543,
      "loss": 0.2046,
      "num_input_tokens_seen": 13476944,
      "step": 15555
    },
    {
      "epoch": 7.336162187647336,
      "grad_norm": 0.0012141769984737039,
      "learning_rate": 0.2012648879243363,
      "loss": 0.2225,
      "num_input_tokens_seen": 13482192,
      "step": 15560
    },
    {
      "epoch": 7.338519566242338,
      "grad_norm": 0.0013456086162477732,
      "learning_rate": 0.20120952606035725,
      "loss": 0.2131,
      "num_input_tokens_seen": 13486128,
      "step": 15565
    },
    {
      "epoch": 7.3408769448373405,
      "grad_norm": 0.005772563628852367,
      "learning_rate": 0.20115415629922576,
      "loss": 0.1939,
      "num_input_tokens_seen": 13491056,
      "step": 15570
    },
    {
      "epoch": 7.343234323432343,
      "grad_norm": 0.001241033896803856,
      "learning_rate": 0.20109877864948048,
      "loss": 0.1871,
      "num_input_tokens_seen": 13494736,
      "step": 15575
    },
    {
      "epoch": 7.345591702027345,
      "grad_norm": 0.0014223781181499362,
      "learning_rate": 0.20104339311966138,
      "loss": 0.2944,
      "num_input_tokens_seen": 13498848,
      "step": 15580
    },
    {
      "epoch": 7.347949080622348,
      "grad_norm": 0.0012310738675296307,
      "learning_rate": 0.2009879997183097,
      "loss": 0.1879,
      "num_input_tokens_seen": 13502752,
      "step": 15585
    },
    {
      "epoch": 7.35030645921735,
      "grad_norm": 0.0038612382486462593,
      "learning_rate": 0.20093259845396763,
      "loss": 0.22,
      "num_input_tokens_seen": 13507296,
      "step": 15590
    },
    {
      "epoch": 7.3526638378123526,
      "grad_norm": 0.0010143116815015674,
      "learning_rate": 0.20087718933517884,
      "loss": 0.1975,
      "num_input_tokens_seen": 13511584,
      "step": 15595
    },
    {
      "epoch": 7.355021216407355,
      "grad_norm": 0.003364418400451541,
      "learning_rate": 0.20082177237048807,
      "loss": 0.2588,
      "num_input_tokens_seen": 13515600,
      "step": 15600
    },
    {
      "epoch": 7.355021216407355,
      "eval_loss": 0.2185654193162918,
      "eval_runtime": 21.9802,
      "eval_samples_per_second": 42.902,
      "eval_steps_per_second": 21.474,
      "num_input_tokens_seen": 13515600,
      "step": 15600
    },
    {
      "epoch": 7.357378595002357,
      "grad_norm": 0.0025536210741847754,
      "learning_rate": 0.20076634756844133,
      "loss": 0.207,
      "num_input_tokens_seen": 13520000,
      "step": 15605
    },
    {
      "epoch": 7.35973597359736,
      "grad_norm": 0.002411969704553485,
      "learning_rate": 0.20071091493758586,
      "loss": 0.2388,
      "num_input_tokens_seen": 13523840,
      "step": 15610
    },
    {
      "epoch": 7.362093352192362,
      "grad_norm": 0.00144630775321275,
      "learning_rate": 0.20065547448647003,
      "loss": 0.2317,
      "num_input_tokens_seen": 13528448,
      "step": 15615
    },
    {
      "epoch": 7.364450730787365,
      "grad_norm": 0.0011975698871538043,
      "learning_rate": 0.20060002622364348,
      "loss": 0.2207,
      "num_input_tokens_seen": 13533104,
      "step": 15620
    },
    {
      "epoch": 7.366808109382367,
      "grad_norm": 0.0012333836639299989,
      "learning_rate": 0.20054457015765695,
      "loss": 0.2094,
      "num_input_tokens_seen": 13537808,
      "step": 15625
    },
    {
      "epoch": 7.369165487977369,
      "grad_norm": 0.0016573378816246986,
      "learning_rate": 0.20048910629706254,
      "loss": 0.2351,
      "num_input_tokens_seen": 13541952,
      "step": 15630
    },
    {
      "epoch": 7.371522866572372,
      "grad_norm": 0.0017191787483170629,
      "learning_rate": 0.20043363465041347,
      "loss": 0.2182,
      "num_input_tokens_seen": 13546208,
      "step": 15635
    },
    {
      "epoch": 7.373880245167374,
      "grad_norm": 0.0035955270286649466,
      "learning_rate": 0.2003781552262641,
      "loss": 0.2804,
      "num_input_tokens_seen": 13550480,
      "step": 15640
    },
    {
      "epoch": 7.376237623762377,
      "grad_norm": 0.0011233547702431679,
      "learning_rate": 0.20032266803317014,
      "loss": 0.1973,
      "num_input_tokens_seen": 13554752,
      "step": 15645
    },
    {
      "epoch": 7.378595002357379,
      "grad_norm": 0.001515724929049611,
      "learning_rate": 0.2002671730796884,
      "loss": 0.2504,
      "num_input_tokens_seen": 13559696,
      "step": 15650
    },
    {
      "epoch": 7.380952380952381,
      "grad_norm": 0.0017262190813198686,
      "learning_rate": 0.20021167037437684,
      "loss": 0.2337,
      "num_input_tokens_seen": 13563856,
      "step": 15655
    },
    {
      "epoch": 7.383309759547384,
      "grad_norm": 0.0013095950707793236,
      "learning_rate": 0.20015615992579472,
      "loss": 0.2322,
      "num_input_tokens_seen": 13567760,
      "step": 15660
    },
    {
      "epoch": 7.385667138142385,
      "grad_norm": 0.00241126143373549,
      "learning_rate": 0.20010064174250244,
      "loss": 0.2137,
      "num_input_tokens_seen": 13572624,
      "step": 15665
    },
    {
      "epoch": 7.388024516737388,
      "grad_norm": 0.0012708373833447695,
      "learning_rate": 0.2000451158330616,
      "loss": 0.2077,
      "num_input_tokens_seen": 13577088,
      "step": 15670
    },
    {
      "epoch": 7.39038189533239,
      "grad_norm": 0.0013766729971393943,
      "learning_rate": 0.199989582206035,
      "loss": 0.2414,
      "num_input_tokens_seen": 13580704,
      "step": 15675
    },
    {
      "epoch": 7.3927392739273925,
      "grad_norm": 0.0015739204827696085,
      "learning_rate": 0.1999340408699866,
      "loss": 0.1947,
      "num_input_tokens_seen": 13584752,
      "step": 15680
    },
    {
      "epoch": 7.395096652522395,
      "grad_norm": 0.0011760651832446456,
      "learning_rate": 0.19987849183348155,
      "loss": 0.2395,
      "num_input_tokens_seen": 13589504,
      "step": 15685
    },
    {
      "epoch": 7.397454031117397,
      "grad_norm": 0.0019183579133823514,
      "learning_rate": 0.19982293510508628,
      "loss": 0.2495,
      "num_input_tokens_seen": 13593312,
      "step": 15690
    },
    {
      "epoch": 7.3998114097124,
      "grad_norm": 0.001145180081948638,
      "learning_rate": 0.19976737069336833,
      "loss": 0.2,
      "num_input_tokens_seen": 13597968,
      "step": 15695
    },
    {
      "epoch": 7.402168788307402,
      "grad_norm": 0.001067962497472763,
      "learning_rate": 0.1997117986068964,
      "loss": 0.2107,
      "num_input_tokens_seen": 13602448,
      "step": 15700
    },
    {
      "epoch": 7.4045261669024045,
      "grad_norm": 0.0013458176981657743,
      "learning_rate": 0.19965621885424037,
      "loss": 0.2008,
      "num_input_tokens_seen": 13606912,
      "step": 15705
    },
    {
      "epoch": 7.406883545497407,
      "grad_norm": 0.0009260313236154616,
      "learning_rate": 0.19960063144397142,
      "loss": 0.2474,
      "num_input_tokens_seen": 13610944,
      "step": 15710
    },
    {
      "epoch": 7.409240924092409,
      "grad_norm": 0.0012564401840791106,
      "learning_rate": 0.19954503638466176,
      "loss": 0.2496,
      "num_input_tokens_seen": 13615184,
      "step": 15715
    },
    {
      "epoch": 7.411598302687412,
      "grad_norm": 0.0012379857944324613,
      "learning_rate": 0.1994894336848848,
      "loss": 0.2025,
      "num_input_tokens_seen": 13619440,
      "step": 15720
    },
    {
      "epoch": 7.413955681282414,
      "grad_norm": 0.0009774168720468879,
      "learning_rate": 0.1994338233532153,
      "loss": 0.2172,
      "num_input_tokens_seen": 13623152,
      "step": 15725
    },
    {
      "epoch": 7.4163130598774165,
      "grad_norm": 0.0011851703748106956,
      "learning_rate": 0.19937820539822904,
      "loss": 0.2465,
      "num_input_tokens_seen": 13627936,
      "step": 15730
    },
    {
      "epoch": 7.418670438472419,
      "grad_norm": 0.0018155736615881324,
      "learning_rate": 0.199322579828503,
      "loss": 0.2547,
      "num_input_tokens_seen": 13631968,
      "step": 15735
    },
    {
      "epoch": 7.421027817067421,
      "grad_norm": 0.0025571875739842653,
      "learning_rate": 0.19926694665261527,
      "loss": 0.2201,
      "num_input_tokens_seen": 13636576,
      "step": 15740
    },
    {
      "epoch": 7.423385195662424,
      "grad_norm": 0.0005860735545866191,
      "learning_rate": 0.19921130587914526,
      "loss": 0.2325,
      "num_input_tokens_seen": 13640688,
      "step": 15745
    },
    {
      "epoch": 7.425742574257426,
      "grad_norm": 0.0009464659378863871,
      "learning_rate": 0.19915565751667344,
      "loss": 0.2276,
      "num_input_tokens_seen": 13644768,
      "step": 15750
    },
    {
      "epoch": 7.428099952852428,
      "grad_norm": 0.0015194168081507087,
      "learning_rate": 0.19910000157378152,
      "loss": 0.1962,
      "num_input_tokens_seen": 13649040,
      "step": 15755
    },
    {
      "epoch": 7.43045733144743,
      "grad_norm": 0.0005551839130930603,
      "learning_rate": 0.1990443380590523,
      "loss": 0.1946,
      "num_input_tokens_seen": 13653344,
      "step": 15760
    },
    {
      "epoch": 7.432814710042432,
      "grad_norm": 0.0019931152928620577,
      "learning_rate": 0.19898866698106984,
      "loss": 0.2267,
      "num_input_tokens_seen": 13657696,
      "step": 15765
    },
    {
      "epoch": 7.435172088637435,
      "grad_norm": 0.000997345196083188,
      "learning_rate": 0.19893298834841933,
      "loss": 0.1653,
      "num_input_tokens_seen": 13661696,
      "step": 15770
    },
    {
      "epoch": 7.437529467232437,
      "grad_norm": 0.0004948952118866146,
      "learning_rate": 0.19887730216968705,
      "loss": 0.2821,
      "num_input_tokens_seen": 13666240,
      "step": 15775
    },
    {
      "epoch": 7.43988684582744,
      "grad_norm": 0.0011417437344789505,
      "learning_rate": 0.19882160845346053,
      "loss": 0.1867,
      "num_input_tokens_seen": 13670944,
      "step": 15780
    },
    {
      "epoch": 7.442244224422442,
      "grad_norm": 0.0009144588257186115,
      "learning_rate": 0.1987659072083285,
      "loss": 0.176,
      "num_input_tokens_seen": 13676320,
      "step": 15785
    },
    {
      "epoch": 7.4446016030174444,
      "grad_norm": 0.0012812716886401176,
      "learning_rate": 0.1987101984428807,
      "loss": 0.2066,
      "num_input_tokens_seen": 13680272,
      "step": 15790
    },
    {
      "epoch": 7.446958981612447,
      "grad_norm": 0.0007047458202578127,
      "learning_rate": 0.19865448216570822,
      "loss": 0.1903,
      "num_input_tokens_seen": 13684880,
      "step": 15795
    },
    {
      "epoch": 7.449316360207449,
      "grad_norm": 0.003035178640857339,
      "learning_rate": 0.19859875838540317,
      "loss": 0.2608,
      "num_input_tokens_seen": 13688640,
      "step": 15800
    },
    {
      "epoch": 7.449316360207449,
      "eval_loss": 0.2198134958744049,
      "eval_runtime": 21.921,
      "eval_samples_per_second": 43.018,
      "eval_steps_per_second": 21.532,
      "num_input_tokens_seen": 13688640,
      "step": 15800
    },
    {
      "epoch": 7.451673738802452,
      "grad_norm": 0.0019790607038885355,
      "learning_rate": 0.1985430271105588,
      "loss": 0.2624,
      "num_input_tokens_seen": 13693616,
      "step": 15805
    },
    {
      "epoch": 7.454031117397454,
      "grad_norm": 0.0008906322764232755,
      "learning_rate": 0.19848728834976961,
      "loss": 0.211,
      "num_input_tokens_seen": 13697392,
      "step": 15810
    },
    {
      "epoch": 7.4563884959924565,
      "grad_norm": 0.0006739745731465518,
      "learning_rate": 0.19843154211163128,
      "loss": 0.2187,
      "num_input_tokens_seen": 13702080,
      "step": 15815
    },
    {
      "epoch": 7.458745874587459,
      "grad_norm": 0.0010772119276225567,
      "learning_rate": 0.1983757884047405,
      "loss": 0.2297,
      "num_input_tokens_seen": 13706192,
      "step": 15820
    },
    {
      "epoch": 7.461103253182461,
      "grad_norm": 0.0006713903858326375,
      "learning_rate": 0.1983200272376952,
      "loss": 0.2225,
      "num_input_tokens_seen": 13710528,
      "step": 15825
    },
    {
      "epoch": 7.463460631777464,
      "grad_norm": 0.001854804577305913,
      "learning_rate": 0.1982642586190945,
      "loss": 0.2074,
      "num_input_tokens_seen": 13714736,
      "step": 15830
    },
    {
      "epoch": 7.465818010372466,
      "grad_norm": 0.0017706318758428097,
      "learning_rate": 0.1982084825575386,
      "loss": 0.1985,
      "num_input_tokens_seen": 13719360,
      "step": 15835
    },
    {
      "epoch": 7.4681753889674685,
      "grad_norm": 0.0009043480968102813,
      "learning_rate": 0.19815269906162883,
      "loss": 0.1822,
      "num_input_tokens_seen": 13722864,
      "step": 15840
    },
    {
      "epoch": 7.470532767562471,
      "grad_norm": 0.0007522740634158254,
      "learning_rate": 0.19809690813996775,
      "loss": 0.2615,
      "num_input_tokens_seen": 13727424,
      "step": 15845
    },
    {
      "epoch": 7.472890146157473,
      "grad_norm": 0.0025366076733917,
      "learning_rate": 0.19804110980115905,
      "loss": 0.2044,
      "num_input_tokens_seen": 13730912,
      "step": 15850
    },
    {
      "epoch": 7.475247524752476,
      "grad_norm": 0.00226388918235898,
      "learning_rate": 0.19798530405380746,
      "loss": 0.2744,
      "num_input_tokens_seen": 13734720,
      "step": 15855
    },
    {
      "epoch": 7.477604903347478,
      "grad_norm": 0.0008240643655881286,
      "learning_rate": 0.19792949090651893,
      "loss": 0.2225,
      "num_input_tokens_seen": 13739520,
      "step": 15860
    },
    {
      "epoch": 7.47996228194248,
      "grad_norm": 0.001074794097803533,
      "learning_rate": 0.19787367036790066,
      "loss": 0.2317,
      "num_input_tokens_seen": 13743504,
      "step": 15865
    },
    {
      "epoch": 7.482319660537482,
      "grad_norm": 0.001268785330466926,
      "learning_rate": 0.19781784244656075,
      "loss": 0.2216,
      "num_input_tokens_seen": 13747216,
      "step": 15870
    },
    {
      "epoch": 7.484677039132484,
      "grad_norm": 0.0009768782183527946,
      "learning_rate": 0.19776200715110864,
      "loss": 0.2474,
      "num_input_tokens_seen": 13751872,
      "step": 15875
    },
    {
      "epoch": 7.487034417727487,
      "grad_norm": 0.001726972870528698,
      "learning_rate": 0.1977061644901548,
      "loss": 0.214,
      "num_input_tokens_seen": 13756848,
      "step": 15880
    },
    {
      "epoch": 7.489391796322489,
      "grad_norm": 0.0008257489535026252,
      "learning_rate": 0.1976503144723109,
      "loss": 0.2281,
      "num_input_tokens_seen": 13761408,
      "step": 15885
    },
    {
      "epoch": 7.491749174917492,
      "grad_norm": 0.0006839131820015609,
      "learning_rate": 0.19759445710618967,
      "loss": 0.2039,
      "num_input_tokens_seen": 13766096,
      "step": 15890
    },
    {
      "epoch": 7.494106553512494,
      "grad_norm": 0.0008757683099247515,
      "learning_rate": 0.19753859240040508,
      "loss": 0.2019,
      "num_input_tokens_seen": 13770896,
      "step": 15895
    },
    {
      "epoch": 7.496463932107496,
      "grad_norm": 0.002707946812734008,
      "learning_rate": 0.1974827203635721,
      "loss": 0.2262,
      "num_input_tokens_seen": 13776048,
      "step": 15900
    },
    {
      "epoch": 7.498821310702499,
      "grad_norm": 0.0009542276966385543,
      "learning_rate": 0.19742684100430694,
      "loss": 0.2519,
      "num_input_tokens_seen": 13780752,
      "step": 15905
    },
    {
      "epoch": 7.501178689297501,
      "grad_norm": 0.0009286348940804601,
      "learning_rate": 0.19737095433122692,
      "loss": 0.224,
      "num_input_tokens_seen": 13784832,
      "step": 15910
    },
    {
      "epoch": 7.503536067892504,
      "grad_norm": 0.001171562122181058,
      "learning_rate": 0.19731506035295046,
      "loss": 0.2312,
      "num_input_tokens_seen": 13789328,
      "step": 15915
    },
    {
      "epoch": 7.505893446487506,
      "grad_norm": 0.002538565546274185,
      "learning_rate": 0.19725915907809702,
      "loss": 0.2013,
      "num_input_tokens_seen": 13793504,
      "step": 15920
    },
    {
      "epoch": 7.508250825082508,
      "grad_norm": 0.0006675312179140747,
      "learning_rate": 0.1972032505152874,
      "loss": 0.2489,
      "num_input_tokens_seen": 13797968,
      "step": 15925
    },
    {
      "epoch": 7.510608203677511,
      "grad_norm": 0.0006826460012234747,
      "learning_rate": 0.19714733467314338,
      "loss": 0.1986,
      "num_input_tokens_seen": 13801472,
      "step": 15930
    },
    {
      "epoch": 7.512965582272513,
      "grad_norm": 0.0006211880245245993,
      "learning_rate": 0.19709141156028784,
      "loss": 0.2176,
      "num_input_tokens_seen": 13806832,
      "step": 15935
    },
    {
      "epoch": 7.515322960867516,
      "grad_norm": 0.0012805040460079908,
      "learning_rate": 0.1970354811853448,
      "loss": 0.2419,
      "num_input_tokens_seen": 13810896,
      "step": 15940
    },
    {
      "epoch": 7.517680339462518,
      "grad_norm": 0.0010723763843998313,
      "learning_rate": 0.19697954355693953,
      "loss": 0.2325,
      "num_input_tokens_seen": 13815856,
      "step": 15945
    },
    {
      "epoch": 7.52003771805752,
      "grad_norm": 0.0007440208573825657,
      "learning_rate": 0.19692359868369827,
      "loss": 0.2164,
      "num_input_tokens_seen": 13819680,
      "step": 15950
    },
    {
      "epoch": 7.522395096652522,
      "grad_norm": 0.0011183223687112331,
      "learning_rate": 0.1968676465742484,
      "loss": 0.2108,
      "num_input_tokens_seen": 13823984,
      "step": 15955
    },
    {
      "epoch": 7.524752475247524,
      "grad_norm": 0.0007324377074837685,
      "learning_rate": 0.19681168723721845,
      "loss": 0.2269,
      "num_input_tokens_seen": 13828256,
      "step": 15960
    },
    {
      "epoch": 7.527109853842527,
      "grad_norm": 0.001567443017847836,
      "learning_rate": 0.19675572068123803,
      "loss": 0.2077,
      "num_input_tokens_seen": 13832112,
      "step": 15965
    },
    {
      "epoch": 7.529467232437529,
      "grad_norm": 0.0010483977384865284,
      "learning_rate": 0.19669974691493794,
      "loss": 0.2402,
      "num_input_tokens_seen": 13836928,
      "step": 15970
    },
    {
      "epoch": 7.5318246110325315,
      "grad_norm": 0.0009805227164179087,
      "learning_rate": 0.19664376594695002,
      "loss": 0.2074,
      "num_input_tokens_seen": 13841232,
      "step": 15975
    },
    {
      "epoch": 7.534181989627534,
      "grad_norm": 0.001306479680351913,
      "learning_rate": 0.19658777778590722,
      "loss": 0.2383,
      "num_input_tokens_seen": 13845056,
      "step": 15980
    },
    {
      "epoch": 7.536539368222536,
      "grad_norm": 0.0009496709099039435,
      "learning_rate": 0.19653178244044364,
      "loss": 0.2252,
      "num_input_tokens_seen": 13849648,
      "step": 15985
    },
    {
      "epoch": 7.538896746817539,
      "grad_norm": 0.001598716713488102,
      "learning_rate": 0.19647577991919443,
      "loss": 0.1804,
      "num_input_tokens_seen": 13853632,
      "step": 15990
    },
    {
      "epoch": 7.541254125412541,
      "grad_norm": 0.0010123283136636019,
      "learning_rate": 0.1964197702307959,
      "loss": 0.2027,
      "num_input_tokens_seen": 13858288,
      "step": 15995
    },
    {
      "epoch": 7.5436115040075435,
      "grad_norm": 0.0007701338035985827,
      "learning_rate": 0.19636375338388545,
      "loss": 0.2053,
      "num_input_tokens_seen": 13863312,
      "step": 16000
    },
    {
      "epoch": 7.5436115040075435,
      "eval_loss": 0.2214398831129074,
      "eval_runtime": 21.9276,
      "eval_samples_per_second": 43.005,
      "eval_steps_per_second": 21.525,
      "num_input_tokens_seen": 13863312,
      "step": 16000
    },
    {
      "epoch": 7.545968882602546,
      "grad_norm": 0.0023055970668792725,
      "learning_rate": 0.1963077293871016,
      "loss": 0.2474,
      "num_input_tokens_seen": 13867280,
      "step": 16005
    },
    {
      "epoch": 7.548326261197548,
      "grad_norm": 0.004393108654767275,
      "learning_rate": 0.19625169824908395,
      "loss": 0.238,
      "num_input_tokens_seen": 13871360,
      "step": 16010
    },
    {
      "epoch": 7.550683639792551,
      "grad_norm": 0.0009342250414192677,
      "learning_rate": 0.19619565997847319,
      "loss": 0.1977,
      "num_input_tokens_seen": 13876032,
      "step": 16015
    },
    {
      "epoch": 7.553041018387553,
      "grad_norm": 0.0018847247119992971,
      "learning_rate": 0.19613961458391113,
      "loss": 0.2091,
      "num_input_tokens_seen": 13880784,
      "step": 16020
    },
    {
      "epoch": 7.5553983969825556,
      "grad_norm": 0.0009684693068265915,
      "learning_rate": 0.19608356207404065,
      "loss": 0.2366,
      "num_input_tokens_seen": 13885280,
      "step": 16025
    },
    {
      "epoch": 7.557755775577558,
      "grad_norm": 0.0016907795798033476,
      "learning_rate": 0.1960275024575058,
      "loss": 0.2043,
      "num_input_tokens_seen": 13889328,
      "step": 16030
    },
    {
      "epoch": 7.56011315417256,
      "grad_norm": 0.000994437257759273,
      "learning_rate": 0.19597143574295164,
      "loss": 0.224,
      "num_input_tokens_seen": 13894256,
      "step": 16035
    },
    {
      "epoch": 7.562470532767563,
      "grad_norm": 0.0011292974231764674,
      "learning_rate": 0.1959153619390244,
      "loss": 0.2052,
      "num_input_tokens_seen": 13898560,
      "step": 16040
    },
    {
      "epoch": 7.564827911362565,
      "grad_norm": 0.0009519473533146083,
      "learning_rate": 0.1958592810543713,
      "loss": 0.2854,
      "num_input_tokens_seen": 13902592,
      "step": 16045
    },
    {
      "epoch": 7.567185289957568,
      "grad_norm": 0.000994957285001874,
      "learning_rate": 0.19580319309764077,
      "loss": 0.2145,
      "num_input_tokens_seen": 13906464,
      "step": 16050
    },
    {
      "epoch": 7.56954266855257,
      "grad_norm": 0.001000052085146308,
      "learning_rate": 0.1957470980774823,
      "loss": 0.2248,
      "num_input_tokens_seen": 13910752,
      "step": 16055
    },
    {
      "epoch": 7.571900047147572,
      "grad_norm": 0.000700485659763217,
      "learning_rate": 0.19569099600254639,
      "loss": 0.2326,
      "num_input_tokens_seen": 13915088,
      "step": 16060
    },
    {
      "epoch": 7.574257425742574,
      "grad_norm": 0.0007162366528064013,
      "learning_rate": 0.1956348868814847,
      "loss": 0.2381,
      "num_input_tokens_seen": 13919104,
      "step": 16065
    },
    {
      "epoch": 7.576614804337576,
      "grad_norm": 0.0011094091460108757,
      "learning_rate": 0.19557877072295,
      "loss": 0.2274,
      "num_input_tokens_seen": 13923392,
      "step": 16070
    },
    {
      "epoch": 7.578972182932579,
      "grad_norm": 0.0016657671658322215,
      "learning_rate": 0.19552264753559603,
      "loss": 0.2037,
      "num_input_tokens_seen": 13928240,
      "step": 16075
    },
    {
      "epoch": 7.581329561527581,
      "grad_norm": 0.0008856115746311843,
      "learning_rate": 0.19546651732807774,
      "loss": 0.2425,
      "num_input_tokens_seen": 13932464,
      "step": 16080
    },
    {
      "epoch": 7.5836869401225835,
      "grad_norm": 0.0009673985769040883,
      "learning_rate": 0.19541038010905112,
      "loss": 0.1742,
      "num_input_tokens_seen": 13936272,
      "step": 16085
    },
    {
      "epoch": 7.586044318717586,
      "grad_norm": 0.0012668371200561523,
      "learning_rate": 0.19535423588717324,
      "loss": 0.1836,
      "num_input_tokens_seen": 13940352,
      "step": 16090
    },
    {
      "epoch": 7.588401697312588,
      "grad_norm": 0.0004977318458259106,
      "learning_rate": 0.19529808467110224,
      "loss": 0.231,
      "num_input_tokens_seen": 13945376,
      "step": 16095
    },
    {
      "epoch": 7.590759075907591,
      "grad_norm": 0.0012938473373651505,
      "learning_rate": 0.19524192646949734,
      "loss": 0.2881,
      "num_input_tokens_seen": 13949488,
      "step": 16100
    },
    {
      "epoch": 7.593116454502593,
      "grad_norm": 0.0020714097190648317,
      "learning_rate": 0.19518576129101878,
      "loss": 0.1997,
      "num_input_tokens_seen": 13954256,
      "step": 16105
    },
    {
      "epoch": 7.5954738330975955,
      "grad_norm": 0.0017766216769814491,
      "learning_rate": 0.19512958914432804,
      "loss": 0.2246,
      "num_input_tokens_seen": 13958832,
      "step": 16110
    },
    {
      "epoch": 7.597831211692598,
      "grad_norm": 0.0012715080520138144,
      "learning_rate": 0.1950734100380875,
      "loss": 0.2115,
      "num_input_tokens_seen": 13963008,
      "step": 16115
    },
    {
      "epoch": 7.6001885902876,
      "grad_norm": 0.0006806178134866059,
      "learning_rate": 0.19501722398096066,
      "loss": 0.231,
      "num_input_tokens_seen": 13966944,
      "step": 16120
    },
    {
      "epoch": 7.602545968882603,
      "grad_norm": 0.00136559980455786,
      "learning_rate": 0.1949610309816122,
      "loss": 0.2319,
      "num_input_tokens_seen": 13971520,
      "step": 16125
    },
    {
      "epoch": 7.604903347477605,
      "grad_norm": 0.0012694124598056078,
      "learning_rate": 0.1949048310487078,
      "loss": 0.2285,
      "num_input_tokens_seen": 13975888,
      "step": 16130
    },
    {
      "epoch": 7.6072607260726075,
      "grad_norm": 0.0020746146328747272,
      "learning_rate": 0.19484862419091406,
      "loss": 0.2305,
      "num_input_tokens_seen": 13980304,
      "step": 16135
    },
    {
      "epoch": 7.60961810466761,
      "grad_norm": 0.0007621843833476305,
      "learning_rate": 0.19479241041689893,
      "loss": 0.2349,
      "num_input_tokens_seen": 13984016,
      "step": 16140
    },
    {
      "epoch": 7.611975483262612,
      "grad_norm": 0.0008407479617744684,
      "learning_rate": 0.19473618973533116,
      "loss": 0.2098,
      "num_input_tokens_seen": 13988464,
      "step": 16145
    },
    {
      "epoch": 7.614332861857615,
      "grad_norm": 0.000796615902800113,
      "learning_rate": 0.19467996215488076,
      "loss": 0.238,
      "num_input_tokens_seen": 13992208,
      "step": 16150
    },
    {
      "epoch": 7.616690240452616,
      "grad_norm": 0.0007784464978612959,
      "learning_rate": 0.1946237276842187,
      "loss": 0.2682,
      "num_input_tokens_seen": 13996688,
      "step": 16155
    },
    {
      "epoch": 7.619047619047619,
      "grad_norm": 0.0009017623378895223,
      "learning_rate": 0.19456748633201712,
      "loss": 0.2144,
      "num_input_tokens_seen": 14001056,
      "step": 16160
    },
    {
      "epoch": 7.621404997642621,
      "grad_norm": 0.0015838656108826399,
      "learning_rate": 0.194511238106949,
      "loss": 0.2081,
      "num_input_tokens_seen": 14004368,
      "step": 16165
    },
    {
      "epoch": 7.623762376237623,
      "grad_norm": 0.000657330674584955,
      "learning_rate": 0.19445498301768863,
      "loss": 0.2182,
      "num_input_tokens_seen": 14007936,
      "step": 16170
    },
    {
      "epoch": 7.626119754832626,
      "grad_norm": 0.0010585157433524728,
      "learning_rate": 0.19439872107291126,
      "loss": 0.2192,
      "num_input_tokens_seen": 14012112,
      "step": 16175
    },
    {
      "epoch": 7.628477133427628,
      "grad_norm": 0.0013791456585749984,
      "learning_rate": 0.1943424522812931,
      "loss": 0.244,
      "num_input_tokens_seen": 14016240,
      "step": 16180
    },
    {
      "epoch": 7.630834512022631,
      "grad_norm": 0.0025927203241735697,
      "learning_rate": 0.19428617665151157,
      "loss": 0.2397,
      "num_input_tokens_seen": 14020448,
      "step": 16185
    },
    {
      "epoch": 7.633191890617633,
      "grad_norm": 0.00101269013248384,
      "learning_rate": 0.19422989419224507,
      "loss": 0.2037,
      "num_input_tokens_seen": 14024864,
      "step": 16190
    },
    {
      "epoch": 7.635549269212635,
      "grad_norm": 0.0005833413451910019,
      "learning_rate": 0.19417360491217303,
      "loss": 0.2104,
      "num_input_tokens_seen": 14029184,
      "step": 16195
    },
    {
      "epoch": 7.637906647807638,
      "grad_norm": 0.0019800260197371244,
      "learning_rate": 0.19411730881997605,
      "loss": 0.231,
      "num_input_tokens_seen": 14032992,
      "step": 16200
    },
    {
      "epoch": 7.637906647807638,
      "eval_loss": 0.21739938855171204,
      "eval_runtime": 21.9077,
      "eval_samples_per_second": 43.044,
      "eval_steps_per_second": 21.545,
      "num_input_tokens_seen": 14032992,
      "step": 16200
    },
    {
      "epoch": 7.64026402640264,
      "grad_norm": 0.0008662063046358526,
      "learning_rate": 0.1940610059243356,
      "loss": 0.2459,
      "num_input_tokens_seen": 14037728,
      "step": 16205
    },
    {
      "epoch": 7.642621404997643,
      "grad_norm": 0.0009490397642366588,
      "learning_rate": 0.19400469623393435,
      "loss": 0.2189,
      "num_input_tokens_seen": 14042192,
      "step": 16210
    },
    {
      "epoch": 7.644978783592645,
      "grad_norm": 0.0008216191199608147,
      "learning_rate": 0.1939483797574559,
      "loss": 0.2072,
      "num_input_tokens_seen": 14046144,
      "step": 16215
    },
    {
      "epoch": 7.6473361621876474,
      "grad_norm": 0.000831150624435395,
      "learning_rate": 0.19389205650358504,
      "loss": 0.2304,
      "num_input_tokens_seen": 14050208,
      "step": 16220
    },
    {
      "epoch": 7.64969354078265,
      "grad_norm": 0.000904123648069799,
      "learning_rate": 0.19383572648100747,
      "loss": 0.2043,
      "num_input_tokens_seen": 14054880,
      "step": 16225
    },
    {
      "epoch": 7.652050919377652,
      "grad_norm": 0.0007262252038344741,
      "learning_rate": 0.19377938969841,
      "loss": 0.2359,
      "num_input_tokens_seen": 14059600,
      "step": 16230
    },
    {
      "epoch": 7.654408297972655,
      "grad_norm": 0.0005963325966149569,
      "learning_rate": 0.1937230461644805,
      "loss": 0.272,
      "num_input_tokens_seen": 14062976,
      "step": 16235
    },
    {
      "epoch": 7.656765676567657,
      "grad_norm": 0.0012865022290498018,
      "learning_rate": 0.19366669588790777,
      "loss": 0.2124,
      "num_input_tokens_seen": 14067120,
      "step": 16240
    },
    {
      "epoch": 7.6591230551626595,
      "grad_norm": 0.0008523368742316961,
      "learning_rate": 0.19361033887738185,
      "loss": 0.1985,
      "num_input_tokens_seen": 14071696,
      "step": 16245
    },
    {
      "epoch": 7.661480433757662,
      "grad_norm": 0.0017915060743689537,
      "learning_rate": 0.19355397514159361,
      "loss": 0.1894,
      "num_input_tokens_seen": 14075760,
      "step": 16250
    },
    {
      "epoch": 7.663837812352664,
      "grad_norm": 0.0009128367528319359,
      "learning_rate": 0.19349760468923508,
      "loss": 0.2233,
      "num_input_tokens_seen": 14080016,
      "step": 16255
    },
    {
      "epoch": 7.666195190947667,
      "grad_norm": 0.0020780207123607397,
      "learning_rate": 0.19344122752899925,
      "loss": 0.2391,
      "num_input_tokens_seen": 14085360,
      "step": 16260
    },
    {
      "epoch": 7.668552569542668,
      "grad_norm": 0.0012987582013010979,
      "learning_rate": 0.1933848436695802,
      "loss": 0.2128,
      "num_input_tokens_seen": 14089280,
      "step": 16265
    },
    {
      "epoch": 7.670909948137671,
      "grad_norm": 0.0007208376773633063,
      "learning_rate": 0.1933284531196731,
      "loss": 0.269,
      "num_input_tokens_seen": 14093728,
      "step": 16270
    },
    {
      "epoch": 7.673267326732673,
      "grad_norm": 0.0008002311224117875,
      "learning_rate": 0.19327205588797403,
      "loss": 0.2114,
      "num_input_tokens_seen": 14098384,
      "step": 16275
    },
    {
      "epoch": 7.675624705327675,
      "grad_norm": 0.0007173403864726424,
      "learning_rate": 0.19321565198318014,
      "loss": 0.2356,
      "num_input_tokens_seen": 14102880,
      "step": 16280
    },
    {
      "epoch": 7.677982083922678,
      "grad_norm": 0.0017524423310533166,
      "learning_rate": 0.1931592414139896,
      "loss": 0.2199,
      "num_input_tokens_seen": 14107184,
      "step": 16285
    },
    {
      "epoch": 7.68033946251768,
      "grad_norm": 0.0006862645386718214,
      "learning_rate": 0.19310282418910169,
      "loss": 0.1944,
      "num_input_tokens_seen": 14111456,
      "step": 16290
    },
    {
      "epoch": 7.682696841112683,
      "grad_norm": 0.0005590326036326587,
      "learning_rate": 0.1930464003172166,
      "loss": 0.1985,
      "num_input_tokens_seen": 14116192,
      "step": 16295
    },
    {
      "epoch": 7.685054219707685,
      "grad_norm": 0.001132928067818284,
      "learning_rate": 0.19298996980703567,
      "loss": 0.1948,
      "num_input_tokens_seen": 14120944,
      "step": 16300
    },
    {
      "epoch": 7.687411598302687,
      "grad_norm": 0.0011681076139211655,
      "learning_rate": 0.19293353266726113,
      "loss": 0.2757,
      "num_input_tokens_seen": 14125120,
      "step": 16305
    },
    {
      "epoch": 7.68976897689769,
      "grad_norm": 0.0005028427694924176,
      "learning_rate": 0.19287708890659633,
      "loss": 0.2377,
      "num_input_tokens_seen": 14129616,
      "step": 16310
    },
    {
      "epoch": 7.692126355492692,
      "grad_norm": 0.001385749434120953,
      "learning_rate": 0.19282063853374556,
      "loss": 0.1887,
      "num_input_tokens_seen": 14133696,
      "step": 16315
    },
    {
      "epoch": 7.694483734087695,
      "grad_norm": 0.0014273737324401736,
      "learning_rate": 0.19276418155741423,
      "loss": 0.1974,
      "num_input_tokens_seen": 14138128,
      "step": 16320
    },
    {
      "epoch": 7.696841112682697,
      "grad_norm": 0.0024340790696442127,
      "learning_rate": 0.19270771798630867,
      "loss": 0.2512,
      "num_input_tokens_seen": 14142608,
      "step": 16325
    },
    {
      "epoch": 7.699198491277699,
      "grad_norm": 0.0007513146847486496,
      "learning_rate": 0.1926512478291363,
      "loss": 0.2376,
      "num_input_tokens_seen": 14147120,
      "step": 16330
    },
    {
      "epoch": 7.701555869872702,
      "grad_norm": 0.0013525652466341853,
      "learning_rate": 0.19259477109460557,
      "loss": 0.2363,
      "num_input_tokens_seen": 14151648,
      "step": 16335
    },
    {
      "epoch": 7.703913248467704,
      "grad_norm": 0.000694717513397336,
      "learning_rate": 0.19253828779142584,
      "loss": 0.2298,
      "num_input_tokens_seen": 14156512,
      "step": 16340
    },
    {
      "epoch": 7.706270627062707,
      "grad_norm": 0.001004462013952434,
      "learning_rate": 0.19248179792830755,
      "loss": 0.2282,
      "num_input_tokens_seen": 14160640,
      "step": 16345
    },
    {
      "epoch": 7.708628005657709,
      "grad_norm": 0.001472437521442771,
      "learning_rate": 0.19242530151396217,
      "loss": 0.2369,
      "num_input_tokens_seen": 14165088,
      "step": 16350
    },
    {
      "epoch": 7.7109853842527105,
      "grad_norm": 0.0012310630409047008,
      "learning_rate": 0.19236879855710215,
      "loss": 0.2401,
      "num_input_tokens_seen": 14169376,
      "step": 16355
    },
    {
      "epoch": 7.713342762847713,
      "grad_norm": 0.002175379078835249,
      "learning_rate": 0.19231228906644096,
      "loss": 0.2224,
      "num_input_tokens_seen": 14173600,
      "step": 16360
    },
    {
      "epoch": 7.715700141442715,
      "grad_norm": 0.0014057999942451715,
      "learning_rate": 0.19225577305069302,
      "loss": 0.2175,
      "num_input_tokens_seen": 14177472,
      "step": 16365
    },
    {
      "epoch": 7.718057520037718,
      "grad_norm": 0.0009695614571683109,
      "learning_rate": 0.1921992505185739,
      "loss": 0.2387,
      "num_input_tokens_seen": 14181040,
      "step": 16370
    },
    {
      "epoch": 7.72041489863272,
      "grad_norm": 0.0018180604092776775,
      "learning_rate": 0.19214272147880004,
      "loss": 0.2625,
      "num_input_tokens_seen": 14184848,
      "step": 16375
    },
    {
      "epoch": 7.7227722772277225,
      "grad_norm": 0.0015521554742008448,
      "learning_rate": 0.19208618594008892,
      "loss": 0.2,
      "num_input_tokens_seen": 14189408,
      "step": 16380
    },
    {
      "epoch": 7.725129655822725,
      "grad_norm": 0.0015190079575404525,
      "learning_rate": 0.19202964391115904,
      "loss": 0.227,
      "num_input_tokens_seen": 14193536,
      "step": 16385
    },
    {
      "epoch": 7.727487034417727,
      "grad_norm": 0.0006034130346961319,
      "learning_rate": 0.1919730954007299,
      "loss": 0.2408,
      "num_input_tokens_seen": 14198064,
      "step": 16390
    },
    {
      "epoch": 7.72984441301273,
      "grad_norm": 0.0013140290975570679,
      "learning_rate": 0.19191654041752199,
      "loss": 0.2136,
      "num_input_tokens_seen": 14202336,
      "step": 16395
    },
    {
      "epoch": 7.732201791607732,
      "grad_norm": 0.0014657570281997323,
      "learning_rate": 0.19185997897025678,
      "loss": 0.2223,
      "num_input_tokens_seen": 14205936,
      "step": 16400
    },
    {
      "epoch": 7.732201791607732,
      "eval_loss": 0.21696949005126953,
      "eval_runtime": 21.877,
      "eval_samples_per_second": 43.105,
      "eval_steps_per_second": 21.575,
      "num_input_tokens_seen": 14205936,
      "step": 16400
    },
    {
      "epoch": 7.7345591702027345,
      "grad_norm": 0.0007751883822493255,
      "learning_rate": 0.19180341106765672,
      "loss": 0.2249,
      "num_input_tokens_seen": 14210128,
      "step": 16405
    },
    {
      "epoch": 7.736916548797737,
      "grad_norm": 0.0006194652523845434,
      "learning_rate": 0.19174683671844536,
      "loss": 0.1619,
      "num_input_tokens_seen": 14214608,
      "step": 16410
    },
    {
      "epoch": 7.739273927392739,
      "grad_norm": 0.0010008709505200386,
      "learning_rate": 0.19169025593134717,
      "loss": 0.2523,
      "num_input_tokens_seen": 14219392,
      "step": 16415
    },
    {
      "epoch": 7.741631305987742,
      "grad_norm": 0.001185420318506658,
      "learning_rate": 0.19163366871508764,
      "loss": 0.2002,
      "num_input_tokens_seen": 14223968,
      "step": 16420
    },
    {
      "epoch": 7.743988684582744,
      "grad_norm": 0.00063501694239676,
      "learning_rate": 0.19157707507839317,
      "loss": 0.2233,
      "num_input_tokens_seen": 14228160,
      "step": 16425
    },
    {
      "epoch": 7.7463460631777465,
      "grad_norm": 0.0007994241314008832,
      "learning_rate": 0.19152047502999123,
      "loss": 0.2074,
      "num_input_tokens_seen": 14233168,
      "step": 16430
    },
    {
      "epoch": 7.748703441772749,
      "grad_norm": 0.000623456435278058,
      "learning_rate": 0.19146386857861025,
      "loss": 0.221,
      "num_input_tokens_seen": 14237200,
      "step": 16435
    },
    {
      "epoch": 7.751060820367751,
      "grad_norm": 0.00451489957049489,
      "learning_rate": 0.19140725573297968,
      "loss": 0.2311,
      "num_input_tokens_seen": 14241712,
      "step": 16440
    },
    {
      "epoch": 7.753418198962754,
      "grad_norm": 0.0007984450785443187,
      "learning_rate": 0.19135063650182987,
      "loss": 0.245,
      "num_input_tokens_seen": 14246112,
      "step": 16445
    },
    {
      "epoch": 7.755775577557756,
      "grad_norm": 0.0007842029444873333,
      "learning_rate": 0.19129401089389234,
      "loss": 0.2117,
      "num_input_tokens_seen": 14249776,
      "step": 16450
    },
    {
      "epoch": 7.7581329561527586,
      "grad_norm": 0.0014097105013206601,
      "learning_rate": 0.19123737891789938,
      "loss": 0.2204,
      "num_input_tokens_seen": 14253232,
      "step": 16455
    },
    {
      "epoch": 7.760490334747761,
      "grad_norm": 0.0007513435557484627,
      "learning_rate": 0.19118074058258439,
      "loss": 0.206,
      "num_input_tokens_seen": 14257152,
      "step": 16460
    },
    {
      "epoch": 7.7628477133427625,
      "grad_norm": 0.0008209226070903242,
      "learning_rate": 0.1911240958966816,
      "loss": 0.2054,
      "num_input_tokens_seen": 14261152,
      "step": 16465
    },
    {
      "epoch": 7.765205091937765,
      "grad_norm": 0.0010022311471402645,
      "learning_rate": 0.19106744486892652,
      "loss": 0.2203,
      "num_input_tokens_seen": 14265632,
      "step": 16470
    },
    {
      "epoch": 7.767562470532767,
      "grad_norm": 0.0019513637525960803,
      "learning_rate": 0.1910107875080553,
      "loss": 0.2276,
      "num_input_tokens_seen": 14269760,
      "step": 16475
    },
    {
      "epoch": 7.76991984912777,
      "grad_norm": 0.001403791131451726,
      "learning_rate": 0.19095412382280533,
      "loss": 0.2693,
      "num_input_tokens_seen": 14273440,
      "step": 16480
    },
    {
      "epoch": 7.772277227722772,
      "grad_norm": 0.0009749647579155862,
      "learning_rate": 0.19089745382191473,
      "loss": 0.219,
      "num_input_tokens_seen": 14278048,
      "step": 16485
    },
    {
      "epoch": 7.7746346063177745,
      "grad_norm": 0.0010335799306631088,
      "learning_rate": 0.19084077751412284,
      "loss": 0.237,
      "num_input_tokens_seen": 14282496,
      "step": 16490
    },
    {
      "epoch": 7.776991984912777,
      "grad_norm": 0.00043918329174630344,
      "learning_rate": 0.19078409490816986,
      "loss": 0.2354,
      "num_input_tokens_seen": 14286608,
      "step": 16495
    },
    {
      "epoch": 7.779349363507779,
      "grad_norm": 0.000610384508036077,
      "learning_rate": 0.19072740601279686,
      "loss": 0.2423,
      "num_input_tokens_seen": 14290560,
      "step": 16500
    },
    {
      "epoch": 7.781706742102782,
      "grad_norm": 0.000394374190364033,
      "learning_rate": 0.19067071083674605,
      "loss": 0.2254,
      "num_input_tokens_seen": 14294512,
      "step": 16505
    },
    {
      "epoch": 7.784064120697784,
      "grad_norm": 0.0007553209434263408,
      "learning_rate": 0.19061400938876052,
      "loss": 0.2267,
      "num_input_tokens_seen": 14298384,
      "step": 16510
    },
    {
      "epoch": 7.7864214992927865,
      "grad_norm": 0.0005539768026210368,
      "learning_rate": 0.1905573016775844,
      "loss": 0.2147,
      "num_input_tokens_seen": 14302704,
      "step": 16515
    },
    {
      "epoch": 7.788778877887789,
      "grad_norm": 0.0013677573297172785,
      "learning_rate": 0.19050058771196263,
      "loss": 0.2283,
      "num_input_tokens_seen": 14307552,
      "step": 16520
    },
    {
      "epoch": 7.791136256482791,
      "grad_norm": 0.0007618629024364054,
      "learning_rate": 0.19044386750064132,
      "loss": 0.2032,
      "num_input_tokens_seen": 14313040,
      "step": 16525
    },
    {
      "epoch": 7.793493635077794,
      "grad_norm": 0.0007178982486948371,
      "learning_rate": 0.19038714105236737,
      "loss": 0.2698,
      "num_input_tokens_seen": 14318384,
      "step": 16530
    },
    {
      "epoch": 7.795851013672796,
      "grad_norm": 0.0006554012652486563,
      "learning_rate": 0.19033040837588874,
      "loss": 0.2312,
      "num_input_tokens_seen": 14322672,
      "step": 16535
    },
    {
      "epoch": 7.7982083922677985,
      "grad_norm": 0.0006028180359862745,
      "learning_rate": 0.1902736694799543,
      "loss": 0.2347,
      "num_input_tokens_seen": 14326944,
      "step": 16540
    },
    {
      "epoch": 7.800565770862801,
      "grad_norm": 0.0009679672075435519,
      "learning_rate": 0.19021692437331392,
      "loss": 0.2303,
      "num_input_tokens_seen": 14331360,
      "step": 16545
    },
    {
      "epoch": 7.802923149457803,
      "grad_norm": 0.0005771837313659489,
      "learning_rate": 0.1901601730647184,
      "loss": 0.2413,
      "num_input_tokens_seen": 14335888,
      "step": 16550
    },
    {
      "epoch": 7.805280528052805,
      "grad_norm": 0.0005262447521090508,
      "learning_rate": 0.19010341556291954,
      "loss": 0.2136,
      "num_input_tokens_seen": 14339856,
      "step": 16555
    },
    {
      "epoch": 7.807637906647807,
      "grad_norm": 0.0008040823740884662,
      "learning_rate": 0.19004665187667,
      "loss": 0.2234,
      "num_input_tokens_seen": 14344160,
      "step": 16560
    },
    {
      "epoch": 7.80999528524281,
      "grad_norm": 0.0006279649096541107,
      "learning_rate": 0.1899898820147235,
      "loss": 0.2258,
      "num_input_tokens_seen": 14348880,
      "step": 16565
    },
    {
      "epoch": 7.812352663837812,
      "grad_norm": 0.0006407352047972381,
      "learning_rate": 0.18993310598583465,
      "loss": 0.2557,
      "num_input_tokens_seen": 14353664,
      "step": 16570
    },
    {
      "epoch": 7.814710042432814,
      "grad_norm": 0.0008111881907097995,
      "learning_rate": 0.18987632379875904,
      "loss": 0.2001,
      "num_input_tokens_seen": 14357184,
      "step": 16575
    },
    {
      "epoch": 7.817067421027817,
      "grad_norm": 0.001462264102883637,
      "learning_rate": 0.18981953546225314,
      "loss": 0.2185,
      "num_input_tokens_seen": 14361408,
      "step": 16580
    },
    {
      "epoch": 7.819424799622819,
      "grad_norm": 0.0008462051046080887,
      "learning_rate": 0.18976274098507445,
      "loss": 0.221,
      "num_input_tokens_seen": 14365456,
      "step": 16585
    },
    {
      "epoch": 7.821782178217822,
      "grad_norm": 0.0006756051443517208,
      "learning_rate": 0.18970594037598146,
      "loss": 0.2054,
      "num_input_tokens_seen": 14369680,
      "step": 16590
    },
    {
      "epoch": 7.824139556812824,
      "grad_norm": 0.0007660582778044045,
      "learning_rate": 0.1896491336437335,
      "loss": 0.2301,
      "num_input_tokens_seen": 14374016,
      "step": 16595
    },
    {
      "epoch": 7.826496935407826,
      "grad_norm": 0.0004699956043623388,
      "learning_rate": 0.18959232079709085,
      "loss": 0.214,
      "num_input_tokens_seen": 14378336,
      "step": 16600
    },
    {
      "epoch": 7.826496935407826,
      "eval_loss": 0.2169797569513321,
      "eval_runtime": 21.8937,
      "eval_samples_per_second": 43.072,
      "eval_steps_per_second": 21.559,
      "num_input_tokens_seen": 14378336,
      "step": 16600
    },
    {
      "epoch": 7.828854314002829,
      "grad_norm": 0.0005402026581577957,
      "learning_rate": 0.18953550184481477,
      "loss": 0.2102,
      "num_input_tokens_seen": 14383152,
      "step": 16605
    },
    {
      "epoch": 7.831211692597831,
      "grad_norm": 0.0014377469196915627,
      "learning_rate": 0.18947867679566752,
      "loss": 0.2281,
      "num_input_tokens_seen": 14387344,
      "step": 16610
    },
    {
      "epoch": 7.833569071192834,
      "grad_norm": 0.0005228299996815622,
      "learning_rate": 0.18942184565841216,
      "loss": 0.2087,
      "num_input_tokens_seen": 14391088,
      "step": 16615
    },
    {
      "epoch": 7.835926449787836,
      "grad_norm": 0.0008317032479681075,
      "learning_rate": 0.18936500844181278,
      "loss": 0.233,
      "num_input_tokens_seen": 14396032,
      "step": 16620
    },
    {
      "epoch": 7.838283828382838,
      "grad_norm": 0.0006651821895502508,
      "learning_rate": 0.18930816515463436,
      "loss": 0.1767,
      "num_input_tokens_seen": 14400944,
      "step": 16625
    },
    {
      "epoch": 7.840641206977841,
      "grad_norm": 0.0007315398543141782,
      "learning_rate": 0.18925131580564297,
      "loss": 0.2229,
      "num_input_tokens_seen": 14405856,
      "step": 16630
    },
    {
      "epoch": 7.842998585572843,
      "grad_norm": 0.000786019314546138,
      "learning_rate": 0.1891944604036054,
      "loss": 0.2161,
      "num_input_tokens_seen": 14410528,
      "step": 16635
    },
    {
      "epoch": 7.845355964167846,
      "grad_norm": 0.0016952683217823505,
      "learning_rate": 0.1891375989572895,
      "loss": 0.1911,
      "num_input_tokens_seen": 14415520,
      "step": 16640
    },
    {
      "epoch": 7.847713342762848,
      "grad_norm": 0.0017846195260062814,
      "learning_rate": 0.18908073147546398,
      "loss": 0.2077,
      "num_input_tokens_seen": 14420144,
      "step": 16645
    },
    {
      "epoch": 7.8500707213578504,
      "grad_norm": 0.0008527045138180256,
      "learning_rate": 0.18902385796689858,
      "loss": 0.2757,
      "num_input_tokens_seen": 14424752,
      "step": 16650
    },
    {
      "epoch": 7.852428099952853,
      "grad_norm": 0.0015322606777772307,
      "learning_rate": 0.18896697844036384,
      "loss": 0.268,
      "num_input_tokens_seen": 14428800,
      "step": 16655
    },
    {
      "epoch": 7.854785478547855,
      "grad_norm": 0.0006297137006185949,
      "learning_rate": 0.18891009290463137,
      "loss": 0.2357,
      "num_input_tokens_seen": 14433488,
      "step": 16660
    },
    {
      "epoch": 7.857142857142857,
      "grad_norm": 0.0009909336222335696,
      "learning_rate": 0.18885320136847353,
      "loss": 0.2386,
      "num_input_tokens_seen": 14437232,
      "step": 16665
    },
    {
      "epoch": 7.859500235737859,
      "grad_norm": 0.0009163377690128982,
      "learning_rate": 0.1887963038406639,
      "loss": 0.2306,
      "num_input_tokens_seen": 14442208,
      "step": 16670
    },
    {
      "epoch": 7.861857614332862,
      "grad_norm": 0.0014180743601173162,
      "learning_rate": 0.18873940032997658,
      "loss": 0.2206,
      "num_input_tokens_seen": 14446208,
      "step": 16675
    },
    {
      "epoch": 7.864214992927864,
      "grad_norm": 0.0006967016961425543,
      "learning_rate": 0.18868249084518693,
      "loss": 0.2058,
      "num_input_tokens_seen": 14450464,
      "step": 16680
    },
    {
      "epoch": 7.866572371522866,
      "grad_norm": 0.0006140832556411624,
      "learning_rate": 0.18862557539507102,
      "loss": 0.2342,
      "num_input_tokens_seen": 14454256,
      "step": 16685
    },
    {
      "epoch": 7.868929750117869,
      "grad_norm": 0.0015372233465313911,
      "learning_rate": 0.18856865398840605,
      "loss": 0.188,
      "num_input_tokens_seen": 14459088,
      "step": 16690
    },
    {
      "epoch": 7.871287128712871,
      "grad_norm": 0.001364298164844513,
      "learning_rate": 0.18851172663396995,
      "loss": 0.2002,
      "num_input_tokens_seen": 14462928,
      "step": 16695
    },
    {
      "epoch": 7.873644507307874,
      "grad_norm": 0.0008025983115658164,
      "learning_rate": 0.1884547933405416,
      "loss": 0.2087,
      "num_input_tokens_seen": 14467824,
      "step": 16700
    },
    {
      "epoch": 7.876001885902876,
      "grad_norm": 0.0010142745450139046,
      "learning_rate": 0.1883978541169009,
      "loss": 0.16,
      "num_input_tokens_seen": 14472000,
      "step": 16705
    },
    {
      "epoch": 7.878359264497878,
      "grad_norm": 0.00030641487683169544,
      "learning_rate": 0.18834090897182854,
      "loss": 0.1899,
      "num_input_tokens_seen": 14475632,
      "step": 16710
    },
    {
      "epoch": 7.880716643092881,
      "grad_norm": 0.0021025799214839935,
      "learning_rate": 0.1882839579141062,
      "loss": 0.2581,
      "num_input_tokens_seen": 14479696,
      "step": 16715
    },
    {
      "epoch": 7.883074021687883,
      "grad_norm": 0.0010305347386747599,
      "learning_rate": 0.18822700095251646,
      "loss": 0.2346,
      "num_input_tokens_seen": 14483776,
      "step": 16720
    },
    {
      "epoch": 7.885431400282886,
      "grad_norm": 0.0007564348052255809,
      "learning_rate": 0.18817003809584273,
      "loss": 0.241,
      "num_input_tokens_seen": 14487472,
      "step": 16725
    },
    {
      "epoch": 7.887788778877888,
      "grad_norm": 0.0013136966153979301,
      "learning_rate": 0.1881130693528695,
      "loss": 0.2353,
      "num_input_tokens_seen": 14491568,
      "step": 16730
    },
    {
      "epoch": 7.89014615747289,
      "grad_norm": 0.00042561019654385746,
      "learning_rate": 0.18805609473238197,
      "loss": 0.2217,
      "num_input_tokens_seen": 14496336,
      "step": 16735
    },
    {
      "epoch": 7.892503536067893,
      "grad_norm": 0.003990917932242155,
      "learning_rate": 0.18799911424316643,
      "loss": 0.2249,
      "num_input_tokens_seen": 14501200,
      "step": 16740
    },
    {
      "epoch": 7.894860914662895,
      "grad_norm": 0.0007849353132769465,
      "learning_rate": 0.18794212789400994,
      "loss": 0.2174,
      "num_input_tokens_seen": 14505296,
      "step": 16745
    },
    {
      "epoch": 7.897218293257898,
      "grad_norm": 0.0006329694297164679,
      "learning_rate": 0.18788513569370052,
      "loss": 0.191,
      "num_input_tokens_seen": 14509536,
      "step": 16750
    },
    {
      "epoch": 7.899575671852899,
      "grad_norm": 0.001568491687066853,
      "learning_rate": 0.1878281376510271,
      "loss": 0.2624,
      "num_input_tokens_seen": 14513920,
      "step": 16755
    },
    {
      "epoch": 7.9019330504479015,
      "grad_norm": 0.0016136080957949162,
      "learning_rate": 0.18777113377477941,
      "loss": 0.2289,
      "num_input_tokens_seen": 14518016,
      "step": 16760
    },
    {
      "epoch": 7.904290429042904,
      "grad_norm": 0.0003511714457999915,
      "learning_rate": 0.1877141240737483,
      "loss": 0.2048,
      "num_input_tokens_seen": 14521296,
      "step": 16765
    },
    {
      "epoch": 7.906647807637906,
      "grad_norm": 0.0008250278769992292,
      "learning_rate": 0.18765710855672527,
      "loss": 0.2035,
      "num_input_tokens_seen": 14525600,
      "step": 16770
    },
    {
      "epoch": 7.909005186232909,
      "grad_norm": 0.0006678131176158786,
      "learning_rate": 0.18760008723250288,
      "loss": 0.2242,
      "num_input_tokens_seen": 14529600,
      "step": 16775
    },
    {
      "epoch": 7.911362564827911,
      "grad_norm": 0.0010437999153509736,
      "learning_rate": 0.18754306010987457,
      "loss": 0.2127,
      "num_input_tokens_seen": 14533744,
      "step": 16780
    },
    {
      "epoch": 7.9137199434229135,
      "grad_norm": 0.002104994608089328,
      "learning_rate": 0.18748602719763457,
      "loss": 0.243,
      "num_input_tokens_seen": 14538672,
      "step": 16785
    },
    {
      "epoch": 7.916077322017916,
      "grad_norm": 0.000894346390850842,
      "learning_rate": 0.18742898850457804,
      "loss": 0.2255,
      "num_input_tokens_seen": 14542560,
      "step": 16790
    },
    {
      "epoch": 7.918434700612918,
      "grad_norm": 0.0015995741123333573,
      "learning_rate": 0.1873719440395012,
      "loss": 0.1968,
      "num_input_tokens_seen": 14547376,
      "step": 16795
    },
    {
      "epoch": 7.920792079207921,
      "grad_norm": 0.0008060087566263974,
      "learning_rate": 0.1873148938112009,
      "loss": 0.23,
      "num_input_tokens_seen": 14551456,
      "step": 16800
    },
    {
      "epoch": 7.920792079207921,
      "eval_loss": 0.21796095371246338,
      "eval_runtime": 21.8888,
      "eval_samples_per_second": 43.081,
      "eval_steps_per_second": 21.564,
      "num_input_tokens_seen": 14551456,
      "step": 16800
    },
    {
      "epoch": 7.923149457802923,
      "grad_norm": 0.0007948409183882177,
      "learning_rate": 0.18725783782847508,
      "loss": 0.197,
      "num_input_tokens_seen": 14555536,
      "step": 16805
    },
    {
      "epoch": 7.9255068363979255,
      "grad_norm": 0.0007202514680102468,
      "learning_rate": 0.1872007761001224,
      "loss": 0.2052,
      "num_input_tokens_seen": 14559824,
      "step": 16810
    },
    {
      "epoch": 7.927864214992928,
      "grad_norm": 0.0004802125331480056,
      "learning_rate": 0.1871437086349426,
      "loss": 0.2233,
      "num_input_tokens_seen": 14563536,
      "step": 16815
    },
    {
      "epoch": 7.93022159358793,
      "grad_norm": 0.0004227221943438053,
      "learning_rate": 0.18708663544173615,
      "loss": 0.2098,
      "num_input_tokens_seen": 14567600,
      "step": 16820
    },
    {
      "epoch": 7.932578972182933,
      "grad_norm": 0.0012340486282482743,
      "learning_rate": 0.18702955652930442,
      "loss": 0.198,
      "num_input_tokens_seen": 14571920,
      "step": 16825
    },
    {
      "epoch": 7.934936350777935,
      "grad_norm": 0.0005463844863697886,
      "learning_rate": 0.18697247190644972,
      "loss": 0.1406,
      "num_input_tokens_seen": 14576016,
      "step": 16830
    },
    {
      "epoch": 7.9372937293729375,
      "grad_norm": 0.0006519047892652452,
      "learning_rate": 0.18691538158197527,
      "loss": 0.2501,
      "num_input_tokens_seen": 14581152,
      "step": 16835
    },
    {
      "epoch": 7.93965110796794,
      "grad_norm": 0.0010418759193271399,
      "learning_rate": 0.1868582855646851,
      "loss": 0.2309,
      "num_input_tokens_seen": 14586000,
      "step": 16840
    },
    {
      "epoch": 7.942008486562942,
      "grad_norm": 0.0009885280160233378,
      "learning_rate": 0.18680118386338404,
      "loss": 0.1987,
      "num_input_tokens_seen": 14589984,
      "step": 16845
    },
    {
      "epoch": 7.944365865157945,
      "grad_norm": 0.0009283453109674156,
      "learning_rate": 0.18674407648687794,
      "loss": 0.2376,
      "num_input_tokens_seen": 14594240,
      "step": 16850
    },
    {
      "epoch": 7.946723243752947,
      "grad_norm": 0.0006845822790637612,
      "learning_rate": 0.1866869634439736,
      "loss": 0.2156,
      "num_input_tokens_seen": 14598880,
      "step": 16855
    },
    {
      "epoch": 7.9490806223479495,
      "grad_norm": 0.0007146000280044973,
      "learning_rate": 0.18662984474347838,
      "loss": 0.2344,
      "num_input_tokens_seen": 14603888,
      "step": 16860
    },
    {
      "epoch": 7.951438000942951,
      "grad_norm": 0.001735373050905764,
      "learning_rate": 0.1865727203942008,
      "loss": 0.2156,
      "num_input_tokens_seen": 14608448,
      "step": 16865
    },
    {
      "epoch": 7.9537953795379535,
      "grad_norm": 0.0004984252736903727,
      "learning_rate": 0.1865155904049501,
      "loss": 0.2489,
      "num_input_tokens_seen": 14612624,
      "step": 16870
    },
    {
      "epoch": 7.956152758132956,
      "grad_norm": 0.0009617778705433011,
      "learning_rate": 0.1864584547845365,
      "loss": 0.2142,
      "num_input_tokens_seen": 14617136,
      "step": 16875
    },
    {
      "epoch": 7.958510136727958,
      "grad_norm": 0.0011309472611173987,
      "learning_rate": 0.186401313541771,
      "loss": 0.2444,
      "num_input_tokens_seen": 14621376,
      "step": 16880
    },
    {
      "epoch": 7.960867515322961,
      "grad_norm": 0.001117267063818872,
      "learning_rate": 0.18634416668546552,
      "loss": 0.2288,
      "num_input_tokens_seen": 14625424,
      "step": 16885
    },
    {
      "epoch": 7.963224893917963,
      "grad_norm": 0.0008237235015258193,
      "learning_rate": 0.1862870142244328,
      "loss": 0.2258,
      "num_input_tokens_seen": 14629488,
      "step": 16890
    },
    {
      "epoch": 7.9655822725129655,
      "grad_norm": 0.0006275955238379538,
      "learning_rate": 0.1862298561674865,
      "loss": 0.2082,
      "num_input_tokens_seen": 14633920,
      "step": 16895
    },
    {
      "epoch": 7.967939651107968,
      "grad_norm": 0.0008190554799512029,
      "learning_rate": 0.18617269252344104,
      "loss": 0.2567,
      "num_input_tokens_seen": 14638144,
      "step": 16900
    },
    {
      "epoch": 7.97029702970297,
      "grad_norm": 0.0008122554863803089,
      "learning_rate": 0.18611552330111186,
      "loss": 0.2576,
      "num_input_tokens_seen": 14642976,
      "step": 16905
    },
    {
      "epoch": 7.972654408297973,
      "grad_norm": 0.000676579016726464,
      "learning_rate": 0.18605834850931507,
      "loss": 0.2435,
      "num_input_tokens_seen": 14647136,
      "step": 16910
    },
    {
      "epoch": 7.975011786892975,
      "grad_norm": 0.000672378926537931,
      "learning_rate": 0.18600116815686787,
      "loss": 0.2135,
      "num_input_tokens_seen": 14652672,
      "step": 16915
    },
    {
      "epoch": 7.9773691654879775,
      "grad_norm": 0.0010219658724963665,
      "learning_rate": 0.1859439822525881,
      "loss": 0.2069,
      "num_input_tokens_seen": 14657952,
      "step": 16920
    },
    {
      "epoch": 7.97972654408298,
      "grad_norm": 0.0005619163857772946,
      "learning_rate": 0.18588679080529455,
      "loss": 0.244,
      "num_input_tokens_seen": 14662768,
      "step": 16925
    },
    {
      "epoch": 7.982083922677982,
      "grad_norm": 0.0007368511287495494,
      "learning_rate": 0.1858295938238069,
      "loss": 0.1957,
      "num_input_tokens_seen": 14667488,
      "step": 16930
    },
    {
      "epoch": 7.984441301272985,
      "grad_norm": 0.0005778747727163136,
      "learning_rate": 0.18577239131694562,
      "loss": 0.2265,
      "num_input_tokens_seen": 14672656,
      "step": 16935
    },
    {
      "epoch": 7.986798679867987,
      "grad_norm": 0.0005841944948770106,
      "learning_rate": 0.18571518329353204,
      "loss": 0.19,
      "num_input_tokens_seen": 14677168,
      "step": 16940
    },
    {
      "epoch": 7.9891560584629895,
      "grad_norm": 0.0004305484180804342,
      "learning_rate": 0.18565796976238838,
      "loss": 0.1969,
      "num_input_tokens_seen": 14681904,
      "step": 16945
    },
    {
      "epoch": 7.991513437057992,
      "grad_norm": 0.0009605127270333469,
      "learning_rate": 0.18560075073233764,
      "loss": 0.2415,
      "num_input_tokens_seen": 14685856,
      "step": 16950
    },
    {
      "epoch": 7.993870815652993,
      "grad_norm": 0.00220647850073874,
      "learning_rate": 0.18554352621220377,
      "loss": 0.2376,
      "num_input_tokens_seen": 14691056,
      "step": 16955
    },
    {
      "epoch": 7.996228194247996,
      "grad_norm": 0.001500558340921998,
      "learning_rate": 0.18548629621081153,
      "loss": 0.23,
      "num_input_tokens_seen": 14695488,
      "step": 16960
    },
    {
      "epoch": 7.998585572842998,
      "grad_norm": 0.0006912950775586069,
      "learning_rate": 0.18542906073698645,
      "loss": 0.2156,
      "num_input_tokens_seen": 14699536,
      "step": 16965
    },
    {
      "epoch": 8.000942951438,
      "grad_norm": 0.0007868616376072168,
      "learning_rate": 0.18537181979955494,
      "loss": 0.2194,
      "num_input_tokens_seen": 14704272,
      "step": 16970
    },
    {
      "epoch": 8.003300330033003,
      "grad_norm": 0.0014794172020629048,
      "learning_rate": 0.18531457340734434,
      "loss": 0.1774,
      "num_input_tokens_seen": 14708352,
      "step": 16975
    },
    {
      "epoch": 8.005657708628005,
      "grad_norm": 0.000944069295655936,
      "learning_rate": 0.1852573215691827,
      "loss": 0.1913,
      "num_input_tokens_seen": 14712880,
      "step": 16980
    },
    {
      "epoch": 8.008015087223008,
      "grad_norm": 0.0010439527686685324,
      "learning_rate": 0.18520006429389904,
      "loss": 0.2206,
      "num_input_tokens_seen": 14717344,
      "step": 16985
    },
    {
      "epoch": 8.01037246581801,
      "grad_norm": 0.0021208319813013077,
      "learning_rate": 0.1851428015903231,
      "loss": 0.2307,
      "num_input_tokens_seen": 14721632,
      "step": 16990
    },
    {
      "epoch": 8.012729844413013,
      "grad_norm": 0.0018792771734297276,
      "learning_rate": 0.1850855334672855,
      "loss": 0.2634,
      "num_input_tokens_seen": 14726080,
      "step": 16995
    },
    {
      "epoch": 8.015087223008015,
      "grad_norm": 0.000571501615922898,
      "learning_rate": 0.1850282599336178,
      "loss": 0.1974,
      "num_input_tokens_seen": 14730672,
      "step": 17000
    },
    {
      "epoch": 8.015087223008015,
      "eval_loss": 0.2183476835489273,
      "eval_runtime": 21.9196,
      "eval_samples_per_second": 43.021,
      "eval_steps_per_second": 21.533,
      "num_input_tokens_seen": 14730672,
      "step": 17000
    },
    {
      "epoch": 8.017444601603017,
      "grad_norm": 0.0005012344918213785,
      "learning_rate": 0.18497098099815215,
      "loss": 0.1977,
      "num_input_tokens_seen": 14734560,
      "step": 17005
    },
    {
      "epoch": 8.01980198019802,
      "grad_norm": 0.0007977562490850687,
      "learning_rate": 0.18491369666972174,
      "loss": 0.2173,
      "num_input_tokens_seen": 14738896,
      "step": 17010
    },
    {
      "epoch": 8.022159358793022,
      "grad_norm": 0.0005976989632472396,
      "learning_rate": 0.1848564069571606,
      "loss": 0.2212,
      "num_input_tokens_seen": 14743120,
      "step": 17015
    },
    {
      "epoch": 8.024516737388025,
      "grad_norm": 0.0005675868014805019,
      "learning_rate": 0.18479911186930348,
      "loss": 0.2483,
      "num_input_tokens_seen": 14747504,
      "step": 17020
    },
    {
      "epoch": 8.026874115983027,
      "grad_norm": 0.0004155739734414965,
      "learning_rate": 0.18474181141498597,
      "loss": 0.2145,
      "num_input_tokens_seen": 14751856,
      "step": 17025
    },
    {
      "epoch": 8.02923149457803,
      "grad_norm": 0.0011215610429644585,
      "learning_rate": 0.18468450560304453,
      "loss": 0.2355,
      "num_input_tokens_seen": 14756288,
      "step": 17030
    },
    {
      "epoch": 8.031588873173032,
      "grad_norm": 0.0009645810932852328,
      "learning_rate": 0.1846271944423165,
      "loss": 0.2386,
      "num_input_tokens_seen": 14760416,
      "step": 17035
    },
    {
      "epoch": 8.033946251768034,
      "grad_norm": 0.0005107417237013578,
      "learning_rate": 0.18456987794163993,
      "loss": 0.2353,
      "num_input_tokens_seen": 14765184,
      "step": 17040
    },
    {
      "epoch": 8.036303630363037,
      "grad_norm": 0.0022268968168646097,
      "learning_rate": 0.18451255610985373,
      "loss": 0.2409,
      "num_input_tokens_seen": 14769792,
      "step": 17045
    },
    {
      "epoch": 8.038661008958039,
      "grad_norm": 0.001414511469192803,
      "learning_rate": 0.18445522895579766,
      "loss": 0.2403,
      "num_input_tokens_seen": 14774560,
      "step": 17050
    },
    {
      "epoch": 8.041018387553041,
      "grad_norm": 0.0006551123224198818,
      "learning_rate": 0.1843978964883123,
      "loss": 0.2176,
      "num_input_tokens_seen": 14778400,
      "step": 17055
    },
    {
      "epoch": 8.043375766148044,
      "grad_norm": 0.0007221862324513495,
      "learning_rate": 0.18434055871623906,
      "loss": 0.2217,
      "num_input_tokens_seen": 14782704,
      "step": 17060
    },
    {
      "epoch": 8.045733144743046,
      "grad_norm": 0.0006119785830378532,
      "learning_rate": 0.18428321564842007,
      "loss": 0.2187,
      "num_input_tokens_seen": 14786688,
      "step": 17065
    },
    {
      "epoch": 8.048090523338049,
      "grad_norm": 0.0012603659415617585,
      "learning_rate": 0.18422586729369841,
      "loss": 0.2177,
      "num_input_tokens_seen": 14791104,
      "step": 17070
    },
    {
      "epoch": 8.050447901933051,
      "grad_norm": 0.0007500843494199216,
      "learning_rate": 0.1841685136609179,
      "loss": 0.2429,
      "num_input_tokens_seen": 14795872,
      "step": 17075
    },
    {
      "epoch": 8.052805280528053,
      "grad_norm": 0.0006081779138185084,
      "learning_rate": 0.18411115475892326,
      "loss": 0.2257,
      "num_input_tokens_seen": 14800368,
      "step": 17080
    },
    {
      "epoch": 8.055162659123056,
      "grad_norm": 0.0007733661332167685,
      "learning_rate": 0.18405379059655982,
      "loss": 0.2381,
      "num_input_tokens_seen": 14805408,
      "step": 17085
    },
    {
      "epoch": 8.057520037718058,
      "grad_norm": 0.000798562599811703,
      "learning_rate": 0.1839964211826739,
      "loss": 0.2055,
      "num_input_tokens_seen": 14809728,
      "step": 17090
    },
    {
      "epoch": 8.05987741631306,
      "grad_norm": 0.0006754283094778657,
      "learning_rate": 0.18393904652611265,
      "loss": 0.2168,
      "num_input_tokens_seen": 14814016,
      "step": 17095
    },
    {
      "epoch": 8.062234794908063,
      "grad_norm": 0.0006707613938488066,
      "learning_rate": 0.18388166663572392,
      "loss": 0.2172,
      "num_input_tokens_seen": 14818784,
      "step": 17100
    },
    {
      "epoch": 8.064592173503065,
      "grad_norm": 0.0014229442458599806,
      "learning_rate": 0.18382428152035643,
      "loss": 0.2461,
      "num_input_tokens_seen": 14824128,
      "step": 17105
    },
    {
      "epoch": 8.066949552098066,
      "grad_norm": 0.0005571647197939456,
      "learning_rate": 0.1837668911888596,
      "loss": 0.2169,
      "num_input_tokens_seen": 14828256,
      "step": 17110
    },
    {
      "epoch": 8.069306930693068,
      "grad_norm": 0.001506650703959167,
      "learning_rate": 0.18370949565008388,
      "loss": 0.2071,
      "num_input_tokens_seen": 14832656,
      "step": 17115
    },
    {
      "epoch": 8.07166430928807,
      "grad_norm": 0.0005854166811332107,
      "learning_rate": 0.1836520949128803,
      "loss": 0.2178,
      "num_input_tokens_seen": 14836544,
      "step": 17120
    },
    {
      "epoch": 8.074021687883073,
      "grad_norm": 0.0007889504777267575,
      "learning_rate": 0.18359468898610076,
      "loss": 0.2163,
      "num_input_tokens_seen": 14841632,
      "step": 17125
    },
    {
      "epoch": 8.076379066478076,
      "grad_norm": 0.0016347586642950773,
      "learning_rate": 0.18353727787859797,
      "loss": 0.2242,
      "num_input_tokens_seen": 14845632,
      "step": 17130
    },
    {
      "epoch": 8.078736445073078,
      "grad_norm": 0.0004919386119581759,
      "learning_rate": 0.18347986159922552,
      "loss": 0.1917,
      "num_input_tokens_seen": 14849824,
      "step": 17135
    },
    {
      "epoch": 8.08109382366808,
      "grad_norm": 0.0008397671044804156,
      "learning_rate": 0.1834224401568377,
      "loss": 0.2483,
      "num_input_tokens_seen": 14854416,
      "step": 17140
    },
    {
      "epoch": 8.083451202263083,
      "grad_norm": 0.001358521985821426,
      "learning_rate": 0.1833650135602896,
      "loss": 0.2407,
      "num_input_tokens_seen": 14857728,
      "step": 17145
    },
    {
      "epoch": 8.085808580858085,
      "grad_norm": 0.0013789423974230886,
      "learning_rate": 0.18330758181843707,
      "loss": 0.1937,
      "num_input_tokens_seen": 14861856,
      "step": 17150
    },
    {
      "epoch": 8.088165959453088,
      "grad_norm": 0.0007847712840884924,
      "learning_rate": 0.18325014494013686,
      "loss": 0.2233,
      "num_input_tokens_seen": 14865280,
      "step": 17155
    },
    {
      "epoch": 8.09052333804809,
      "grad_norm": 0.0005809231079183519,
      "learning_rate": 0.18319270293424647,
      "loss": 0.2076,
      "num_input_tokens_seen": 14869936,
      "step": 17160
    },
    {
      "epoch": 8.092880716643092,
      "grad_norm": 0.001483680447563529,
      "learning_rate": 0.18313525580962417,
      "loss": 0.207,
      "num_input_tokens_seen": 14873936,
      "step": 17165
    },
    {
      "epoch": 8.095238095238095,
      "grad_norm": 0.0007495217141695321,
      "learning_rate": 0.18307780357512896,
      "loss": 0.2297,
      "num_input_tokens_seen": 14878208,
      "step": 17170
    },
    {
      "epoch": 8.097595473833097,
      "grad_norm": 0.0007839567260816693,
      "learning_rate": 0.1830203462396208,
      "loss": 0.2125,
      "num_input_tokens_seen": 14882672,
      "step": 17175
    },
    {
      "epoch": 8.0999528524281,
      "grad_norm": 0.0007993195904418826,
      "learning_rate": 0.18296288381196033,
      "loss": 0.2295,
      "num_input_tokens_seen": 14887632,
      "step": 17180
    },
    {
      "epoch": 8.102310231023102,
      "grad_norm": 0.0014070695033296943,
      "learning_rate": 0.1829054163010089,
      "loss": 0.1975,
      "num_input_tokens_seen": 14891664,
      "step": 17185
    },
    {
      "epoch": 8.104667609618105,
      "grad_norm": 0.0004898497718386352,
      "learning_rate": 0.18284794371562874,
      "loss": 0.2214,
      "num_input_tokens_seen": 14896256,
      "step": 17190
    },
    {
      "epoch": 8.107024988213107,
      "grad_norm": 0.0005600343574769795,
      "learning_rate": 0.18279046606468288,
      "loss": 0.2247,
      "num_input_tokens_seen": 14900352,
      "step": 17195
    },
    {
      "epoch": 8.10938236680811,
      "grad_norm": 0.0006317143561318517,
      "learning_rate": 0.1827329833570351,
      "loss": 0.2445,
      "num_input_tokens_seen": 14904544,
      "step": 17200
    },
    {
      "epoch": 8.10938236680811,
      "eval_loss": 0.21596461534500122,
      "eval_runtime": 21.9243,
      "eval_samples_per_second": 43.012,
      "eval_steps_per_second": 21.529,
      "num_input_tokens_seen": 14904544,
      "step": 17200
    },
    {
      "epoch": 8.111739745403112,
      "grad_norm": 0.0006996337906457484,
      "learning_rate": 0.18267549560154991,
      "loss": 0.2483,
      "num_input_tokens_seen": 14908528,
      "step": 17205
    },
    {
      "epoch": 8.114097123998114,
      "grad_norm": 0.0017482263501733541,
      "learning_rate": 0.18261800280709267,
      "loss": 0.2201,
      "num_input_tokens_seen": 14912864,
      "step": 17210
    },
    {
      "epoch": 8.116454502593117,
      "grad_norm": 0.0017100503901019692,
      "learning_rate": 0.18256050498252957,
      "loss": 0.2187,
      "num_input_tokens_seen": 14917984,
      "step": 17215
    },
    {
      "epoch": 8.118811881188119,
      "grad_norm": 0.001442396780475974,
      "learning_rate": 0.18250300213672735,
      "loss": 0.2402,
      "num_input_tokens_seen": 14921760,
      "step": 17220
    },
    {
      "epoch": 8.121169259783121,
      "grad_norm": 0.0012662067310884595,
      "learning_rate": 0.18244549427855378,
      "loss": 0.2659,
      "num_input_tokens_seen": 14926160,
      "step": 17225
    },
    {
      "epoch": 8.123526638378124,
      "grad_norm": 0.0018833104986697435,
      "learning_rate": 0.1823879814168772,
      "loss": 0.2332,
      "num_input_tokens_seen": 14930368,
      "step": 17230
    },
    {
      "epoch": 8.125884016973126,
      "grad_norm": 0.0005143648595549166,
      "learning_rate": 0.18233046356056692,
      "loss": 0.2383,
      "num_input_tokens_seen": 14934672,
      "step": 17235
    },
    {
      "epoch": 8.128241395568129,
      "grad_norm": 0.0005064297583885491,
      "learning_rate": 0.18227294071849284,
      "loss": 0.2256,
      "num_input_tokens_seen": 14939696,
      "step": 17240
    },
    {
      "epoch": 8.130598774163131,
      "grad_norm": 0.0012462330050766468,
      "learning_rate": 0.18221541289952578,
      "loss": 0.1988,
      "num_input_tokens_seen": 14943568,
      "step": 17245
    },
    {
      "epoch": 8.132956152758133,
      "grad_norm": 0.000742485688533634,
      "learning_rate": 0.18215788011253717,
      "loss": 0.2265,
      "num_input_tokens_seen": 14948672,
      "step": 17250
    },
    {
      "epoch": 8.135313531353136,
      "grad_norm": 0.0010609545279294252,
      "learning_rate": 0.18210034236639935,
      "loss": 0.2026,
      "num_input_tokens_seen": 14953824,
      "step": 17255
    },
    {
      "epoch": 8.137670909948138,
      "grad_norm": 0.0010143255349248648,
      "learning_rate": 0.1820427996699853,
      "loss": 0.1671,
      "num_input_tokens_seen": 14957936,
      "step": 17260
    },
    {
      "epoch": 8.14002828854314,
      "grad_norm": 0.00041120799141936004,
      "learning_rate": 0.1819852520321689,
      "loss": 0.2356,
      "num_input_tokens_seen": 14962224,
      "step": 17265
    },
    {
      "epoch": 8.142385667138143,
      "grad_norm": 0.0008708821260370314,
      "learning_rate": 0.18192769946182466,
      "loss": 0.2237,
      "num_input_tokens_seen": 14966288,
      "step": 17270
    },
    {
      "epoch": 8.144743045733145,
      "grad_norm": 0.0010673468932509422,
      "learning_rate": 0.18187014196782794,
      "loss": 0.1769,
      "num_input_tokens_seen": 14971184,
      "step": 17275
    },
    {
      "epoch": 8.147100424328148,
      "grad_norm": 0.0003312168410047889,
      "learning_rate": 0.18181257955905486,
      "loss": 0.1729,
      "num_input_tokens_seen": 14976112,
      "step": 17280
    },
    {
      "epoch": 8.14945780292315,
      "grad_norm": 0.001018760958686471,
      "learning_rate": 0.18175501224438217,
      "loss": 0.2412,
      "num_input_tokens_seen": 14980448,
      "step": 17285
    },
    {
      "epoch": 8.151815181518153,
      "grad_norm": 0.001600234187208116,
      "learning_rate": 0.18169744003268756,
      "loss": 0.2634,
      "num_input_tokens_seen": 14985024,
      "step": 17290
    },
    {
      "epoch": 8.154172560113155,
      "grad_norm": 0.0006341019761748612,
      "learning_rate": 0.18163986293284937,
      "loss": 0.2502,
      "num_input_tokens_seen": 14989680,
      "step": 17295
    },
    {
      "epoch": 8.156529938708157,
      "grad_norm": 0.0015918717253953218,
      "learning_rate": 0.18158228095374673,
      "loss": 0.2165,
      "num_input_tokens_seen": 14994000,
      "step": 17300
    },
    {
      "epoch": 8.15888731730316,
      "grad_norm": 0.0008630157099105418,
      "learning_rate": 0.18152469410425945,
      "loss": 0.2274,
      "num_input_tokens_seen": 14998416,
      "step": 17305
    },
    {
      "epoch": 8.16124469589816,
      "grad_norm": 0.0012875405373051763,
      "learning_rate": 0.18146710239326813,
      "loss": 0.2354,
      "num_input_tokens_seen": 15003776,
      "step": 17310
    },
    {
      "epoch": 8.163602074493163,
      "grad_norm": 0.0007028420804999769,
      "learning_rate": 0.18140950582965423,
      "loss": 0.2218,
      "num_input_tokens_seen": 15008224,
      "step": 17315
    },
    {
      "epoch": 8.165959453088165,
      "grad_norm": 0.0019235782092437148,
      "learning_rate": 0.1813519044222998,
      "loss": 0.2278,
      "num_input_tokens_seen": 15012528,
      "step": 17320
    },
    {
      "epoch": 8.168316831683168,
      "grad_norm": 0.0010205862345173955,
      "learning_rate": 0.18129429818008772,
      "loss": 0.235,
      "num_input_tokens_seen": 15016656,
      "step": 17325
    },
    {
      "epoch": 8.17067421027817,
      "grad_norm": 0.0007929943967610598,
      "learning_rate": 0.18123668711190163,
      "loss": 0.2159,
      "num_input_tokens_seen": 15020752,
      "step": 17330
    },
    {
      "epoch": 8.173031588873172,
      "grad_norm": 0.0006862014997750521,
      "learning_rate": 0.18117907122662583,
      "loss": 0.2191,
      "num_input_tokens_seen": 15025216,
      "step": 17335
    },
    {
      "epoch": 8.175388967468175,
      "grad_norm": 0.0012741342652589083,
      "learning_rate": 0.1811214505331454,
      "loss": 0.1819,
      "num_input_tokens_seen": 15029184,
      "step": 17340
    },
    {
      "epoch": 8.177746346063177,
      "grad_norm": 0.0004548447031993419,
      "learning_rate": 0.1810638250403462,
      "loss": 0.2161,
      "num_input_tokens_seen": 15033008,
      "step": 17345
    },
    {
      "epoch": 8.18010372465818,
      "grad_norm": 0.0004260679997969419,
      "learning_rate": 0.1810061947571148,
      "loss": 0.2246,
      "num_input_tokens_seen": 15036560,
      "step": 17350
    },
    {
      "epoch": 8.182461103253182,
      "grad_norm": 0.0003665993281174451,
      "learning_rate": 0.1809485596923385,
      "loss": 0.1691,
      "num_input_tokens_seen": 15040336,
      "step": 17355
    },
    {
      "epoch": 8.184818481848184,
      "grad_norm": 0.0004515019536484033,
      "learning_rate": 0.18089091985490546,
      "loss": 0.2465,
      "num_input_tokens_seen": 15044256,
      "step": 17360
    },
    {
      "epoch": 8.187175860443187,
      "grad_norm": 0.001216707518324256,
      "learning_rate": 0.18083327525370432,
      "loss": 0.2349,
      "num_input_tokens_seen": 15047744,
      "step": 17365
    },
    {
      "epoch": 8.18953323903819,
      "grad_norm": 0.000654072966426611,
      "learning_rate": 0.18077562589762464,
      "loss": 0.2375,
      "num_input_tokens_seen": 15051616,
      "step": 17370
    },
    {
      "epoch": 8.191890617633192,
      "grad_norm": 0.0006570627447217703,
      "learning_rate": 0.1807179717955567,
      "loss": 0.2366,
      "num_input_tokens_seen": 15055632,
      "step": 17375
    },
    {
      "epoch": 8.194247996228194,
      "grad_norm": 0.0010528704151511192,
      "learning_rate": 0.1806603129563915,
      "loss": 0.2,
      "num_input_tokens_seen": 15059536,
      "step": 17380
    },
    {
      "epoch": 8.196605374823196,
      "grad_norm": 0.00138662604149431,
      "learning_rate": 0.1806026493890208,
      "loss": 0.2018,
      "num_input_tokens_seen": 15064336,
      "step": 17385
    },
    {
      "epoch": 8.198962753418199,
      "grad_norm": 0.0006904808105900884,
      "learning_rate": 0.18054498110233688,
      "loss": 0.2144,
      "num_input_tokens_seen": 15069456,
      "step": 17390
    },
    {
      "epoch": 8.201320132013201,
      "grad_norm": 0.0006260880618356168,
      "learning_rate": 0.1804873081052331,
      "loss": 0.1757,
      "num_input_tokens_seen": 15074416,
      "step": 17395
    },
    {
      "epoch": 8.203677510608204,
      "grad_norm": 0.001168601680546999,
      "learning_rate": 0.18042963040660326,
      "loss": 0.1646,
      "num_input_tokens_seen": 15078832,
      "step": 17400
    },
    {
      "epoch": 8.203677510608204,
      "eval_loss": 0.2226901650428772,
      "eval_runtime": 21.9877,
      "eval_samples_per_second": 42.888,
      "eval_steps_per_second": 21.467,
      "num_input_tokens_seen": 15078832,
      "step": 17400
    },
    {
      "epoch": 8.206034889203206,
      "grad_norm": 0.00063800293719396,
      "learning_rate": 0.180371948015342,
      "loss": 0.1843,
      "num_input_tokens_seen": 15083120,
      "step": 17405
    },
    {
      "epoch": 8.208392267798208,
      "grad_norm": 0.0005953225190751255,
      "learning_rate": 0.18031426094034472,
      "loss": 0.2623,
      "num_input_tokens_seen": 15088272,
      "step": 17410
    },
    {
      "epoch": 8.21074964639321,
      "grad_norm": 0.0011070126201957464,
      "learning_rate": 0.18025656919050737,
      "loss": 0.1672,
      "num_input_tokens_seen": 15092080,
      "step": 17415
    },
    {
      "epoch": 8.213107024988213,
      "grad_norm": 0.0012170132249593735,
      "learning_rate": 0.18019887277472688,
      "loss": 0.2125,
      "num_input_tokens_seen": 15095952,
      "step": 17420
    },
    {
      "epoch": 8.215464403583216,
      "grad_norm": 0.002036644145846367,
      "learning_rate": 0.18014117170190067,
      "loss": 0.2789,
      "num_input_tokens_seen": 15100128,
      "step": 17425
    },
    {
      "epoch": 8.217821782178218,
      "grad_norm": 0.0014281088951975107,
      "learning_rate": 0.18008346598092703,
      "loss": 0.2381,
      "num_input_tokens_seen": 15104416,
      "step": 17430
    },
    {
      "epoch": 8.22017916077322,
      "grad_norm": 0.0008759621996432543,
      "learning_rate": 0.18002575562070489,
      "loss": 0.2311,
      "num_input_tokens_seen": 15108592,
      "step": 17435
    },
    {
      "epoch": 8.222536539368223,
      "grad_norm": 0.000924751628190279,
      "learning_rate": 0.1799680406301339,
      "loss": 0.2133,
      "num_input_tokens_seen": 15112608,
      "step": 17440
    },
    {
      "epoch": 8.224893917963225,
      "grad_norm": 0.0012491259258240461,
      "learning_rate": 0.17991032101811447,
      "loss": 0.2237,
      "num_input_tokens_seen": 15116608,
      "step": 17445
    },
    {
      "epoch": 8.227251296558228,
      "grad_norm": 0.0004877631727140397,
      "learning_rate": 0.1798525967935476,
      "loss": 0.2203,
      "num_input_tokens_seen": 15120064,
      "step": 17450
    },
    {
      "epoch": 8.22960867515323,
      "grad_norm": 0.0006171252462081611,
      "learning_rate": 0.17979486796533517,
      "loss": 0.2401,
      "num_input_tokens_seen": 15124608,
      "step": 17455
    },
    {
      "epoch": 8.231966053748232,
      "grad_norm": 0.00073317188071087,
      "learning_rate": 0.1797371345423797,
      "loss": 0.1908,
      "num_input_tokens_seen": 15129376,
      "step": 17460
    },
    {
      "epoch": 8.234323432343235,
      "grad_norm": 0.0014318376779556274,
      "learning_rate": 0.17967939653358436,
      "loss": 0.2625,
      "num_input_tokens_seen": 15133936,
      "step": 17465
    },
    {
      "epoch": 8.236680810938237,
      "grad_norm": 0.0006916575948707759,
      "learning_rate": 0.17962165394785315,
      "loss": 0.2042,
      "num_input_tokens_seen": 15138272,
      "step": 17470
    },
    {
      "epoch": 8.23903818953324,
      "grad_norm": 0.0007186068105511367,
      "learning_rate": 0.17956390679409057,
      "loss": 0.2386,
      "num_input_tokens_seen": 15143056,
      "step": 17475
    },
    {
      "epoch": 8.241395568128242,
      "grad_norm": 0.0008136983960866928,
      "learning_rate": 0.1795061550812021,
      "loss": 0.2273,
      "num_input_tokens_seen": 15147168,
      "step": 17480
    },
    {
      "epoch": 8.243752946723244,
      "grad_norm": 0.000652920629363507,
      "learning_rate": 0.1794483988180937,
      "loss": 0.223,
      "num_input_tokens_seen": 15151392,
      "step": 17485
    },
    {
      "epoch": 8.246110325318247,
      "grad_norm": 0.0007319622673094273,
      "learning_rate": 0.17939063801367214,
      "loss": 0.217,
      "num_input_tokens_seen": 15156160,
      "step": 17490
    },
    {
      "epoch": 8.24846770391325,
      "grad_norm": 0.0014420924708247185,
      "learning_rate": 0.17933287267684483,
      "loss": 0.2457,
      "num_input_tokens_seen": 15159856,
      "step": 17495
    },
    {
      "epoch": 8.250825082508252,
      "grad_norm": 0.001195083255879581,
      "learning_rate": 0.17927510281651995,
      "loss": 0.2212,
      "num_input_tokens_seen": 15163392,
      "step": 17500
    },
    {
      "epoch": 8.253182461103254,
      "grad_norm": 0.002028409857302904,
      "learning_rate": 0.17921732844160634,
      "loss": 0.2395,
      "num_input_tokens_seen": 15167472,
      "step": 17505
    },
    {
      "epoch": 8.255539839698255,
      "grad_norm": 0.0003986610972788185,
      "learning_rate": 0.17915954956101351,
      "loss": 0.2278,
      "num_input_tokens_seen": 15171424,
      "step": 17510
    },
    {
      "epoch": 8.257897218293257,
      "grad_norm": 0.001574523514136672,
      "learning_rate": 0.17910176618365165,
      "loss": 0.2366,
      "num_input_tokens_seen": 15177968,
      "step": 17515
    },
    {
      "epoch": 8.26025459688826,
      "grad_norm": 0.0010000461479648948,
      "learning_rate": 0.17904397831843177,
      "loss": 0.2242,
      "num_input_tokens_seen": 15182432,
      "step": 17520
    },
    {
      "epoch": 8.262611975483262,
      "grad_norm": 0.0005931671475991607,
      "learning_rate": 0.17898618597426547,
      "loss": 0.2129,
      "num_input_tokens_seen": 15186352,
      "step": 17525
    },
    {
      "epoch": 8.264969354078264,
      "grad_norm": 0.001251478330232203,
      "learning_rate": 0.17892838916006495,
      "loss": 0.1803,
      "num_input_tokens_seen": 15191744,
      "step": 17530
    },
    {
      "epoch": 8.267326732673267,
      "grad_norm": 0.0009137343149632215,
      "learning_rate": 0.17887058788474333,
      "loss": 0.1468,
      "num_input_tokens_seen": 15196224,
      "step": 17535
    },
    {
      "epoch": 8.269684111268269,
      "grad_norm": 0.0007736209081485868,
      "learning_rate": 0.17881278215721427,
      "loss": 0.1954,
      "num_input_tokens_seen": 15201008,
      "step": 17540
    },
    {
      "epoch": 8.272041489863271,
      "grad_norm": 0.0014450362650677562,
      "learning_rate": 0.1787549719863921,
      "loss": 0.3063,
      "num_input_tokens_seen": 15205424,
      "step": 17545
    },
    {
      "epoch": 8.274398868458274,
      "grad_norm": 0.0004421088087838143,
      "learning_rate": 0.17869715738119188,
      "loss": 0.2827,
      "num_input_tokens_seen": 15209504,
      "step": 17550
    },
    {
      "epoch": 8.276756247053276,
      "grad_norm": 0.002065566834062338,
      "learning_rate": 0.17863933835052936,
      "loss": 0.2187,
      "num_input_tokens_seen": 15214256,
      "step": 17555
    },
    {
      "epoch": 8.279113625648279,
      "grad_norm": 0.002030690899118781,
      "learning_rate": 0.17858151490332097,
      "loss": 0.2252,
      "num_input_tokens_seen": 15218080,
      "step": 17560
    },
    {
      "epoch": 8.281471004243281,
      "grad_norm": 0.0010913773439824581,
      "learning_rate": 0.17852368704848381,
      "loss": 0.2253,
      "num_input_tokens_seen": 15221856,
      "step": 17565
    },
    {
      "epoch": 8.283828382838283,
      "grad_norm": 0.000587529968470335,
      "learning_rate": 0.17846585479493565,
      "loss": 0.2282,
      "num_input_tokens_seen": 15226192,
      "step": 17570
    },
    {
      "epoch": 8.286185761433286,
      "grad_norm": 0.0006090219249017537,
      "learning_rate": 0.178408018151595,
      "loss": 0.2343,
      "num_input_tokens_seen": 15230912,
      "step": 17575
    },
    {
      "epoch": 8.288543140028288,
      "grad_norm": 0.0041658394038677216,
      "learning_rate": 0.17835017712738085,
      "loss": 0.2207,
      "num_input_tokens_seen": 15236560,
      "step": 17580
    },
    {
      "epoch": 8.29090051862329,
      "grad_norm": 0.0014563678996637464,
      "learning_rate": 0.17829233173121323,
      "loss": 0.2296,
      "num_input_tokens_seen": 15241456,
      "step": 17585
    },
    {
      "epoch": 8.293257897218293,
      "grad_norm": 0.0011527508031576872,
      "learning_rate": 0.17823448197201244,
      "loss": 0.23,
      "num_input_tokens_seen": 15245744,
      "step": 17590
    },
    {
      "epoch": 8.295615275813295,
      "grad_norm": 0.000673772010486573,
      "learning_rate": 0.1781766278586997,
      "loss": 0.2308,
      "num_input_tokens_seen": 15249824,
      "step": 17595
    },
    {
      "epoch": 8.297972654408298,
      "grad_norm": 0.0013623657869175076,
      "learning_rate": 0.1781187694001969,
      "loss": 0.2175,
      "num_input_tokens_seen": 15254544,
      "step": 17600
    },
    {
      "epoch": 8.297972654408298,
      "eval_loss": 0.21614058315753937,
      "eval_runtime": 21.9041,
      "eval_samples_per_second": 43.051,
      "eval_steps_per_second": 21.548,
      "num_input_tokens_seen": 15254544,
      "step": 17600
    },
    {
      "epoch": 8.3003300330033,
      "grad_norm": 0.0015921415761113167,
      "learning_rate": 0.1780609066054265,
      "loss": 0.1982,
      "num_input_tokens_seen": 15258352,
      "step": 17605
    },
    {
      "epoch": 8.302687411598303,
      "grad_norm": 0.0007842971244826913,
      "learning_rate": 0.17800303948331164,
      "loss": 0.241,
      "num_input_tokens_seen": 15261856,
      "step": 17610
    },
    {
      "epoch": 8.305044790193305,
      "grad_norm": 0.0012833571527153254,
      "learning_rate": 0.1779451680427762,
      "loss": 0.2418,
      "num_input_tokens_seen": 15266240,
      "step": 17615
    },
    {
      "epoch": 8.307402168788308,
      "grad_norm": 0.0007401234470307827,
      "learning_rate": 0.17788729229274464,
      "loss": 0.2064,
      "num_input_tokens_seen": 15270000,
      "step": 17620
    },
    {
      "epoch": 8.30975954738331,
      "grad_norm": 0.000766481040045619,
      "learning_rate": 0.17782941224214222,
      "loss": 0.2025,
      "num_input_tokens_seen": 15274704,
      "step": 17625
    },
    {
      "epoch": 8.312116925978312,
      "grad_norm": 0.000734795059543103,
      "learning_rate": 0.17777152789989464,
      "loss": 0.2405,
      "num_input_tokens_seen": 15278448,
      "step": 17630
    },
    {
      "epoch": 8.314474304573315,
      "grad_norm": 0.001417364226654172,
      "learning_rate": 0.17771363927492845,
      "loss": 0.2083,
      "num_input_tokens_seen": 15282384,
      "step": 17635
    },
    {
      "epoch": 8.316831683168317,
      "grad_norm": 0.0008811577572487295,
      "learning_rate": 0.17765574637617085,
      "loss": 0.2216,
      "num_input_tokens_seen": 15286544,
      "step": 17640
    },
    {
      "epoch": 8.31918906176332,
      "grad_norm": 0.0009883709717541933,
      "learning_rate": 0.17759784921254962,
      "loss": 0.2032,
      "num_input_tokens_seen": 15290128,
      "step": 17645
    },
    {
      "epoch": 8.321546440358322,
      "grad_norm": 0.0017199733993038535,
      "learning_rate": 0.1775399477929932,
      "loss": 0.231,
      "num_input_tokens_seen": 15295424,
      "step": 17650
    },
    {
      "epoch": 8.323903818953324,
      "grad_norm": 0.002116607967764139,
      "learning_rate": 0.17748204212643076,
      "loss": 0.2032,
      "num_input_tokens_seen": 15299344,
      "step": 17655
    },
    {
      "epoch": 8.326261197548327,
      "grad_norm": 0.0005680570611730218,
      "learning_rate": 0.17742413222179204,
      "loss": 0.1751,
      "num_input_tokens_seen": 15303872,
      "step": 17660
    },
    {
      "epoch": 8.32861857614333,
      "grad_norm": 0.0005705593503080308,
      "learning_rate": 0.17736621808800754,
      "loss": 0.2437,
      "num_input_tokens_seen": 15307824,
      "step": 17665
    },
    {
      "epoch": 8.330975954738332,
      "grad_norm": 0.0007248824113048613,
      "learning_rate": 0.17730829973400827,
      "loss": 0.2552,
      "num_input_tokens_seen": 15311776,
      "step": 17670
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.0012016367400065064,
      "learning_rate": 0.17725037716872602,
      "loss": 0.2527,
      "num_input_tokens_seen": 15316752,
      "step": 17675
    },
    {
      "epoch": 8.335690711928336,
      "grad_norm": 0.0016305889002978802,
      "learning_rate": 0.17719245040109313,
      "loss": 0.2053,
      "num_input_tokens_seen": 15320512,
      "step": 17680
    },
    {
      "epoch": 8.338048090523339,
      "grad_norm": 0.0008650401141494513,
      "learning_rate": 0.17713451944004271,
      "loss": 0.2318,
      "num_input_tokens_seen": 15324512,
      "step": 17685
    },
    {
      "epoch": 8.340405469118341,
      "grad_norm": 0.0008470658212900162,
      "learning_rate": 0.17707658429450843,
      "loss": 0.2008,
      "num_input_tokens_seen": 15329552,
      "step": 17690
    },
    {
      "epoch": 8.342762847713344,
      "grad_norm": 0.0005339845665730536,
      "learning_rate": 0.1770186449734245,
      "loss": 0.2078,
      "num_input_tokens_seen": 15333744,
      "step": 17695
    },
    {
      "epoch": 8.345120226308346,
      "grad_norm": 0.0012775103095918894,
      "learning_rate": 0.17696070148572599,
      "loss": 0.1954,
      "num_input_tokens_seen": 15338000,
      "step": 17700
    },
    {
      "epoch": 8.347477604903348,
      "grad_norm": 0.0009918904397636652,
      "learning_rate": 0.17690275384034856,
      "loss": 0.206,
      "num_input_tokens_seen": 15341776,
      "step": 17705
    },
    {
      "epoch": 8.34983498349835,
      "grad_norm": 0.002046305686235428,
      "learning_rate": 0.17684480204622835,
      "loss": 0.2724,
      "num_input_tokens_seen": 15346016,
      "step": 17710
    },
    {
      "epoch": 8.352192362093351,
      "grad_norm": 0.0005304171936586499,
      "learning_rate": 0.1767868461123023,
      "loss": 0.2623,
      "num_input_tokens_seen": 15350016,
      "step": 17715
    },
    {
      "epoch": 8.354549740688354,
      "grad_norm": 0.0006565371295437217,
      "learning_rate": 0.176728886047508,
      "loss": 0.2344,
      "num_input_tokens_seen": 15353536,
      "step": 17720
    },
    {
      "epoch": 8.356907119283356,
      "grad_norm": 0.0006002521840855479,
      "learning_rate": 0.17667092186078362,
      "loss": 0.2449,
      "num_input_tokens_seen": 15357696,
      "step": 17725
    },
    {
      "epoch": 8.359264497878359,
      "grad_norm": 0.0008706214721314609,
      "learning_rate": 0.17661295356106785,
      "loss": 0.2312,
      "num_input_tokens_seen": 15361936,
      "step": 17730
    },
    {
      "epoch": 8.361621876473361,
      "grad_norm": 0.0012711867457255721,
      "learning_rate": 0.1765549811573002,
      "loss": 0.2289,
      "num_input_tokens_seen": 15366000,
      "step": 17735
    },
    {
      "epoch": 8.363979255068363,
      "grad_norm": 0.0008479977841489017,
      "learning_rate": 0.17649700465842078,
      "loss": 0.2317,
      "num_input_tokens_seen": 15369792,
      "step": 17740
    },
    {
      "epoch": 8.366336633663366,
      "grad_norm": 0.0006554234423674643,
      "learning_rate": 0.17643902407337023,
      "loss": 0.2221,
      "num_input_tokens_seen": 15374800,
      "step": 17745
    },
    {
      "epoch": 8.368694012258368,
      "grad_norm": 0.0006481492309831083,
      "learning_rate": 0.17638103941108993,
      "loss": 0.224,
      "num_input_tokens_seen": 15378976,
      "step": 17750
    },
    {
      "epoch": 8.37105139085337,
      "grad_norm": 0.0013533122837543488,
      "learning_rate": 0.1763230506805218,
      "loss": 0.217,
      "num_input_tokens_seen": 15383968,
      "step": 17755
    },
    {
      "epoch": 8.373408769448373,
      "grad_norm": 0.0005582069279626012,
      "learning_rate": 0.1762650578906085,
      "loss": 0.2361,
      "num_input_tokens_seen": 15387776,
      "step": 17760
    },
    {
      "epoch": 8.375766148043375,
      "grad_norm": 0.0005073976935818791,
      "learning_rate": 0.1762070610502932,
      "loss": 0.2298,
      "num_input_tokens_seen": 15392624,
      "step": 17765
    },
    {
      "epoch": 8.378123526638378,
      "grad_norm": 0.00040189785067923367,
      "learning_rate": 0.17614906016851975,
      "loss": 0.2253,
      "num_input_tokens_seen": 15396768,
      "step": 17770
    },
    {
      "epoch": 8.38048090523338,
      "grad_norm": 0.0006163145881146193,
      "learning_rate": 0.17609105525423258,
      "loss": 0.2276,
      "num_input_tokens_seen": 15400592,
      "step": 17775
    },
    {
      "epoch": 8.382838283828383,
      "grad_norm": 0.0034978787880390882,
      "learning_rate": 0.1760330463163768,
      "loss": 0.2302,
      "num_input_tokens_seen": 15405136,
      "step": 17780
    },
    {
      "epoch": 8.385195662423385,
      "grad_norm": 0.0015213626902550459,
      "learning_rate": 0.17597503336389816,
      "loss": 0.2323,
      "num_input_tokens_seen": 15409408,
      "step": 17785
    },
    {
      "epoch": 8.387553041018387,
      "grad_norm": 0.0006921447929926217,
      "learning_rate": 0.17591701640574298,
      "loss": 0.2125,
      "num_input_tokens_seen": 15414112,
      "step": 17790
    },
    {
      "epoch": 8.38991041961339,
      "grad_norm": 0.0012722890824079514,
      "learning_rate": 0.17585899545085815,
      "loss": 0.2256,
      "num_input_tokens_seen": 15418160,
      "step": 17795
    },
    {
      "epoch": 8.392267798208392,
      "grad_norm": 0.0013010416878387332,
      "learning_rate": 0.17580097050819124,
      "loss": 0.2334,
      "num_input_tokens_seen": 15422256,
      "step": 17800
    },
    {
      "epoch": 8.392267798208392,
      "eval_loss": 0.21769794821739197,
      "eval_runtime": 21.9375,
      "eval_samples_per_second": 42.986,
      "eval_steps_per_second": 21.516,
      "num_input_tokens_seen": 15422256,
      "step": 17800
    },
    {
      "epoch": 8.394625176803395,
      "grad_norm": 0.0007857599412091076,
      "learning_rate": 0.17574294158669046,
      "loss": 0.2206,
      "num_input_tokens_seen": 15426528,
      "step": 17805
    },
    {
      "epoch": 8.396982555398397,
      "grad_norm": 0.00071547063998878,
      "learning_rate": 0.17568490869530456,
      "loss": 0.2268,
      "num_input_tokens_seen": 15430704,
      "step": 17810
    },
    {
      "epoch": 8.3993399339934,
      "grad_norm": 0.0007274668896570802,
      "learning_rate": 0.17562687184298295,
      "loss": 0.1982,
      "num_input_tokens_seen": 15434880,
      "step": 17815
    },
    {
      "epoch": 8.401697312588402,
      "grad_norm": 0.0005184274050407112,
      "learning_rate": 0.1755688310386757,
      "loss": 0.2271,
      "num_input_tokens_seen": 15438992,
      "step": 17820
    },
    {
      "epoch": 8.404054691183404,
      "grad_norm": 0.0014656211715191603,
      "learning_rate": 0.17551078629133335,
      "loss": 0.2391,
      "num_input_tokens_seen": 15443600,
      "step": 17825
    },
    {
      "epoch": 8.406412069778407,
      "grad_norm": 0.0006041778833605349,
      "learning_rate": 0.17545273760990718,
      "loss": 0.2435,
      "num_input_tokens_seen": 15448688,
      "step": 17830
    },
    {
      "epoch": 8.408769448373409,
      "grad_norm": 0.0013833378907293081,
      "learning_rate": 0.17539468500334904,
      "loss": 0.2207,
      "num_input_tokens_seen": 15452880,
      "step": 17835
    },
    {
      "epoch": 8.411126826968411,
      "grad_norm": 0.0007196586229838431,
      "learning_rate": 0.17533662848061132,
      "loss": 0.2291,
      "num_input_tokens_seen": 15457488,
      "step": 17840
    },
    {
      "epoch": 8.413484205563414,
      "grad_norm": 0.0005932339117862284,
      "learning_rate": 0.1752785680506471,
      "loss": 0.2166,
      "num_input_tokens_seen": 15461520,
      "step": 17845
    },
    {
      "epoch": 8.415841584158416,
      "grad_norm": 0.0006514521664939821,
      "learning_rate": 0.17522050372241,
      "loss": 0.215,
      "num_input_tokens_seen": 15465824,
      "step": 17850
    },
    {
      "epoch": 8.418198962753419,
      "grad_norm": 0.0013751299120485783,
      "learning_rate": 0.17516243550485425,
      "loss": 0.2075,
      "num_input_tokens_seen": 15470368,
      "step": 17855
    },
    {
      "epoch": 8.420556341348421,
      "grad_norm": 0.0006146053201518953,
      "learning_rate": 0.17510436340693478,
      "loss": 0.255,
      "num_input_tokens_seen": 15475168,
      "step": 17860
    },
    {
      "epoch": 8.422913719943423,
      "grad_norm": 0.0008043135749176145,
      "learning_rate": 0.175046287437607,
      "loss": 0.2348,
      "num_input_tokens_seen": 15479056,
      "step": 17865
    },
    {
      "epoch": 8.425271098538426,
      "grad_norm": 0.0006233906606212258,
      "learning_rate": 0.17498820760582695,
      "loss": 0.2273,
      "num_input_tokens_seen": 15484176,
      "step": 17870
    },
    {
      "epoch": 8.427628477133428,
      "grad_norm": 0.0007953964523039758,
      "learning_rate": 0.1749301239205512,
      "loss": 0.2182,
      "num_input_tokens_seen": 15487952,
      "step": 17875
    },
    {
      "epoch": 8.42998585572843,
      "grad_norm": 0.0007979943184182048,
      "learning_rate": 0.1748720363907371,
      "loss": 0.2276,
      "num_input_tokens_seen": 15492432,
      "step": 17880
    },
    {
      "epoch": 8.432343234323433,
      "grad_norm": 0.0018163628410547972,
      "learning_rate": 0.17481394502534242,
      "loss": 0.233,
      "num_input_tokens_seen": 15497120,
      "step": 17885
    },
    {
      "epoch": 8.434700612918435,
      "grad_norm": 0.0005012198816984892,
      "learning_rate": 0.17475584983332562,
      "loss": 0.2155,
      "num_input_tokens_seen": 15502160,
      "step": 17890
    },
    {
      "epoch": 8.437057991513438,
      "grad_norm": 0.0007532468298450112,
      "learning_rate": 0.17469775082364558,
      "loss": 0.2328,
      "num_input_tokens_seen": 15505920,
      "step": 17895
    },
    {
      "epoch": 8.43941537010844,
      "grad_norm": 0.0006603209185414016,
      "learning_rate": 0.17463964800526205,
      "loss": 0.2119,
      "num_input_tokens_seen": 15510432,
      "step": 17900
    },
    {
      "epoch": 8.441772748703443,
      "grad_norm": 0.0006522470503114164,
      "learning_rate": 0.17458154138713522,
      "loss": 0.1916,
      "num_input_tokens_seen": 15514688,
      "step": 17905
    },
    {
      "epoch": 8.444130127298443,
      "grad_norm": 0.0005579644348472357,
      "learning_rate": 0.17452343097822576,
      "loss": 0.2034,
      "num_input_tokens_seen": 15519424,
      "step": 17910
    },
    {
      "epoch": 8.446487505893446,
      "grad_norm": 0.0011432182509452105,
      "learning_rate": 0.17446531678749497,
      "loss": 0.2051,
      "num_input_tokens_seen": 15522912,
      "step": 17915
    },
    {
      "epoch": 8.448844884488448,
      "grad_norm": 0.001259755459614098,
      "learning_rate": 0.17440719882390496,
      "loss": 0.2582,
      "num_input_tokens_seen": 15526928,
      "step": 17920
    },
    {
      "epoch": 8.45120226308345,
      "grad_norm": 0.0010582468239590526,
      "learning_rate": 0.17434907709641814,
      "loss": 0.1299,
      "num_input_tokens_seen": 15530944,
      "step": 17925
    },
    {
      "epoch": 8.453559641678453,
      "grad_norm": 0.002214054809883237,
      "learning_rate": 0.17429095161399769,
      "loss": 0.2595,
      "num_input_tokens_seen": 15535408,
      "step": 17930
    },
    {
      "epoch": 8.455917020273455,
      "grad_norm": 0.0005403182585723698,
      "learning_rate": 0.1742328223856072,
      "loss": 0.1764,
      "num_input_tokens_seen": 15540544,
      "step": 17935
    },
    {
      "epoch": 8.458274398868458,
      "grad_norm": 0.0005449664895422757,
      "learning_rate": 0.174174689420211,
      "loss": 0.2507,
      "num_input_tokens_seen": 15544112,
      "step": 17940
    },
    {
      "epoch": 8.46063177746346,
      "grad_norm": 0.0004789628437720239,
      "learning_rate": 0.1741165527267739,
      "loss": 0.1652,
      "num_input_tokens_seen": 15548560,
      "step": 17945
    },
    {
      "epoch": 8.462989156058462,
      "grad_norm": 0.0010395179269835353,
      "learning_rate": 0.17405841231426125,
      "loss": 0.1884,
      "num_input_tokens_seen": 15553920,
      "step": 17950
    },
    {
      "epoch": 8.465346534653465,
      "grad_norm": 0.0008846030104905367,
      "learning_rate": 0.1740002681916391,
      "loss": 0.2387,
      "num_input_tokens_seen": 15557712,
      "step": 17955
    },
    {
      "epoch": 8.467703913248467,
      "grad_norm": 0.0008814088650979102,
      "learning_rate": 0.17394212036787401,
      "loss": 0.2507,
      "num_input_tokens_seen": 15562544,
      "step": 17960
    },
    {
      "epoch": 8.47006129184347,
      "grad_norm": 0.0015101421158760786,
      "learning_rate": 0.1738839688519331,
      "loss": 0.2085,
      "num_input_tokens_seen": 15566416,
      "step": 17965
    },
    {
      "epoch": 8.472418670438472,
      "grad_norm": 0.0006444223108701408,
      "learning_rate": 0.17382581365278402,
      "loss": 0.2332,
      "num_input_tokens_seen": 15571792,
      "step": 17970
    },
    {
      "epoch": 8.474776049033474,
      "grad_norm": 0.00211532530374825,
      "learning_rate": 0.17376765477939507,
      "loss": 0.2356,
      "num_input_tokens_seen": 15575968,
      "step": 17975
    },
    {
      "epoch": 8.477133427628477,
      "grad_norm": 0.001247377716936171,
      "learning_rate": 0.1737094922407351,
      "loss": 0.2461,
      "num_input_tokens_seen": 15579296,
      "step": 17980
    },
    {
      "epoch": 8.47949080622348,
      "grad_norm": 0.001665474264882505,
      "learning_rate": 0.1736513260457734,
      "loss": 0.2235,
      "num_input_tokens_seen": 15583408,
      "step": 17985
    },
    {
      "epoch": 8.481848184818482,
      "grad_norm": 0.0011555610690265894,
      "learning_rate": 0.17359315620348006,
      "loss": 0.2272,
      "num_input_tokens_seen": 15587680,
      "step": 17990
    },
    {
      "epoch": 8.484205563413484,
      "grad_norm": 0.0011111653875559568,
      "learning_rate": 0.17353498272282547,
      "loss": 0.2365,
      "num_input_tokens_seen": 15591344,
      "step": 17995
    },
    {
      "epoch": 8.486562942008486,
      "grad_norm": 0.0018261030782014132,
      "learning_rate": 0.17347680561278087,
      "loss": 0.2287,
      "num_input_tokens_seen": 15595776,
      "step": 18000
    },
    {
      "epoch": 8.486562942008486,
      "eval_loss": 0.2266455441713333,
      "eval_runtime": 21.9511,
      "eval_samples_per_second": 42.959,
      "eval_steps_per_second": 21.502,
      "num_input_tokens_seen": 15595776,
      "step": 18000
    },
    {
      "epoch": 8.488920320603489,
      "grad_norm": 0.0012799183605238795,
      "learning_rate": 0.1734186248823178,
      "loss": 0.2289,
      "num_input_tokens_seen": 15600048,
      "step": 18005
    },
    {
      "epoch": 8.491277699198491,
      "grad_norm": 0.0013307837070897222,
      "learning_rate": 0.17336044054040844,
      "loss": 0.2267,
      "num_input_tokens_seen": 15604464,
      "step": 18010
    },
    {
      "epoch": 8.493635077793494,
      "grad_norm": 0.0006515239365398884,
      "learning_rate": 0.1733022525960256,
      "loss": 0.215,
      "num_input_tokens_seen": 15608880,
      "step": 18015
    },
    {
      "epoch": 8.495992456388496,
      "grad_norm": 0.0006984287756495178,
      "learning_rate": 0.1732440610581426,
      "loss": 0.2066,
      "num_input_tokens_seen": 15613184,
      "step": 18020
    },
    {
      "epoch": 8.498349834983498,
      "grad_norm": 0.0007471186690963805,
      "learning_rate": 0.17318586593573326,
      "loss": 0.2164,
      "num_input_tokens_seen": 15616768,
      "step": 18025
    },
    {
      "epoch": 8.500707213578501,
      "grad_norm": 0.0005731939454562962,
      "learning_rate": 0.17312766723777204,
      "loss": 0.1775,
      "num_input_tokens_seen": 15620656,
      "step": 18030
    },
    {
      "epoch": 8.503064592173503,
      "grad_norm": 0.0011987403267994523,
      "learning_rate": 0.1730694649732339,
      "loss": 0.2684,
      "num_input_tokens_seen": 15624624,
      "step": 18035
    },
    {
      "epoch": 8.505421970768506,
      "grad_norm": 0.0010582548566162586,
      "learning_rate": 0.17301125915109428,
      "loss": 0.1788,
      "num_input_tokens_seen": 15629008,
      "step": 18040
    },
    {
      "epoch": 8.507779349363508,
      "grad_norm": 0.001136177801527083,
      "learning_rate": 0.17295304978032938,
      "loss": 0.2469,
      "num_input_tokens_seen": 15633984,
      "step": 18045
    },
    {
      "epoch": 8.51013672795851,
      "grad_norm": 0.0007597638759762049,
      "learning_rate": 0.17289483686991577,
      "loss": 0.2483,
      "num_input_tokens_seen": 15638064,
      "step": 18050
    },
    {
      "epoch": 8.512494106553513,
      "grad_norm": 0.0009094870765693486,
      "learning_rate": 0.1728366204288306,
      "loss": 0.218,
      "num_input_tokens_seen": 15642080,
      "step": 18055
    },
    {
      "epoch": 8.514851485148515,
      "grad_norm": 0.0015877906698733568,
      "learning_rate": 0.17277840046605153,
      "loss": 0.2355,
      "num_input_tokens_seen": 15647120,
      "step": 18060
    },
    {
      "epoch": 8.517208863743518,
      "grad_norm": 0.0012023497838526964,
      "learning_rate": 0.17272017699055686,
      "loss": 0.2225,
      "num_input_tokens_seen": 15651296,
      "step": 18065
    },
    {
      "epoch": 8.51956624233852,
      "grad_norm": 0.0007763237808831036,
      "learning_rate": 0.17266195001132542,
      "loss": 0.2111,
      "num_input_tokens_seen": 15655088,
      "step": 18070
    },
    {
      "epoch": 8.521923620933523,
      "grad_norm": 0.0007487460970878601,
      "learning_rate": 0.17260371953733647,
      "loss": 0.2101,
      "num_input_tokens_seen": 15659600,
      "step": 18075
    },
    {
      "epoch": 8.524280999528525,
      "grad_norm": 0.0006525563658215106,
      "learning_rate": 0.1725454855775699,
      "loss": 0.2143,
      "num_input_tokens_seen": 15663648,
      "step": 18080
    },
    {
      "epoch": 8.526638378123527,
      "grad_norm": 0.0005706198862753808,
      "learning_rate": 0.17248724814100616,
      "loss": 0.1839,
      "num_input_tokens_seen": 15668224,
      "step": 18085
    },
    {
      "epoch": 8.52899575671853,
      "grad_norm": 0.0009220067877322435,
      "learning_rate": 0.17242900723662619,
      "loss": 0.2287,
      "num_input_tokens_seen": 15672128,
      "step": 18090
    },
    {
      "epoch": 8.531353135313532,
      "grad_norm": 0.000576293095946312,
      "learning_rate": 0.1723707628734114,
      "loss": 0.2357,
      "num_input_tokens_seen": 15675712,
      "step": 18095
    },
    {
      "epoch": 8.533710513908535,
      "grad_norm": 0.0009397325338795781,
      "learning_rate": 0.1723125150603438,
      "loss": 0.232,
      "num_input_tokens_seen": 15680560,
      "step": 18100
    },
    {
      "epoch": 8.536067892503535,
      "grad_norm": 0.014482175931334496,
      "learning_rate": 0.1722542638064061,
      "loss": 0.2429,
      "num_input_tokens_seen": 15685008,
      "step": 18105
    },
    {
      "epoch": 8.53842527109854,
      "grad_norm": 0.0011561346473172307,
      "learning_rate": 0.17219600912058117,
      "loss": 0.2263,
      "num_input_tokens_seen": 15689344,
      "step": 18110
    },
    {
      "epoch": 8.54078264969354,
      "grad_norm": 0.0009198483894579113,
      "learning_rate": 0.17213775101185272,
      "loss": 0.2238,
      "num_input_tokens_seen": 15693296,
      "step": 18115
    },
    {
      "epoch": 8.543140028288542,
      "grad_norm": 0.00163687311578542,
      "learning_rate": 0.17207948948920485,
      "loss": 0.2042,
      "num_input_tokens_seen": 15697264,
      "step": 18120
    },
    {
      "epoch": 8.545497406883545,
      "grad_norm": 0.0005884267156943679,
      "learning_rate": 0.17202122456162228,
      "loss": 0.1856,
      "num_input_tokens_seen": 15702704,
      "step": 18125
    },
    {
      "epoch": 8.547854785478547,
      "grad_norm": 0.0009090225212275982,
      "learning_rate": 0.17196295623809013,
      "loss": 0.3012,
      "num_input_tokens_seen": 15707376,
      "step": 18130
    },
    {
      "epoch": 8.55021216407355,
      "grad_norm": 0.001534249517135322,
      "learning_rate": 0.1719046845275941,
      "loss": 0.2534,
      "num_input_tokens_seen": 15711408,
      "step": 18135
    },
    {
      "epoch": 8.552569542668552,
      "grad_norm": 0.00697338255122304,
      "learning_rate": 0.17184640943912044,
      "loss": 0.2148,
      "num_input_tokens_seen": 15716096,
      "step": 18140
    },
    {
      "epoch": 8.554926921263554,
      "grad_norm": 0.0008019887609407306,
      "learning_rate": 0.1717881309816559,
      "loss": 0.2391,
      "num_input_tokens_seen": 15720656,
      "step": 18145
    },
    {
      "epoch": 8.557284299858557,
      "grad_norm": 0.0009679835056886077,
      "learning_rate": 0.1717298491641878,
      "loss": 0.2368,
      "num_input_tokens_seen": 15724176,
      "step": 18150
    },
    {
      "epoch": 8.55964167845356,
      "grad_norm": 0.001741203828714788,
      "learning_rate": 0.17167156399570385,
      "loss": 0.2105,
      "num_input_tokens_seen": 15728144,
      "step": 18155
    },
    {
      "epoch": 8.561999057048562,
      "grad_norm": 0.0017620599828660488,
      "learning_rate": 0.17161327548519242,
      "loss": 0.2011,
      "num_input_tokens_seen": 15733328,
      "step": 18160
    },
    {
      "epoch": 8.564356435643564,
      "grad_norm": 0.0012654566671699286,
      "learning_rate": 0.1715549836416423,
      "loss": 0.2087,
      "num_input_tokens_seen": 15737712,
      "step": 18165
    },
    {
      "epoch": 8.566713814238566,
      "grad_norm": 0.0012035195250064135,
      "learning_rate": 0.17149668847404279,
      "loss": 0.29,
      "num_input_tokens_seen": 15741952,
      "step": 18170
    },
    {
      "epoch": 8.569071192833569,
      "grad_norm": 0.0030324305407702923,
      "learning_rate": 0.1714383899913838,
      "loss": 0.2007,
      "num_input_tokens_seen": 15746464,
      "step": 18175
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 0.00086741684935987,
      "learning_rate": 0.17138008820265563,
      "loss": 0.2471,
      "num_input_tokens_seen": 15750480,
      "step": 18180
    },
    {
      "epoch": 8.573785950023574,
      "grad_norm": 0.0009010234498418868,
      "learning_rate": 0.17132178311684917,
      "loss": 0.2101,
      "num_input_tokens_seen": 15754736,
      "step": 18185
    },
    {
      "epoch": 8.576143328618576,
      "grad_norm": 0.0008850029553286731,
      "learning_rate": 0.1712634747429559,
      "loss": 0.2393,
      "num_input_tokens_seen": 15759280,
      "step": 18190
    },
    {
      "epoch": 8.578500707213578,
      "grad_norm": 0.0007456456078216434,
      "learning_rate": 0.17120516308996753,
      "loss": 0.2251,
      "num_input_tokens_seen": 15763312,
      "step": 18195
    },
    {
      "epoch": 8.58085808580858,
      "grad_norm": 0.0010238420218229294,
      "learning_rate": 0.17114684816687653,
      "loss": 0.2202,
      "num_input_tokens_seen": 15768288,
      "step": 18200
    },
    {
      "epoch": 8.58085808580858,
      "eval_loss": 0.219070166349411,
      "eval_runtime": 21.9087,
      "eval_samples_per_second": 43.042,
      "eval_steps_per_second": 21.544,
      "num_input_tokens_seen": 15768288,
      "step": 18200
    },
    {
      "epoch": 8.583215464403583,
      "grad_norm": 0.0007150620222091675,
      "learning_rate": 0.17108852998267585,
      "loss": 0.2247,
      "num_input_tokens_seen": 15772000,
      "step": 18205
    },
    {
      "epoch": 8.585572842998586,
      "grad_norm": 0.0005563875311054289,
      "learning_rate": 0.17103020854635878,
      "loss": 0.2364,
      "num_input_tokens_seen": 15775280,
      "step": 18210
    },
    {
      "epoch": 8.587930221593588,
      "grad_norm": 0.001316106179729104,
      "learning_rate": 0.1709718838669193,
      "loss": 0.2391,
      "num_input_tokens_seen": 15780464,
      "step": 18215
    },
    {
      "epoch": 8.59028760018859,
      "grad_norm": 0.0009191678836941719,
      "learning_rate": 0.17091355595335173,
      "loss": 0.221,
      "num_input_tokens_seen": 15785040,
      "step": 18220
    },
    {
      "epoch": 8.592644978783593,
      "grad_norm": 0.0007271976210176945,
      "learning_rate": 0.17085522481465107,
      "loss": 0.2217,
      "num_input_tokens_seen": 15789072,
      "step": 18225
    },
    {
      "epoch": 8.595002357378595,
      "grad_norm": 0.0008339803316630423,
      "learning_rate": 0.17079689045981264,
      "loss": 0.2226,
      "num_input_tokens_seen": 15793744,
      "step": 18230
    },
    {
      "epoch": 8.597359735973598,
      "grad_norm": 0.0007401821203529835,
      "learning_rate": 0.17073855289783238,
      "loss": 0.2519,
      "num_input_tokens_seen": 15798080,
      "step": 18235
    },
    {
      "epoch": 8.5997171145686,
      "grad_norm": 0.0005845606210641563,
      "learning_rate": 0.1706802121377066,
      "loss": 0.2165,
      "num_input_tokens_seen": 15802192,
      "step": 18240
    },
    {
      "epoch": 8.602074493163602,
      "grad_norm": 0.0015291826566681266,
      "learning_rate": 0.17062186818843225,
      "loss": 0.23,
      "num_input_tokens_seen": 15806464,
      "step": 18245
    },
    {
      "epoch": 8.604431871758605,
      "grad_norm": 0.001709967153146863,
      "learning_rate": 0.17056352105900668,
      "loss": 0.2268,
      "num_input_tokens_seen": 15811232,
      "step": 18250
    },
    {
      "epoch": 8.606789250353607,
      "grad_norm": 0.0005686606746166945,
      "learning_rate": 0.17050517075842772,
      "loss": 0.2246,
      "num_input_tokens_seen": 15815600,
      "step": 18255
    },
    {
      "epoch": 8.60914662894861,
      "grad_norm": 0.0006128829554654658,
      "learning_rate": 0.17044681729569375,
      "loss": 0.2431,
      "num_input_tokens_seen": 15819456,
      "step": 18260
    },
    {
      "epoch": 8.611504007543612,
      "grad_norm": 0.0005904178251512349,
      "learning_rate": 0.17038846067980365,
      "loss": 0.2208,
      "num_input_tokens_seen": 15823200,
      "step": 18265
    },
    {
      "epoch": 8.613861386138614,
      "grad_norm": 0.0011399361537769437,
      "learning_rate": 0.17033010091975664,
      "loss": 0.2323,
      "num_input_tokens_seen": 15828576,
      "step": 18270
    },
    {
      "epoch": 8.616218764733617,
      "grad_norm": 0.0016350378282368183,
      "learning_rate": 0.17027173802455262,
      "loss": 0.2061,
      "num_input_tokens_seen": 15832720,
      "step": 18275
    },
    {
      "epoch": 8.61857614332862,
      "grad_norm": 0.0006240732036530972,
      "learning_rate": 0.1702133720031918,
      "loss": 0.2173,
      "num_input_tokens_seen": 15837472,
      "step": 18280
    },
    {
      "epoch": 8.620933521923622,
      "grad_norm": 0.0007576360949315131,
      "learning_rate": 0.17015500286467503,
      "loss": 0.2129,
      "num_input_tokens_seen": 15841792,
      "step": 18285
    },
    {
      "epoch": 8.623290900518624,
      "grad_norm": 0.00043002175516448915,
      "learning_rate": 0.17009663061800354,
      "loss": 0.2077,
      "num_input_tokens_seen": 15846256,
      "step": 18290
    },
    {
      "epoch": 8.625648279113626,
      "grad_norm": 0.0008850198937579989,
      "learning_rate": 0.17003825527217903,
      "loss": 0.2193,
      "num_input_tokens_seen": 15850416,
      "step": 18295
    },
    {
      "epoch": 8.628005657708629,
      "grad_norm": 0.0007569605368189514,
      "learning_rate": 0.16997987683620377,
      "loss": 0.2623,
      "num_input_tokens_seen": 15854912,
      "step": 18300
    },
    {
      "epoch": 8.630363036303631,
      "grad_norm": 0.0006765298312529922,
      "learning_rate": 0.16992149531908043,
      "loss": 0.2293,
      "num_input_tokens_seen": 15859648,
      "step": 18305
    },
    {
      "epoch": 8.632720414898632,
      "grad_norm": 0.0009301642421633005,
      "learning_rate": 0.16986311072981214,
      "loss": 0.2356,
      "num_input_tokens_seen": 15863312,
      "step": 18310
    },
    {
      "epoch": 8.635077793493634,
      "grad_norm": 0.0005708853714168072,
      "learning_rate": 0.16980472307740255,
      "loss": 0.2704,
      "num_input_tokens_seen": 15868656,
      "step": 18315
    },
    {
      "epoch": 8.637435172088637,
      "grad_norm": 0.0004610781907103956,
      "learning_rate": 0.1697463323708558,
      "loss": 0.2358,
      "num_input_tokens_seen": 15872688,
      "step": 18320
    },
    {
      "epoch": 8.639792550683639,
      "grad_norm": 0.0005200142622925341,
      "learning_rate": 0.16968793861917641,
      "loss": 0.2278,
      "num_input_tokens_seen": 15876880,
      "step": 18325
    },
    {
      "epoch": 8.642149929278641,
      "grad_norm": 0.0006966299843043089,
      "learning_rate": 0.16962954183136952,
      "loss": 0.2247,
      "num_input_tokens_seen": 15881072,
      "step": 18330
    },
    {
      "epoch": 8.644507307873644,
      "grad_norm": 0.0014031528262421489,
      "learning_rate": 0.16957114201644058,
      "loss": 0.24,
      "num_input_tokens_seen": 15885120,
      "step": 18335
    },
    {
      "epoch": 8.646864686468646,
      "grad_norm": 0.0013513099402189255,
      "learning_rate": 0.16951273918339563,
      "loss": 0.209,
      "num_input_tokens_seen": 15889584,
      "step": 18340
    },
    {
      "epoch": 8.649222065063649,
      "grad_norm": 0.0007043613004498184,
      "learning_rate": 0.16945433334124105,
      "loss": 0.2267,
      "num_input_tokens_seen": 15893568,
      "step": 18345
    },
    {
      "epoch": 8.651579443658651,
      "grad_norm": 0.0013094445457682014,
      "learning_rate": 0.1693959244989838,
      "loss": 0.2176,
      "num_input_tokens_seen": 15897952,
      "step": 18350
    },
    {
      "epoch": 8.653936822253653,
      "grad_norm": 0.0007430961122736335,
      "learning_rate": 0.16933751266563127,
      "loss": 0.1973,
      "num_input_tokens_seen": 15901712,
      "step": 18355
    },
    {
      "epoch": 8.656294200848656,
      "grad_norm": 0.0008211465319618583,
      "learning_rate": 0.16927909785019118,
      "loss": 0.2055,
      "num_input_tokens_seen": 15906160,
      "step": 18360
    },
    {
      "epoch": 8.658651579443658,
      "grad_norm": 0.00038109449087642133,
      "learning_rate": 0.169220680061672,
      "loss": 0.1889,
      "num_input_tokens_seen": 15910256,
      "step": 18365
    },
    {
      "epoch": 8.66100895803866,
      "grad_norm": 0.0016713551012799144,
      "learning_rate": 0.16916225930908244,
      "loss": 0.2383,
      "num_input_tokens_seen": 15914592,
      "step": 18370
    },
    {
      "epoch": 8.663366336633663,
      "grad_norm": 0.0005188613431528211,
      "learning_rate": 0.16910383560143163,
      "loss": 0.1625,
      "num_input_tokens_seen": 15920224,
      "step": 18375
    },
    {
      "epoch": 8.665723715228665,
      "grad_norm": 0.001049801241606474,
      "learning_rate": 0.16904540894772935,
      "loss": 0.2608,
      "num_input_tokens_seen": 15924944,
      "step": 18380
    },
    {
      "epoch": 8.668081093823668,
      "grad_norm": 0.0015595387667417526,
      "learning_rate": 0.16898697935698562,
      "loss": 0.2254,
      "num_input_tokens_seen": 15929264,
      "step": 18385
    },
    {
      "epoch": 8.67043847241867,
      "grad_norm": 0.0006288376753218472,
      "learning_rate": 0.1689285468382111,
      "loss": 0.2103,
      "num_input_tokens_seen": 15933264,
      "step": 18390
    },
    {
      "epoch": 8.672795851013673,
      "grad_norm": 0.0012521749595180154,
      "learning_rate": 0.16887011140041677,
      "loss": 0.2058,
      "num_input_tokens_seen": 15937680,
      "step": 18395
    },
    {
      "epoch": 8.675153229608675,
      "grad_norm": 0.0004965679836459458,
      "learning_rate": 0.1688116730526141,
      "loss": 0.2279,
      "num_input_tokens_seen": 15941776,
      "step": 18400
    },
    {
      "epoch": 8.675153229608675,
      "eval_loss": 0.21795068681240082,
      "eval_runtime": 21.8939,
      "eval_samples_per_second": 43.071,
      "eval_steps_per_second": 21.559,
      "num_input_tokens_seen": 15941776,
      "step": 18400
    },
    {
      "epoch": 8.677510608203677,
      "grad_norm": 0.0014082754496484995,
      "learning_rate": 0.1687532318038151,
      "loss": 0.2047,
      "num_input_tokens_seen": 15945232,
      "step": 18405
    },
    {
      "epoch": 8.67986798679868,
      "grad_norm": 0.0014392426237463951,
      "learning_rate": 0.16869478766303206,
      "loss": 0.2363,
      "num_input_tokens_seen": 15948944,
      "step": 18410
    },
    {
      "epoch": 8.682225365393682,
      "grad_norm": 0.0015883008018136024,
      "learning_rate": 0.16863634063927788,
      "loss": 0.2113,
      "num_input_tokens_seen": 15952848,
      "step": 18415
    },
    {
      "epoch": 8.684582743988685,
      "grad_norm": 0.0004736390837933868,
      "learning_rate": 0.16857789074156568,
      "loss": 0.1829,
      "num_input_tokens_seen": 15957744,
      "step": 18420
    },
    {
      "epoch": 8.686940122583687,
      "grad_norm": 0.0005550820496864617,
      "learning_rate": 0.16851943797890928,
      "loss": 0.2316,
      "num_input_tokens_seen": 15962352,
      "step": 18425
    },
    {
      "epoch": 8.68929750117869,
      "grad_norm": 0.0008728143875487149,
      "learning_rate": 0.16846098236032284,
      "loss": 0.2552,
      "num_input_tokens_seen": 15967456,
      "step": 18430
    },
    {
      "epoch": 8.691654879773692,
      "grad_norm": 0.0004981233505532146,
      "learning_rate": 0.16840252389482097,
      "loss": 0.2198,
      "num_input_tokens_seen": 15971824,
      "step": 18435
    },
    {
      "epoch": 8.694012258368694,
      "grad_norm": 0.0006245926488190889,
      "learning_rate": 0.16834406259141857,
      "loss": 0.2227,
      "num_input_tokens_seen": 15976576,
      "step": 18440
    },
    {
      "epoch": 8.696369636963697,
      "grad_norm": 0.0006486743804998696,
      "learning_rate": 0.16828559845913124,
      "loss": 0.2398,
      "num_input_tokens_seen": 15980272,
      "step": 18445
    },
    {
      "epoch": 8.698727015558699,
      "grad_norm": 0.0006653268937952816,
      "learning_rate": 0.16822713150697488,
      "loss": 0.2014,
      "num_input_tokens_seen": 15985248,
      "step": 18450
    },
    {
      "epoch": 8.701084394153701,
      "grad_norm": 0.000704884878359735,
      "learning_rate": 0.16816866174396575,
      "loss": 0.2015,
      "num_input_tokens_seen": 15989104,
      "step": 18455
    },
    {
      "epoch": 8.703441772748704,
      "grad_norm": 0.0004543059039860964,
      "learning_rate": 0.16811018917912057,
      "loss": 0.2135,
      "num_input_tokens_seen": 15992736,
      "step": 18460
    },
    {
      "epoch": 8.705799151343706,
      "grad_norm": 0.0004836838343180716,
      "learning_rate": 0.16805171382145673,
      "loss": 0.2062,
      "num_input_tokens_seen": 15996912,
      "step": 18465
    },
    {
      "epoch": 8.708156529938709,
      "grad_norm": 0.0004295568505767733,
      "learning_rate": 0.16799323567999175,
      "loss": 0.2179,
      "num_input_tokens_seen": 16001568,
      "step": 18470
    },
    {
      "epoch": 8.710513908533711,
      "grad_norm": 0.0017301524057984352,
      "learning_rate": 0.16793475476374367,
      "loss": 0.2384,
      "num_input_tokens_seen": 16006288,
      "step": 18475
    },
    {
      "epoch": 8.712871287128714,
      "grad_norm": 0.0006898589199408889,
      "learning_rate": 0.1678762710817311,
      "loss": 0.2147,
      "num_input_tokens_seen": 16010976,
      "step": 18480
    },
    {
      "epoch": 8.715228665723716,
      "grad_norm": 0.0007041610078886151,
      "learning_rate": 0.1678177846429728,
      "loss": 0.2056,
      "num_input_tokens_seen": 16015840,
      "step": 18485
    },
    {
      "epoch": 8.717586044318718,
      "grad_norm": 0.0008175145485438406,
      "learning_rate": 0.16775929545648827,
      "loss": 0.174,
      "num_input_tokens_seen": 16019888,
      "step": 18490
    },
    {
      "epoch": 8.71994342291372,
      "grad_norm": 0.0011335917515680194,
      "learning_rate": 0.16770080353129715,
      "loss": 0.2524,
      "num_input_tokens_seen": 16023936,
      "step": 18495
    },
    {
      "epoch": 8.722300801508723,
      "grad_norm": 0.0008496858063153923,
      "learning_rate": 0.16764230887641968,
      "loss": 0.239,
      "num_input_tokens_seen": 16028208,
      "step": 18500
    },
    {
      "epoch": 8.724658180103724,
      "grad_norm": 0.001566729275509715,
      "learning_rate": 0.1675838115008765,
      "loss": 0.2237,
      "num_input_tokens_seen": 16032464,
      "step": 18505
    },
    {
      "epoch": 8.727015558698728,
      "grad_norm": 0.0008703620405867696,
      "learning_rate": 0.1675253114136886,
      "loss": 0.1987,
      "num_input_tokens_seen": 16036960,
      "step": 18510
    },
    {
      "epoch": 8.729372937293729,
      "grad_norm": 0.0011228607036173344,
      "learning_rate": 0.16746680862387747,
      "loss": 0.1883,
      "num_input_tokens_seen": 16042112,
      "step": 18515
    },
    {
      "epoch": 8.731730315888731,
      "grad_norm": 0.00037527523818425834,
      "learning_rate": 0.16740830314046493,
      "loss": 0.2017,
      "num_input_tokens_seen": 16045904,
      "step": 18520
    },
    {
      "epoch": 8.734087694483733,
      "grad_norm": 0.0003812809882219881,
      "learning_rate": 0.1673497949724733,
      "loss": 0.2248,
      "num_input_tokens_seen": 16049824,
      "step": 18525
    },
    {
      "epoch": 8.736445073078736,
      "grad_norm": 0.0007668108446523547,
      "learning_rate": 0.16729128412892522,
      "loss": 0.2269,
      "num_input_tokens_seen": 16053728,
      "step": 18530
    },
    {
      "epoch": 8.738802451673738,
      "grad_norm": 0.0011326927924528718,
      "learning_rate": 0.16723277061884384,
      "loss": 0.2094,
      "num_input_tokens_seen": 16058032,
      "step": 18535
    },
    {
      "epoch": 8.74115983026874,
      "grad_norm": 0.00044401141349226236,
      "learning_rate": 0.16717425445125267,
      "loss": 0.2029,
      "num_input_tokens_seen": 16062384,
      "step": 18540
    },
    {
      "epoch": 8.743517208863743,
      "grad_norm": 0.0004710571956820786,
      "learning_rate": 0.16711573563517565,
      "loss": 0.2414,
      "num_input_tokens_seen": 16067312,
      "step": 18545
    },
    {
      "epoch": 8.745874587458745,
      "grad_norm": 0.0015167276142165065,
      "learning_rate": 0.1670572141796371,
      "loss": 0.2186,
      "num_input_tokens_seen": 16071904,
      "step": 18550
    },
    {
      "epoch": 8.748231966053748,
      "grad_norm": 0.000627753499429673,
      "learning_rate": 0.16699869009366175,
      "loss": 0.2046,
      "num_input_tokens_seen": 16076880,
      "step": 18555
    },
    {
      "epoch": 8.75058934464875,
      "grad_norm": 0.0004848121025133878,
      "learning_rate": 0.1669401633862748,
      "loss": 0.223,
      "num_input_tokens_seen": 16080704,
      "step": 18560
    },
    {
      "epoch": 8.752946723243753,
      "grad_norm": 0.0007873569848015904,
      "learning_rate": 0.16688163406650178,
      "loss": 0.2463,
      "num_input_tokens_seen": 16084992,
      "step": 18565
    },
    {
      "epoch": 8.755304101838755,
      "grad_norm": 0.0006370774353854358,
      "learning_rate": 0.1668231021433686,
      "loss": 0.2359,
      "num_input_tokens_seen": 16090080,
      "step": 18570
    },
    {
      "epoch": 8.757661480433757,
      "grad_norm": 0.0005954314838163555,
      "learning_rate": 0.1667645676259017,
      "loss": 0.2413,
      "num_input_tokens_seen": 16094240,
      "step": 18575
    },
    {
      "epoch": 8.76001885902876,
      "grad_norm": 0.0004563817346934229,
      "learning_rate": 0.1667060305231277,
      "loss": 0.213,
      "num_input_tokens_seen": 16099184,
      "step": 18580
    },
    {
      "epoch": 8.762376237623762,
      "grad_norm": 0.0005711234407499433,
      "learning_rate": 0.16664749084407396,
      "loss": 0.2184,
      "num_input_tokens_seen": 16103680,
      "step": 18585
    },
    {
      "epoch": 8.764733616218765,
      "grad_norm": 0.0005423026741482317,
      "learning_rate": 0.16658894859776788,
      "loss": 0.2304,
      "num_input_tokens_seen": 16107536,
      "step": 18590
    },
    {
      "epoch": 8.767090994813767,
      "grad_norm": 0.0015658442862331867,
      "learning_rate": 0.16653040379323752,
      "loss": 0.2158,
      "num_input_tokens_seen": 16111104,
      "step": 18595
    },
    {
      "epoch": 8.76944837340877,
      "grad_norm": 0.0005919443792663515,
      "learning_rate": 0.16647185643951107,
      "loss": 0.2137,
      "num_input_tokens_seen": 16115152,
      "step": 18600
    },
    {
      "epoch": 8.76944837340877,
      "eval_loss": 0.217378169298172,
      "eval_runtime": 21.8988,
      "eval_samples_per_second": 43.062,
      "eval_steps_per_second": 21.554,
      "num_input_tokens_seen": 16115152,
      "step": 18600
    },
    {
      "epoch": 8.771805752003772,
      "grad_norm": 0.0004206119629088789,
      "learning_rate": 0.1664133065456174,
      "loss": 0.2107,
      "num_input_tokens_seen": 16119024,
      "step": 18605
    },
    {
      "epoch": 8.774163130598774,
      "grad_norm": 0.001129704061895609,
      "learning_rate": 0.1663547541205856,
      "loss": 0.2279,
      "num_input_tokens_seen": 16123632,
      "step": 18610
    },
    {
      "epoch": 8.776520509193777,
      "grad_norm": 0.0008168525528162718,
      "learning_rate": 0.16629619917344518,
      "loss": 0.2286,
      "num_input_tokens_seen": 16127584,
      "step": 18615
    },
    {
      "epoch": 8.778877887788779,
      "grad_norm": 0.0014294319553300738,
      "learning_rate": 0.16623764171322605,
      "loss": 0.2053,
      "num_input_tokens_seen": 16131536,
      "step": 18620
    },
    {
      "epoch": 8.781235266383781,
      "grad_norm": 0.0014796837931498885,
      "learning_rate": 0.1661790817489585,
      "loss": 0.2725,
      "num_input_tokens_seen": 16136960,
      "step": 18625
    },
    {
      "epoch": 8.783592644978784,
      "grad_norm": 0.0008465485298074782,
      "learning_rate": 0.16612051928967328,
      "loss": 0.2209,
      "num_input_tokens_seen": 16140784,
      "step": 18630
    },
    {
      "epoch": 8.785950023573786,
      "grad_norm": 0.0007864964427426457,
      "learning_rate": 0.16606195434440138,
      "loss": 0.2044,
      "num_input_tokens_seen": 16144464,
      "step": 18635
    },
    {
      "epoch": 8.788307402168789,
      "grad_norm": 0.0008262985502369702,
      "learning_rate": 0.16600338692217426,
      "loss": 0.2175,
      "num_input_tokens_seen": 16148048,
      "step": 18640
    },
    {
      "epoch": 8.790664780763791,
      "grad_norm": 0.000739801733288914,
      "learning_rate": 0.16594481703202374,
      "loss": 0.2154,
      "num_input_tokens_seen": 16152016,
      "step": 18645
    },
    {
      "epoch": 8.793022159358793,
      "grad_norm": 0.0005678630550391972,
      "learning_rate": 0.1658862446829821,
      "loss": 0.2131,
      "num_input_tokens_seen": 16156848,
      "step": 18650
    },
    {
      "epoch": 8.795379537953796,
      "grad_norm": 0.0007014997536316514,
      "learning_rate": 0.16582766988408187,
      "loss": 0.2233,
      "num_input_tokens_seen": 16160960,
      "step": 18655
    },
    {
      "epoch": 8.797736916548798,
      "grad_norm": 0.0005888852756470442,
      "learning_rate": 0.16576909264435608,
      "loss": 0.2107,
      "num_input_tokens_seen": 16164960,
      "step": 18660
    },
    {
      "epoch": 8.8000942951438,
      "grad_norm": 0.001299911760725081,
      "learning_rate": 0.16571051297283798,
      "loss": 0.2076,
      "num_input_tokens_seen": 16169504,
      "step": 18665
    },
    {
      "epoch": 8.802451673738803,
      "grad_norm": 0.0006208215490914881,
      "learning_rate": 0.16565193087856137,
      "loss": 0.2086,
      "num_input_tokens_seen": 16173168,
      "step": 18670
    },
    {
      "epoch": 8.804809052333805,
      "grad_norm": 0.0011324415681883693,
      "learning_rate": 0.16559334637056033,
      "loss": 0.2036,
      "num_input_tokens_seen": 16177696,
      "step": 18675
    },
    {
      "epoch": 8.807166430928808,
      "grad_norm": 0.0009565473301336169,
      "learning_rate": 0.16553475945786933,
      "loss": 0.2283,
      "num_input_tokens_seen": 16182000,
      "step": 18680
    },
    {
      "epoch": 8.80952380952381,
      "grad_norm": 0.0007318815332837403,
      "learning_rate": 0.16547617014952318,
      "loss": 0.2125,
      "num_input_tokens_seen": 16186544,
      "step": 18685
    },
    {
      "epoch": 8.811881188118813,
      "grad_norm": 0.0008638911531306803,
      "learning_rate": 0.1654175784545571,
      "loss": 0.1975,
      "num_input_tokens_seen": 16191216,
      "step": 18690
    },
    {
      "epoch": 8.814238566713815,
      "grad_norm": 0.0016324676107615232,
      "learning_rate": 0.1653589843820067,
      "loss": 0.2472,
      "num_input_tokens_seen": 16196160,
      "step": 18695
    },
    {
      "epoch": 8.816595945308817,
      "grad_norm": 0.0007655143272131681,
      "learning_rate": 0.1653003879409079,
      "loss": 0.2305,
      "num_input_tokens_seen": 16200160,
      "step": 18700
    },
    {
      "epoch": 8.81895332390382,
      "grad_norm": 0.0015668487176299095,
      "learning_rate": 0.165241789140297,
      "loss": 0.2207,
      "num_input_tokens_seen": 16203376,
      "step": 18705
    },
    {
      "epoch": 8.82131070249882,
      "grad_norm": 0.00101180886849761,
      "learning_rate": 0.16518318798921064,
      "loss": 0.2099,
      "num_input_tokens_seen": 16207472,
      "step": 18710
    },
    {
      "epoch": 8.823668081093825,
      "grad_norm": 0.0016081476351246238,
      "learning_rate": 0.16512458449668593,
      "loss": 0.2296,
      "num_input_tokens_seen": 16212144,
      "step": 18715
    },
    {
      "epoch": 8.826025459688825,
      "grad_norm": 0.0006952566327527165,
      "learning_rate": 0.1650659786717602,
      "loss": 0.214,
      "num_input_tokens_seen": 16216432,
      "step": 18720
    },
    {
      "epoch": 8.828382838283828,
      "grad_norm": 0.004767382517457008,
      "learning_rate": 0.1650073705234712,
      "loss": 0.2681,
      "num_input_tokens_seen": 16220848,
      "step": 18725
    },
    {
      "epoch": 8.83074021687883,
      "grad_norm": 0.0005493024364113808,
      "learning_rate": 0.16494876006085712,
      "loss": 0.2304,
      "num_input_tokens_seen": 16224768,
      "step": 18730
    },
    {
      "epoch": 8.833097595473832,
      "grad_norm": 0.0005917272064834833,
      "learning_rate": 0.16489014729295634,
      "loss": 0.2137,
      "num_input_tokens_seen": 16228896,
      "step": 18735
    },
    {
      "epoch": 8.835454974068835,
      "grad_norm": 0.0008148672641254961,
      "learning_rate": 0.16483153222880775,
      "loss": 0.2355,
      "num_input_tokens_seen": 16233088,
      "step": 18740
    },
    {
      "epoch": 8.837812352663837,
      "grad_norm": 0.0009700177470222116,
      "learning_rate": 0.16477291487745052,
      "loss": 0.2166,
      "num_input_tokens_seen": 16237648,
      "step": 18745
    },
    {
      "epoch": 8.84016973125884,
      "grad_norm": 0.001340088085271418,
      "learning_rate": 0.16471429524792416,
      "loss": 0.2088,
      "num_input_tokens_seen": 16241824,
      "step": 18750
    },
    {
      "epoch": 8.842527109853842,
      "grad_norm": 0.0008712430135346949,
      "learning_rate": 0.16465567334926856,
      "loss": 0.1874,
      "num_input_tokens_seen": 16245856,
      "step": 18755
    },
    {
      "epoch": 8.844884488448844,
      "grad_norm": 0.001651065074838698,
      "learning_rate": 0.16459704919052395,
      "loss": 0.2256,
      "num_input_tokens_seen": 16250480,
      "step": 18760
    },
    {
      "epoch": 8.847241867043847,
      "grad_norm": 0.0011034810449928045,
      "learning_rate": 0.16453842278073086,
      "loss": 0.2442,
      "num_input_tokens_seen": 16254864,
      "step": 18765
    },
    {
      "epoch": 8.84959924563885,
      "grad_norm": 0.0005329998675733805,
      "learning_rate": 0.16447979412893038,
      "loss": 0.2389,
      "num_input_tokens_seen": 16258704,
      "step": 18770
    },
    {
      "epoch": 8.851956624233852,
      "grad_norm": 0.0006073230179026723,
      "learning_rate": 0.16442116324416367,
      "loss": 0.2346,
      "num_input_tokens_seen": 16262864,
      "step": 18775
    },
    {
      "epoch": 8.854314002828854,
      "grad_norm": 0.0010504636447876692,
      "learning_rate": 0.1643625301354723,
      "loss": 0.226,
      "num_input_tokens_seen": 16267632,
      "step": 18780
    },
    {
      "epoch": 8.856671381423856,
      "grad_norm": 0.000914181349799037,
      "learning_rate": 0.16430389481189828,
      "loss": 0.2193,
      "num_input_tokens_seen": 16271552,
      "step": 18785
    },
    {
      "epoch": 8.859028760018859,
      "grad_norm": 0.0009132477571256459,
      "learning_rate": 0.164245257282484,
      "loss": 0.212,
      "num_input_tokens_seen": 16275680,
      "step": 18790
    },
    {
      "epoch": 8.861386138613861,
      "grad_norm": 0.0006554905558004975,
      "learning_rate": 0.16418661755627195,
      "loss": 0.2112,
      "num_input_tokens_seen": 16279744,
      "step": 18795
    },
    {
      "epoch": 8.863743517208864,
      "grad_norm": 0.00128890888299793,
      "learning_rate": 0.16412797564230527,
      "loss": 0.2182,
      "num_input_tokens_seen": 16284384,
      "step": 18800
    },
    {
      "epoch": 8.863743517208864,
      "eval_loss": 0.2159968912601471,
      "eval_runtime": 21.9784,
      "eval_samples_per_second": 42.906,
      "eval_steps_per_second": 21.476,
      "num_input_tokens_seen": 16284384,
      "step": 18800
    },
    {
      "epoch": 8.866100895803866,
      "grad_norm": 0.0006238169153220952,
      "learning_rate": 0.16406933154962713,
      "loss": 0.2689,
      "num_input_tokens_seen": 16288832,
      "step": 18805
    },
    {
      "epoch": 8.868458274398868,
      "grad_norm": 0.0007709947531111538,
      "learning_rate": 0.16401068528728133,
      "loss": 0.2171,
      "num_input_tokens_seen": 16293152,
      "step": 18810
    },
    {
      "epoch": 8.87081565299387,
      "grad_norm": 0.0015059354482218623,
      "learning_rate": 0.16395203686431173,
      "loss": 0.2079,
      "num_input_tokens_seen": 16297616,
      "step": 18815
    },
    {
      "epoch": 8.873173031588873,
      "grad_norm": 0.0006565727526322007,
      "learning_rate": 0.16389338628976277,
      "loss": 0.2117,
      "num_input_tokens_seen": 16302224,
      "step": 18820
    },
    {
      "epoch": 8.875530410183876,
      "grad_norm": 0.002393738366663456,
      "learning_rate": 0.163834733572679,
      "loss": 0.1912,
      "num_input_tokens_seen": 16307392,
      "step": 18825
    },
    {
      "epoch": 8.877887788778878,
      "grad_norm": 0.009231410920619965,
      "learning_rate": 0.16377607872210545,
      "loss": 0.2494,
      "num_input_tokens_seen": 16312528,
      "step": 18830
    },
    {
      "epoch": 8.88024516737388,
      "grad_norm": 0.001489865593612194,
      "learning_rate": 0.16371742174708748,
      "loss": 0.2103,
      "num_input_tokens_seen": 16316736,
      "step": 18835
    },
    {
      "epoch": 8.882602545968883,
      "grad_norm": 0.0008405112312175333,
      "learning_rate": 0.16365876265667065,
      "loss": 0.1659,
      "num_input_tokens_seen": 16321152,
      "step": 18840
    },
    {
      "epoch": 8.884959924563885,
      "grad_norm": 0.0011108461767435074,
      "learning_rate": 0.163600101459901,
      "loss": 0.1761,
      "num_input_tokens_seen": 16325344,
      "step": 18845
    },
    {
      "epoch": 8.887317303158888,
      "grad_norm": 0.0005577962729148567,
      "learning_rate": 0.16354143816582484,
      "loss": 0.1763,
      "num_input_tokens_seen": 16328928,
      "step": 18850
    },
    {
      "epoch": 8.88967468175389,
      "grad_norm": 0.0005054153152741492,
      "learning_rate": 0.1634827727834887,
      "loss": 0.3091,
      "num_input_tokens_seen": 16333024,
      "step": 18855
    },
    {
      "epoch": 8.892032060348892,
      "grad_norm": 0.0007885924424044788,
      "learning_rate": 0.16342410532193954,
      "loss": 0.2695,
      "num_input_tokens_seen": 16337136,
      "step": 18860
    },
    {
      "epoch": 8.894389438943895,
      "grad_norm": 0.0006701434031128883,
      "learning_rate": 0.16336543579022464,
      "loss": 0.216,
      "num_input_tokens_seen": 16341472,
      "step": 18865
    },
    {
      "epoch": 8.896746817538897,
      "grad_norm": 0.0011918238596990705,
      "learning_rate": 0.16330676419739157,
      "loss": 0.2275,
      "num_input_tokens_seen": 16346480,
      "step": 18870
    },
    {
      "epoch": 8.8991041961339,
      "grad_norm": 0.0005918386159464717,
      "learning_rate": 0.1632480905524883,
      "loss": 0.2124,
      "num_input_tokens_seen": 16350272,
      "step": 18875
    },
    {
      "epoch": 8.901461574728902,
      "grad_norm": 0.0007370043313130736,
      "learning_rate": 0.16318941486456293,
      "loss": 0.1946,
      "num_input_tokens_seen": 16354160,
      "step": 18880
    },
    {
      "epoch": 8.903818953323904,
      "grad_norm": 0.0010830272221937776,
      "learning_rate": 0.16313073714266405,
      "loss": 0.2151,
      "num_input_tokens_seen": 16359248,
      "step": 18885
    },
    {
      "epoch": 8.906176331918907,
      "grad_norm": 0.002053155330941081,
      "learning_rate": 0.16307205739584052,
      "loss": 0.2693,
      "num_input_tokens_seen": 16363664,
      "step": 18890
    },
    {
      "epoch": 8.90853371051391,
      "grad_norm": 0.001284484169445932,
      "learning_rate": 0.16301337563314144,
      "loss": 0.226,
      "num_input_tokens_seen": 16367632,
      "step": 18895
    },
    {
      "epoch": 8.910891089108912,
      "grad_norm": 0.0005017881048843265,
      "learning_rate": 0.1629546918636163,
      "loss": 0.159,
      "num_input_tokens_seen": 16371536,
      "step": 18900
    },
    {
      "epoch": 8.913248467703912,
      "grad_norm": 0.0009290958405472338,
      "learning_rate": 0.16289600609631485,
      "loss": 0.2383,
      "num_input_tokens_seen": 16375888,
      "step": 18905
    },
    {
      "epoch": 8.915605846298917,
      "grad_norm": 0.0008688514353707433,
      "learning_rate": 0.16283731834028722,
      "loss": 0.1865,
      "num_input_tokens_seen": 16380816,
      "step": 18910
    },
    {
      "epoch": 8.917963224893917,
      "grad_norm": 0.0005103500443510711,
      "learning_rate": 0.16277862860458378,
      "loss": 0.2238,
      "num_input_tokens_seen": 16384944,
      "step": 18915
    },
    {
      "epoch": 8.92032060348892,
      "grad_norm": 0.0010723761515691876,
      "learning_rate": 0.16271993689825526,
      "loss": 0.2085,
      "num_input_tokens_seen": 16389296,
      "step": 18920
    },
    {
      "epoch": 8.922677982083922,
      "grad_norm": 0.0013141264207661152,
      "learning_rate": 0.1626612432303526,
      "loss": 0.254,
      "num_input_tokens_seen": 16393440,
      "step": 18925
    },
    {
      "epoch": 8.925035360678924,
      "grad_norm": 0.0005712908459827304,
      "learning_rate": 0.1626025476099271,
      "loss": 0.2255,
      "num_input_tokens_seen": 16397872,
      "step": 18930
    },
    {
      "epoch": 8.927392739273927,
      "grad_norm": 0.0007861306075938046,
      "learning_rate": 0.1625438500460304,
      "loss": 0.2286,
      "num_input_tokens_seen": 16401968,
      "step": 18935
    },
    {
      "epoch": 8.92975011786893,
      "grad_norm": 0.001409479882568121,
      "learning_rate": 0.16248515054771442,
      "loss": 0.2219,
      "num_input_tokens_seen": 16406000,
      "step": 18940
    },
    {
      "epoch": 8.932107496463932,
      "grad_norm": 0.0013923911610618234,
      "learning_rate": 0.16242644912403123,
      "loss": 0.2193,
      "num_input_tokens_seen": 16410336,
      "step": 18945
    },
    {
      "epoch": 8.934464875058934,
      "grad_norm": 0.0005249543464742601,
      "learning_rate": 0.1623677457840335,
      "loss": 0.2195,
      "num_input_tokens_seen": 16414896,
      "step": 18950
    },
    {
      "epoch": 8.936822253653936,
      "grad_norm": 0.0006817495450377464,
      "learning_rate": 0.16230904053677397,
      "loss": 0.2182,
      "num_input_tokens_seen": 16419056,
      "step": 18955
    },
    {
      "epoch": 8.939179632248939,
      "grad_norm": 0.002415242837741971,
      "learning_rate": 0.16225033339130568,
      "loss": 0.2357,
      "num_input_tokens_seen": 16423088,
      "step": 18960
    },
    {
      "epoch": 8.941537010843941,
      "grad_norm": 0.0007602553232572973,
      "learning_rate": 0.16219162435668197,
      "loss": 0.213,
      "num_input_tokens_seen": 16427152,
      "step": 18965
    },
    {
      "epoch": 8.943894389438944,
      "grad_norm": 0.0003763705608434975,
      "learning_rate": 0.16213291344195666,
      "loss": 0.1726,
      "num_input_tokens_seen": 16430688,
      "step": 18970
    },
    {
      "epoch": 8.946251768033946,
      "grad_norm": 0.0018044696189463139,
      "learning_rate": 0.16207420065618358,
      "loss": 0.2407,
      "num_input_tokens_seen": 16435040,
      "step": 18975
    },
    {
      "epoch": 8.948609146628948,
      "grad_norm": 0.0014765002997592092,
      "learning_rate": 0.16201548600841706,
      "loss": 0.2262,
      "num_input_tokens_seen": 16438960,
      "step": 18980
    },
    {
      "epoch": 8.95096652522395,
      "grad_norm": 0.001295358524657786,
      "learning_rate": 0.16195676950771154,
      "loss": 0.1891,
      "num_input_tokens_seen": 16442656,
      "step": 18985
    },
    {
      "epoch": 8.953323903818953,
      "grad_norm": 0.0005954341613687575,
      "learning_rate": 0.16189805116312198,
      "loss": 0.2915,
      "num_input_tokens_seen": 16446864,
      "step": 18990
    },
    {
      "epoch": 8.955681282413956,
      "grad_norm": 0.0020496181678026915,
      "learning_rate": 0.16183933098370337,
      "loss": 0.2403,
      "num_input_tokens_seen": 16453104,
      "step": 18995
    },
    {
      "epoch": 8.958038661008958,
      "grad_norm": 0.0007852452108636498,
      "learning_rate": 0.16178060897851115,
      "loss": 0.2249,
      "num_input_tokens_seen": 16457552,
      "step": 19000
    },
    {
      "epoch": 8.958038661008958,
      "eval_loss": 0.22157442569732666,
      "eval_runtime": 21.8927,
      "eval_samples_per_second": 43.074,
      "eval_steps_per_second": 21.56,
      "num_input_tokens_seen": 16457552,
      "step": 19000
    },
    {
      "epoch": 8.96039603960396,
      "grad_norm": 0.0015846642199903727,
      "learning_rate": 0.16172188515660096,
      "loss": 0.2346,
      "num_input_tokens_seen": 16461392,
      "step": 19005
    },
    {
      "epoch": 8.962753418198963,
      "grad_norm": 0.0014008423313498497,
      "learning_rate": 0.16166315952702878,
      "loss": 0.1988,
      "num_input_tokens_seen": 16466320,
      "step": 19010
    },
    {
      "epoch": 8.965110796793965,
      "grad_norm": 0.0008225043420679867,
      "learning_rate": 0.16160443209885084,
      "loss": 0.1992,
      "num_input_tokens_seen": 16470816,
      "step": 19015
    },
    {
      "epoch": 8.967468175388968,
      "grad_norm": 0.0008845989359542727,
      "learning_rate": 0.16154570288112363,
      "loss": 0.2425,
      "num_input_tokens_seen": 16474976,
      "step": 19020
    },
    {
      "epoch": 8.96982555398397,
      "grad_norm": 0.0009841021383181214,
      "learning_rate": 0.16148697188290395,
      "loss": 0.2282,
      "num_input_tokens_seen": 16479216,
      "step": 19025
    },
    {
      "epoch": 8.972182932578972,
      "grad_norm": 0.0007237554527819157,
      "learning_rate": 0.16142823911324888,
      "loss": 0.2296,
      "num_input_tokens_seen": 16483536,
      "step": 19030
    },
    {
      "epoch": 8.974540311173975,
      "grad_norm": 0.0012564309872686863,
      "learning_rate": 0.16136950458121568,
      "loss": 0.2529,
      "num_input_tokens_seen": 16488304,
      "step": 19035
    },
    {
      "epoch": 8.976897689768977,
      "grad_norm": 0.001177856232970953,
      "learning_rate": 0.16131076829586205,
      "loss": 0.222,
      "num_input_tokens_seen": 16492688,
      "step": 19040
    },
    {
      "epoch": 8.97925506836398,
      "grad_norm": 0.0008911918848752975,
      "learning_rate": 0.1612520302662457,
      "loss": 0.2295,
      "num_input_tokens_seen": 16496576,
      "step": 19045
    },
    {
      "epoch": 8.981612446958982,
      "grad_norm": 0.0008841329836286604,
      "learning_rate": 0.16119329050142497,
      "loss": 0.2411,
      "num_input_tokens_seen": 16501120,
      "step": 19050
    },
    {
      "epoch": 8.983969825553984,
      "grad_norm": 0.001054307329468429,
      "learning_rate": 0.16113454901045818,
      "loss": 0.2322,
      "num_input_tokens_seen": 16505520,
      "step": 19055
    },
    {
      "epoch": 8.986327204148987,
      "grad_norm": 0.0005575273535214365,
      "learning_rate": 0.16107580580240397,
      "loss": 0.2309,
      "num_input_tokens_seen": 16510032,
      "step": 19060
    },
    {
      "epoch": 8.98868458274399,
      "grad_norm": 0.0005366908735595644,
      "learning_rate": 0.16101706088632134,
      "loss": 0.2223,
      "num_input_tokens_seen": 16514416,
      "step": 19065
    },
    {
      "epoch": 8.991041961338992,
      "grad_norm": 0.0003956820582970977,
      "learning_rate": 0.16095831427126947,
      "loss": 0.2232,
      "num_input_tokens_seen": 16518256,
      "step": 19070
    },
    {
      "epoch": 8.993399339933994,
      "grad_norm": 0.00056469562696293,
      "learning_rate": 0.16089956596630783,
      "loss": 0.2416,
      "num_input_tokens_seen": 16522784,
      "step": 19075
    },
    {
      "epoch": 8.995756718528996,
      "grad_norm": 0.0011187553172931075,
      "learning_rate": 0.16084081598049618,
      "loss": 0.2248,
      "num_input_tokens_seen": 16527056,
      "step": 19080
    },
    {
      "epoch": 8.998114097123999,
      "grad_norm": 0.0004213513748254627,
      "learning_rate": 0.1607820643228944,
      "loss": 0.2224,
      "num_input_tokens_seen": 16531328,
      "step": 19085
    },
    {
      "epoch": 9.000471475719001,
      "grad_norm": 0.0010986063862219453,
      "learning_rate": 0.16072331100256285,
      "loss": 0.2314,
      "num_input_tokens_seen": 16535808,
      "step": 19090
    },
    {
      "epoch": 9.002828854314004,
      "grad_norm": 0.0004284385940991342,
      "learning_rate": 0.16066455602856197,
      "loss": 0.2117,
      "num_input_tokens_seen": 16541664,
      "step": 19095
    },
    {
      "epoch": 9.005186232909006,
      "grad_norm": 0.000663784914650023,
      "learning_rate": 0.16060579940995257,
      "loss": 0.2371,
      "num_input_tokens_seen": 16545808,
      "step": 19100
    },
    {
      "epoch": 9.007543611504008,
      "grad_norm": 0.0006084454944357276,
      "learning_rate": 0.16054704115579557,
      "loss": 0.2126,
      "num_input_tokens_seen": 16550400,
      "step": 19105
    },
    {
      "epoch": 9.009900990099009,
      "grad_norm": 0.0014110076008364558,
      "learning_rate": 0.1604882812751523,
      "loss": 0.2234,
      "num_input_tokens_seen": 16554240,
      "step": 19110
    },
    {
      "epoch": 9.012258368694011,
      "grad_norm": 0.0005746796377934515,
      "learning_rate": 0.16042951977708425,
      "loss": 0.2196,
      "num_input_tokens_seen": 16558672,
      "step": 19115
    },
    {
      "epoch": 9.014615747289014,
      "grad_norm": 0.0006133854622021317,
      "learning_rate": 0.16037075667065318,
      "loss": 0.2038,
      "num_input_tokens_seen": 16563168,
      "step": 19120
    },
    {
      "epoch": 9.016973125884016,
      "grad_norm": 0.0007921397918835282,
      "learning_rate": 0.1603119919649211,
      "loss": 0.207,
      "num_input_tokens_seen": 16567152,
      "step": 19125
    },
    {
      "epoch": 9.019330504479019,
      "grad_norm": 0.0004568667500279844,
      "learning_rate": 0.16025322566895028,
      "loss": 0.2207,
      "num_input_tokens_seen": 16571408,
      "step": 19130
    },
    {
      "epoch": 9.021687883074021,
      "grad_norm": 0.0005554032977670431,
      "learning_rate": 0.16019445779180322,
      "loss": 0.2327,
      "num_input_tokens_seen": 16575664,
      "step": 19135
    },
    {
      "epoch": 9.024045261669023,
      "grad_norm": 0.001470726216211915,
      "learning_rate": 0.16013568834254271,
      "loss": 0.211,
      "num_input_tokens_seen": 16579744,
      "step": 19140
    },
    {
      "epoch": 9.026402640264026,
      "grad_norm": 0.0005826438427902758,
      "learning_rate": 0.1600769173302316,
      "loss": 0.2262,
      "num_input_tokens_seen": 16583936,
      "step": 19145
    },
    {
      "epoch": 9.028760018859028,
      "grad_norm": 0.0007293298840522766,
      "learning_rate": 0.16001814476393322,
      "loss": 0.2165,
      "num_input_tokens_seen": 16588272,
      "step": 19150
    },
    {
      "epoch": 9.03111739745403,
      "grad_norm": 0.0006621292559430003,
      "learning_rate": 0.15995937065271104,
      "loss": 0.2155,
      "num_input_tokens_seen": 16591712,
      "step": 19155
    },
    {
      "epoch": 9.033474776049033,
      "grad_norm": 0.0016877889866009355,
      "learning_rate": 0.15990059500562873,
      "loss": 0.2061,
      "num_input_tokens_seen": 16595824,
      "step": 19160
    },
    {
      "epoch": 9.035832154644035,
      "grad_norm": 0.0007410135585814714,
      "learning_rate": 0.15984181783175025,
      "loss": 0.2428,
      "num_input_tokens_seen": 16600304,
      "step": 19165
    },
    {
      "epoch": 9.038189533239038,
      "grad_norm": 0.001243826001882553,
      "learning_rate": 0.1597830391401398,
      "loss": 0.2182,
      "num_input_tokens_seen": 16604912,
      "step": 19170
    },
    {
      "epoch": 9.04054691183404,
      "grad_norm": 0.0013863013591617346,
      "learning_rate": 0.15972425893986178,
      "loss": 0.2332,
      "num_input_tokens_seen": 16609376,
      "step": 19175
    },
    {
      "epoch": 9.042904290429043,
      "grad_norm": 0.0007335345726460218,
      "learning_rate": 0.15966547723998084,
      "loss": 0.2405,
      "num_input_tokens_seen": 16613296,
      "step": 19180
    },
    {
      "epoch": 9.045261669024045,
      "grad_norm": 0.001419559121131897,
      "learning_rate": 0.15960669404956176,
      "loss": 0.2328,
      "num_input_tokens_seen": 16617824,
      "step": 19185
    },
    {
      "epoch": 9.047619047619047,
      "grad_norm": 0.0015403294237330556,
      "learning_rate": 0.1595479093776698,
      "loss": 0.1817,
      "num_input_tokens_seen": 16622768,
      "step": 19190
    },
    {
      "epoch": 9.04997642621405,
      "grad_norm": 0.001643997966311872,
      "learning_rate": 0.15948912323337022,
      "loss": 0.2317,
      "num_input_tokens_seen": 16628464,
      "step": 19195
    },
    {
      "epoch": 9.052333804809052,
      "grad_norm": 0.0017574172234162688,
      "learning_rate": 0.1594303356257286,
      "loss": 0.2489,
      "num_input_tokens_seen": 16632272,
      "step": 19200
    },
    {
      "epoch": 9.052333804809052,
      "eval_loss": 0.2172006219625473,
      "eval_runtime": 21.8936,
      "eval_samples_per_second": 43.072,
      "eval_steps_per_second": 21.559,
      "num_input_tokens_seen": 16632272,
      "step": 19200
    },
    {
      "epoch": 9.054691183404055,
      "grad_norm": 0.0013368241488933563,
      "learning_rate": 0.15937154656381072,
      "loss": 0.2237,
      "num_input_tokens_seen": 16635680,
      "step": 19205
    },
    {
      "epoch": 9.057048561999057,
      "grad_norm": 0.0006881268927827477,
      "learning_rate": 0.15931275605668258,
      "loss": 0.1929,
      "num_input_tokens_seen": 16640208,
      "step": 19210
    },
    {
      "epoch": 9.05940594059406,
      "grad_norm": 0.0016955672763288021,
      "learning_rate": 0.1592539641134104,
      "loss": 0.2559,
      "num_input_tokens_seen": 16644784,
      "step": 19215
    },
    {
      "epoch": 9.061763319189062,
      "grad_norm": 0.0010982337407767773,
      "learning_rate": 0.1591951707430607,
      "loss": 0.239,
      "num_input_tokens_seen": 16648912,
      "step": 19220
    },
    {
      "epoch": 9.064120697784064,
      "grad_norm": 0.001591330743394792,
      "learning_rate": 0.15913637595470007,
      "loss": 0.2055,
      "num_input_tokens_seen": 16653216,
      "step": 19225
    },
    {
      "epoch": 9.066478076379067,
      "grad_norm": 0.0021961117163300514,
      "learning_rate": 0.15907757975739548,
      "loss": 0.2158,
      "num_input_tokens_seen": 16657296,
      "step": 19230
    },
    {
      "epoch": 9.068835454974069,
      "grad_norm": 0.000696578121278435,
      "learning_rate": 0.159018782160214,
      "loss": 0.2125,
      "num_input_tokens_seen": 16661856,
      "step": 19235
    },
    {
      "epoch": 9.071192833569071,
      "grad_norm": 0.0006848767516203225,
      "learning_rate": 0.158959983172223,
      "loss": 0.2019,
      "num_input_tokens_seen": 16665792,
      "step": 19240
    },
    {
      "epoch": 9.073550212164074,
      "grad_norm": 0.0007311970693990588,
      "learning_rate": 0.15890118280249,
      "loss": 0.2066,
      "num_input_tokens_seen": 16670336,
      "step": 19245
    },
    {
      "epoch": 9.075907590759076,
      "grad_norm": 0.0010550773004069924,
      "learning_rate": 0.15884238106008275,
      "loss": 0.2288,
      "num_input_tokens_seen": 16674832,
      "step": 19250
    },
    {
      "epoch": 9.078264969354079,
      "grad_norm": 0.0008420665981248021,
      "learning_rate": 0.15878357795406922,
      "loss": 0.2487,
      "num_input_tokens_seen": 16679696,
      "step": 19255
    },
    {
      "epoch": 9.080622347949081,
      "grad_norm": 0.0013266054447740316,
      "learning_rate": 0.15872477349351757,
      "loss": 0.22,
      "num_input_tokens_seen": 16683344,
      "step": 19260
    },
    {
      "epoch": 9.082979726544083,
      "grad_norm": 0.0015388745814561844,
      "learning_rate": 0.15866596768749622,
      "loss": 0.2216,
      "num_input_tokens_seen": 16687920,
      "step": 19265
    },
    {
      "epoch": 9.085337105139086,
      "grad_norm": 0.0007284139865078032,
      "learning_rate": 0.15860716054507373,
      "loss": 0.2519,
      "num_input_tokens_seen": 16692288,
      "step": 19270
    },
    {
      "epoch": 9.087694483734088,
      "grad_norm": 0.0011999985435977578,
      "learning_rate": 0.1585483520753189,
      "loss": 0.2283,
      "num_input_tokens_seen": 16696288,
      "step": 19275
    },
    {
      "epoch": 9.09005186232909,
      "grad_norm": 0.0006251749582588673,
      "learning_rate": 0.1584895422873008,
      "loss": 0.2246,
      "num_input_tokens_seen": 16700800,
      "step": 19280
    },
    {
      "epoch": 9.092409240924093,
      "grad_norm": 0.0006919074221514165,
      "learning_rate": 0.1584307311900886,
      "loss": 0.215,
      "num_input_tokens_seen": 16705424,
      "step": 19285
    },
    {
      "epoch": 9.094766619519095,
      "grad_norm": 0.0013530576834455132,
      "learning_rate": 0.1583719187927517,
      "loss": 0.2067,
      "num_input_tokens_seen": 16709648,
      "step": 19290
    },
    {
      "epoch": 9.097123998114098,
      "grad_norm": 0.0005306766252033412,
      "learning_rate": 0.15831310510435967,
      "loss": 0.1886,
      "num_input_tokens_seen": 16713552,
      "step": 19295
    },
    {
      "epoch": 9.0994813767091,
      "grad_norm": 0.000619783706497401,
      "learning_rate": 0.15825429013398243,
      "loss": 0.2857,
      "num_input_tokens_seen": 16717808,
      "step": 19300
    },
    {
      "epoch": 9.101838755304103,
      "grad_norm": 0.0005635828129015863,
      "learning_rate": 0.15819547389068986,
      "loss": 0.2173,
      "num_input_tokens_seen": 16721680,
      "step": 19305
    },
    {
      "epoch": 9.104196133899103,
      "grad_norm": 0.0006941886967979372,
      "learning_rate": 0.1581366563835522,
      "loss": 0.1976,
      "num_input_tokens_seen": 16725728,
      "step": 19310
    },
    {
      "epoch": 9.106553512494106,
      "grad_norm": 0.0006347576272673905,
      "learning_rate": 0.15807783762163993,
      "loss": 0.2243,
      "num_input_tokens_seen": 16730224,
      "step": 19315
    },
    {
      "epoch": 9.108910891089108,
      "grad_norm": 0.0012645877432078123,
      "learning_rate": 0.15801901761402365,
      "loss": 0.1793,
      "num_input_tokens_seen": 16734128,
      "step": 19320
    },
    {
      "epoch": 9.11126826968411,
      "grad_norm": 0.0011194159742444754,
      "learning_rate": 0.157960196369774,
      "loss": 0.1707,
      "num_input_tokens_seen": 16739024,
      "step": 19325
    },
    {
      "epoch": 9.113625648279113,
      "grad_norm": 0.0008897940861061215,
      "learning_rate": 0.157901373897962,
      "loss": 0.2312,
      "num_input_tokens_seen": 16742800,
      "step": 19330
    },
    {
      "epoch": 9.115983026874115,
      "grad_norm": 0.0010471489513292909,
      "learning_rate": 0.15784255020765892,
      "loss": 0.2578,
      "num_input_tokens_seen": 16748096,
      "step": 19335
    },
    {
      "epoch": 9.118340405469118,
      "grad_norm": 0.001466831425204873,
      "learning_rate": 0.157783725307936,
      "loss": 0.2035,
      "num_input_tokens_seen": 16752224,
      "step": 19340
    },
    {
      "epoch": 9.12069778406412,
      "grad_norm": 0.000609258480835706,
      "learning_rate": 0.15772489920786484,
      "loss": 0.2365,
      "num_input_tokens_seen": 16757056,
      "step": 19345
    },
    {
      "epoch": 9.123055162659123,
      "grad_norm": 0.0004949611029587686,
      "learning_rate": 0.15766607191651713,
      "loss": 0.225,
      "num_input_tokens_seen": 16760768,
      "step": 19350
    },
    {
      "epoch": 9.125412541254125,
      "grad_norm": 0.0013947695260867476,
      "learning_rate": 0.1576072434429648,
      "loss": 0.2067,
      "num_input_tokens_seen": 16765152,
      "step": 19355
    },
    {
      "epoch": 9.127769919849127,
      "grad_norm": 0.0006468924693763256,
      "learning_rate": 0.15754841379627998,
      "loss": 0.2333,
      "num_input_tokens_seen": 16769664,
      "step": 19360
    },
    {
      "epoch": 9.13012729844413,
      "grad_norm": 0.0010903155198320746,
      "learning_rate": 0.15748958298553484,
      "loss": 0.1976,
      "num_input_tokens_seen": 16774624,
      "step": 19365
    },
    {
      "epoch": 9.132484677039132,
      "grad_norm": 0.0005933353095315397,
      "learning_rate": 0.1574307510198019,
      "loss": 0.1987,
      "num_input_tokens_seen": 16778752,
      "step": 19370
    },
    {
      "epoch": 9.134842055634135,
      "grad_norm": 0.000687772873789072,
      "learning_rate": 0.15737191790815375,
      "loss": 0.2285,
      "num_input_tokens_seen": 16783648,
      "step": 19375
    },
    {
      "epoch": 9.137199434229137,
      "grad_norm": 0.0008050749311223626,
      "learning_rate": 0.15731308365966323,
      "loss": 0.2418,
      "num_input_tokens_seen": 16788640,
      "step": 19380
    },
    {
      "epoch": 9.13955681282414,
      "grad_norm": 0.0009536448051221669,
      "learning_rate": 0.15725424828340331,
      "loss": 0.2098,
      "num_input_tokens_seen": 16792560,
      "step": 19385
    },
    {
      "epoch": 9.141914191419142,
      "grad_norm": 0.0007546758279204369,
      "learning_rate": 0.15719541178844715,
      "loss": 0.2184,
      "num_input_tokens_seen": 16796624,
      "step": 19390
    },
    {
      "epoch": 9.144271570014144,
      "grad_norm": 0.000612596224527806,
      "learning_rate": 0.15713657418386806,
      "loss": 0.226,
      "num_input_tokens_seen": 16801712,
      "step": 19395
    },
    {
      "epoch": 9.146628948609147,
      "grad_norm": 0.0008675246499478817,
      "learning_rate": 0.15707773547873957,
      "loss": 0.2361,
      "num_input_tokens_seen": 16806304,
      "step": 19400
    },
    {
      "epoch": 9.146628948609147,
      "eval_loss": 0.21671287715435028,
      "eval_runtime": 21.9288,
      "eval_samples_per_second": 43.003,
      "eval_steps_per_second": 21.524,
      "num_input_tokens_seen": 16806304,
      "step": 19400
    },
    {
      "epoch": 9.148986327204149,
      "grad_norm": 0.0007517562480643392,
      "learning_rate": 0.1570188956821353,
      "loss": 0.2182,
      "num_input_tokens_seen": 16811440,
      "step": 19405
    },
    {
      "epoch": 9.151343705799151,
      "grad_norm": 0.0006993401912041008,
      "learning_rate": 0.1569600548031291,
      "loss": 0.2049,
      "num_input_tokens_seen": 16814816,
      "step": 19410
    },
    {
      "epoch": 9.153701084394154,
      "grad_norm": 0.0006031584343872964,
      "learning_rate": 0.156901212850795,
      "loss": 0.2054,
      "num_input_tokens_seen": 16818672,
      "step": 19415
    },
    {
      "epoch": 9.156058462989156,
      "grad_norm": 0.001139874686487019,
      "learning_rate": 0.15684236983420716,
      "loss": 0.2253,
      "num_input_tokens_seen": 16822880,
      "step": 19420
    },
    {
      "epoch": 9.158415841584159,
      "grad_norm": 0.0005851318128407001,
      "learning_rate": 0.1567835257624399,
      "loss": 0.2293,
      "num_input_tokens_seen": 16826544,
      "step": 19425
    },
    {
      "epoch": 9.160773220179161,
      "grad_norm": 0.0012328233569860458,
      "learning_rate": 0.1567246806445677,
      "loss": 0.2253,
      "num_input_tokens_seen": 16831024,
      "step": 19430
    },
    {
      "epoch": 9.163130598774163,
      "grad_norm": 0.0007991245947778225,
      "learning_rate": 0.15666583448966526,
      "loss": 0.1871,
      "num_input_tokens_seen": 16835632,
      "step": 19435
    },
    {
      "epoch": 9.165487977369166,
      "grad_norm": 0.0005156891420483589,
      "learning_rate": 0.1566069873068074,
      "loss": 0.2206,
      "num_input_tokens_seen": 16839744,
      "step": 19440
    },
    {
      "epoch": 9.167845355964168,
      "grad_norm": 0.0014800290809944272,
      "learning_rate": 0.156548139105069,
      "loss": 0.2401,
      "num_input_tokens_seen": 16844256,
      "step": 19445
    },
    {
      "epoch": 9.17020273455917,
      "grad_norm": 0.0005292773712426424,
      "learning_rate": 0.15648928989352529,
      "loss": 0.1986,
      "num_input_tokens_seen": 16848400,
      "step": 19450
    },
    {
      "epoch": 9.172560113154173,
      "grad_norm": 0.0005309787811711431,
      "learning_rate": 0.15643043968125156,
      "loss": 0.2066,
      "num_input_tokens_seen": 16852064,
      "step": 19455
    },
    {
      "epoch": 9.174917491749175,
      "grad_norm": 0.0007944013923406601,
      "learning_rate": 0.15637158847732316,
      "loss": 0.2195,
      "num_input_tokens_seen": 16856528,
      "step": 19460
    },
    {
      "epoch": 9.177274870344178,
      "grad_norm": 0.0006142548518255353,
      "learning_rate": 0.15631273629081582,
      "loss": 0.2073,
      "num_input_tokens_seen": 16860736,
      "step": 19465
    },
    {
      "epoch": 9.17963224893918,
      "grad_norm": 0.0006446227780543268,
      "learning_rate": 0.15625388313080518,
      "loss": 0.2295,
      "num_input_tokens_seen": 16864704,
      "step": 19470
    },
    {
      "epoch": 9.181989627534183,
      "grad_norm": 0.001815654686652124,
      "learning_rate": 0.15619502900636714,
      "loss": 0.2303,
      "num_input_tokens_seen": 16869536,
      "step": 19475
    },
    {
      "epoch": 9.184347006129185,
      "grad_norm": 0.0005904544377699494,
      "learning_rate": 0.15613617392657783,
      "loss": 0.2348,
      "num_input_tokens_seen": 16874416,
      "step": 19480
    },
    {
      "epoch": 9.186704384724187,
      "grad_norm": 0.0010089308489114046,
      "learning_rate": 0.15607731790051335,
      "loss": 0.2227,
      "num_input_tokens_seen": 16878800,
      "step": 19485
    },
    {
      "epoch": 9.18906176331919,
      "grad_norm": 0.0020343728829175234,
      "learning_rate": 0.15601846093725008,
      "loss": 0.2327,
      "num_input_tokens_seen": 16883856,
      "step": 19490
    },
    {
      "epoch": 9.191419141914192,
      "grad_norm": 0.0007493772427551448,
      "learning_rate": 0.1559596030458645,
      "loss": 0.1971,
      "num_input_tokens_seen": 16887808,
      "step": 19495
    },
    {
      "epoch": 9.193776520509195,
      "grad_norm": 0.001080323476344347,
      "learning_rate": 0.1559007442354333,
      "loss": 0.1862,
      "num_input_tokens_seen": 16891536,
      "step": 19500
    },
    {
      "epoch": 9.196133899104197,
      "grad_norm": 0.0008865422569215298,
      "learning_rate": 0.15584188451503314,
      "loss": 0.2151,
      "num_input_tokens_seen": 16896192,
      "step": 19505
    },
    {
      "epoch": 9.198491277699198,
      "grad_norm": 0.0008593936800025403,
      "learning_rate": 0.15578302389374094,
      "loss": 0.2016,
      "num_input_tokens_seen": 16900880,
      "step": 19510
    },
    {
      "epoch": 9.2008486562942,
      "grad_norm": 0.0010523648234084249,
      "learning_rate": 0.1557241623806338,
      "loss": 0.1814,
      "num_input_tokens_seen": 16904832,
      "step": 19515
    },
    {
      "epoch": 9.203206034889202,
      "grad_norm": 0.0010727095650509,
      "learning_rate": 0.15566529998478887,
      "loss": 0.2307,
      "num_input_tokens_seen": 16909552,
      "step": 19520
    },
    {
      "epoch": 9.205563413484205,
      "grad_norm": 0.00045209444942884147,
      "learning_rate": 0.15560643671528354,
      "loss": 0.2243,
      "num_input_tokens_seen": 16913872,
      "step": 19525
    },
    {
      "epoch": 9.207920792079207,
      "grad_norm": 0.000814487284515053,
      "learning_rate": 0.15554757258119514,
      "loss": 0.202,
      "num_input_tokens_seen": 16918176,
      "step": 19530
    },
    {
      "epoch": 9.21027817067421,
      "grad_norm": 0.0005932566709816456,
      "learning_rate": 0.1554887075916014,
      "loss": 0.2179,
      "num_input_tokens_seen": 16922144,
      "step": 19535
    },
    {
      "epoch": 9.212635549269212,
      "grad_norm": 0.0006946565699763596,
      "learning_rate": 0.15542984175558,
      "loss": 0.2359,
      "num_input_tokens_seen": 16926752,
      "step": 19540
    },
    {
      "epoch": 9.214992927864214,
      "grad_norm": 0.0008088828180916607,
      "learning_rate": 0.1553709750822087,
      "loss": 0.2183,
      "num_input_tokens_seen": 16931168,
      "step": 19545
    },
    {
      "epoch": 9.217350306459217,
      "grad_norm": 0.0009399834671057761,
      "learning_rate": 0.15531210758056554,
      "loss": 0.2314,
      "num_input_tokens_seen": 16935440,
      "step": 19550
    },
    {
      "epoch": 9.21970768505422,
      "grad_norm": 0.0013915252638980746,
      "learning_rate": 0.15525323925972867,
      "loss": 0.2504,
      "num_input_tokens_seen": 16940512,
      "step": 19555
    },
    {
      "epoch": 9.222065063649222,
      "grad_norm": 0.0008073599310591817,
      "learning_rate": 0.15519437012877627,
      "loss": 0.2318,
      "num_input_tokens_seen": 16944688,
      "step": 19560
    },
    {
      "epoch": 9.224422442244224,
      "grad_norm": 0.001296561909839511,
      "learning_rate": 0.15513550019678676,
      "loss": 0.2278,
      "num_input_tokens_seen": 16948416,
      "step": 19565
    },
    {
      "epoch": 9.226779820839226,
      "grad_norm": 0.0017550145275890827,
      "learning_rate": 0.15507662947283854,
      "loss": 0.2206,
      "num_input_tokens_seen": 16952800,
      "step": 19570
    },
    {
      "epoch": 9.229137199434229,
      "grad_norm": 0.0007011658162809908,
      "learning_rate": 0.15501775796601028,
      "loss": 0.2,
      "num_input_tokens_seen": 16956032,
      "step": 19575
    },
    {
      "epoch": 9.231494578029231,
      "grad_norm": 0.0007046512328088284,
      "learning_rate": 0.15495888568538066,
      "loss": 0.1873,
      "num_input_tokens_seen": 16961152,
      "step": 19580
    },
    {
      "epoch": 9.233851956624234,
      "grad_norm": 0.0005122357397340238,
      "learning_rate": 0.1549000126400286,
      "loss": 0.1993,
      "num_input_tokens_seen": 16965792,
      "step": 19585
    },
    {
      "epoch": 9.236209335219236,
      "grad_norm": 0.0021124614868313074,
      "learning_rate": 0.15484113883903294,
      "loss": 0.2007,
      "num_input_tokens_seen": 16971008,
      "step": 19590
    },
    {
      "epoch": 9.238566713814238,
      "grad_norm": 0.0018380408873781562,
      "learning_rate": 0.15478226429147288,
      "loss": 0.2849,
      "num_input_tokens_seen": 16974304,
      "step": 19595
    },
    {
      "epoch": 9.24092409240924,
      "grad_norm": 0.0014694529818370938,
      "learning_rate": 0.15472338900642757,
      "loss": 0.2656,
      "num_input_tokens_seen": 16979072,
      "step": 19600
    },
    {
      "epoch": 9.24092409240924,
      "eval_loss": 0.21500974893569946,
      "eval_runtime": 21.9022,
      "eval_samples_per_second": 43.055,
      "eval_steps_per_second": 21.55,
      "num_input_tokens_seen": 16979072,
      "step": 19600
    },
    {
      "epoch": 9.243281471004243,
      "grad_norm": 0.0014381809160113335,
      "learning_rate": 0.15466451299297632,
      "loss": 0.2309,
      "num_input_tokens_seen": 16983264,
      "step": 19605
    },
    {
      "epoch": 9.245638849599246,
      "grad_norm": 0.0008513167849741876,
      "learning_rate": 0.15460563626019852,
      "loss": 0.2223,
      "num_input_tokens_seen": 16987328,
      "step": 19610
    },
    {
      "epoch": 9.247996228194248,
      "grad_norm": 0.0010599633678793907,
      "learning_rate": 0.15454675881717375,
      "loss": 0.2346,
      "num_input_tokens_seen": 16991472,
      "step": 19615
    },
    {
      "epoch": 9.25035360678925,
      "grad_norm": 0.0006905505433678627,
      "learning_rate": 0.1544878806729816,
      "loss": 0.2295,
      "num_input_tokens_seen": 16995536,
      "step": 19620
    },
    {
      "epoch": 9.252710985384253,
      "grad_norm": 0.0009247318957932293,
      "learning_rate": 0.1544290018367019,
      "loss": 0.2254,
      "num_input_tokens_seen": 16999216,
      "step": 19625
    },
    {
      "epoch": 9.255068363979255,
      "grad_norm": 0.0006859771674498916,
      "learning_rate": 0.15437012231741445,
      "loss": 0.2309,
      "num_input_tokens_seen": 17003168,
      "step": 19630
    },
    {
      "epoch": 9.257425742574258,
      "grad_norm": 0.0008022209513001144,
      "learning_rate": 0.1543112421241992,
      "loss": 0.2115,
      "num_input_tokens_seen": 17007056,
      "step": 19635
    },
    {
      "epoch": 9.25978312116926,
      "grad_norm": 0.001319844275712967,
      "learning_rate": 0.15425236126613626,
      "loss": 0.2342,
      "num_input_tokens_seen": 17011008,
      "step": 19640
    },
    {
      "epoch": 9.262140499764262,
      "grad_norm": 0.0006238723872229457,
      "learning_rate": 0.15419347975230577,
      "loss": 0.2237,
      "num_input_tokens_seen": 17014976,
      "step": 19645
    },
    {
      "epoch": 9.264497878359265,
      "grad_norm": 0.0007251785136759281,
      "learning_rate": 0.154134597591788,
      "loss": 0.2206,
      "num_input_tokens_seen": 17019920,
      "step": 19650
    },
    {
      "epoch": 9.266855256954267,
      "grad_norm": 0.000666293315589428,
      "learning_rate": 0.1540757147936633,
      "loss": 0.2171,
      "num_input_tokens_seen": 17024112,
      "step": 19655
    },
    {
      "epoch": 9.26921263554927,
      "grad_norm": 0.0023159752599895,
      "learning_rate": 0.1540168313670122,
      "loss": 0.232,
      "num_input_tokens_seen": 17028208,
      "step": 19660
    },
    {
      "epoch": 9.271570014144272,
      "grad_norm": 0.0013567954301834106,
      "learning_rate": 0.1539579473209152,
      "loss": 0.2176,
      "num_input_tokens_seen": 17032288,
      "step": 19665
    },
    {
      "epoch": 9.273927392739274,
      "grad_norm": 0.0009036398841999471,
      "learning_rate": 0.15389906266445294,
      "loss": 0.225,
      "num_input_tokens_seen": 17036864,
      "step": 19670
    },
    {
      "epoch": 9.276284771334277,
      "grad_norm": 0.0015848031034693122,
      "learning_rate": 0.15384017740670627,
      "loss": 0.1994,
      "num_input_tokens_seen": 17040512,
      "step": 19675
    },
    {
      "epoch": 9.27864214992928,
      "grad_norm": 0.0007027833489701152,
      "learning_rate": 0.15378129155675602,
      "loss": 0.2263,
      "num_input_tokens_seen": 17044544,
      "step": 19680
    },
    {
      "epoch": 9.280999528524282,
      "grad_norm": 0.0009053319226950407,
      "learning_rate": 0.15372240512368307,
      "loss": 0.1584,
      "num_input_tokens_seen": 17048800,
      "step": 19685
    },
    {
      "epoch": 9.283356907119284,
      "grad_norm": 0.0019408568041399121,
      "learning_rate": 0.1536635181165684,
      "loss": 0.1935,
      "num_input_tokens_seen": 17053232,
      "step": 19690
    },
    {
      "epoch": 9.285714285714286,
      "grad_norm": 0.0012799431569874287,
      "learning_rate": 0.15360463054449328,
      "loss": 0.1992,
      "num_input_tokens_seen": 17057328,
      "step": 19695
    },
    {
      "epoch": 9.288071664309289,
      "grad_norm": 0.0005950471386313438,
      "learning_rate": 0.1535457424165388,
      "loss": 0.2248,
      "num_input_tokens_seen": 17062208,
      "step": 19700
    },
    {
      "epoch": 9.290429042904291,
      "grad_norm": 0.0013661581324413419,
      "learning_rate": 0.15348685374178628,
      "loss": 0.2365,
      "num_input_tokens_seen": 17067200,
      "step": 19705
    },
    {
      "epoch": 9.292786421499294,
      "grad_norm": 0.0007819689344614744,
      "learning_rate": 0.1534279645293171,
      "loss": 0.2245,
      "num_input_tokens_seen": 17071200,
      "step": 19710
    },
    {
      "epoch": 9.295143800094294,
      "grad_norm": 0.0009435624233447015,
      "learning_rate": 0.1533690747882127,
      "loss": 0.2283,
      "num_input_tokens_seen": 17075408,
      "step": 19715
    },
    {
      "epoch": 9.297501178689297,
      "grad_norm": 0.0007729819626547396,
      "learning_rate": 0.15331018452755465,
      "loss": 0.2309,
      "num_input_tokens_seen": 17079328,
      "step": 19720
    },
    {
      "epoch": 9.299858557284299,
      "grad_norm": 0.0006764448480680585,
      "learning_rate": 0.15325129375642457,
      "loss": 0.2177,
      "num_input_tokens_seen": 17083808,
      "step": 19725
    },
    {
      "epoch": 9.302215935879302,
      "grad_norm": 0.004101473838090897,
      "learning_rate": 0.15319240248390406,
      "loss": 0.254,
      "num_input_tokens_seen": 17088944,
      "step": 19730
    },
    {
      "epoch": 9.304573314474304,
      "grad_norm": 0.002013708231970668,
      "learning_rate": 0.153133510719075,
      "loss": 0.2328,
      "num_input_tokens_seen": 17093200,
      "step": 19735
    },
    {
      "epoch": 9.306930693069306,
      "grad_norm": 0.00077823520405218,
      "learning_rate": 0.15307461847101922,
      "loss": 0.2117,
      "num_input_tokens_seen": 17097712,
      "step": 19740
    },
    {
      "epoch": 9.309288071664309,
      "grad_norm": 0.0007947483099997044,
      "learning_rate": 0.15301572574881864,
      "loss": 0.2377,
      "num_input_tokens_seen": 17102736,
      "step": 19745
    },
    {
      "epoch": 9.311645450259311,
      "grad_norm": 0.0006836146931163967,
      "learning_rate": 0.15295683256155523,
      "loss": 0.234,
      "num_input_tokens_seen": 17106240,
      "step": 19750
    },
    {
      "epoch": 9.314002828854314,
      "grad_norm": 0.0017669089138507843,
      "learning_rate": 0.15289793891831113,
      "loss": 0.2445,
      "num_input_tokens_seen": 17110816,
      "step": 19755
    },
    {
      "epoch": 9.316360207449316,
      "grad_norm": 0.0008257796871475875,
      "learning_rate": 0.15283904482816837,
      "loss": 0.2449,
      "num_input_tokens_seen": 17114848,
      "step": 19760
    },
    {
      "epoch": 9.318717586044318,
      "grad_norm": 0.0006342238048091531,
      "learning_rate": 0.15278015030020928,
      "loss": 0.2192,
      "num_input_tokens_seen": 17119488,
      "step": 19765
    },
    {
      "epoch": 9.32107496463932,
      "grad_norm": 0.000719713163562119,
      "learning_rate": 0.152721255343516,
      "loss": 0.2454,
      "num_input_tokens_seen": 17123888,
      "step": 19770
    },
    {
      "epoch": 9.323432343234323,
      "grad_norm": 0.001504553947597742,
      "learning_rate": 0.15266235996717098,
      "loss": 0.2145,
      "num_input_tokens_seen": 17128144,
      "step": 19775
    },
    {
      "epoch": 9.325789721829326,
      "grad_norm": 0.0007254997617565095,
      "learning_rate": 0.15260346418025664,
      "loss": 0.2266,
      "num_input_tokens_seen": 17131776,
      "step": 19780
    },
    {
      "epoch": 9.328147100424328,
      "grad_norm": 0.0007850494002923369,
      "learning_rate": 0.15254456799185537,
      "loss": 0.2145,
      "num_input_tokens_seen": 17135792,
      "step": 19785
    },
    {
      "epoch": 9.33050447901933,
      "grad_norm": 0.0006241676746867597,
      "learning_rate": 0.15248567141104974,
      "loss": 0.1923,
      "num_input_tokens_seen": 17140272,
      "step": 19790
    },
    {
      "epoch": 9.332861857614333,
      "grad_norm": 0.0005241729086264968,
      "learning_rate": 0.15242677444692232,
      "loss": 0.1863,
      "num_input_tokens_seen": 17145216,
      "step": 19795
    },
    {
      "epoch": 9.335219236209335,
      "grad_norm": 0.0010514442110434175,
      "learning_rate": 0.15236787710855584,
      "loss": 0.2317,
      "num_input_tokens_seen": 17150160,
      "step": 19800
    },
    {
      "epoch": 9.335219236209335,
      "eval_loss": 0.2187410593032837,
      "eval_runtime": 21.9395,
      "eval_samples_per_second": 42.982,
      "eval_steps_per_second": 21.514,
      "num_input_tokens_seen": 17150160,
      "step": 19800
    },
    {
      "epoch": 9.337576614804338,
      "grad_norm": 0.00046312736230902374,
      "learning_rate": 0.1523089794050329,
      "loss": 0.1852,
      "num_input_tokens_seen": 17154256,
      "step": 19805
    },
    {
      "epoch": 9.33993399339934,
      "grad_norm": 0.0006244800169952214,
      "learning_rate": 0.15225008134543633,
      "loss": 0.2253,
      "num_input_tokens_seen": 17157808,
      "step": 19810
    },
    {
      "epoch": 9.342291371994342,
      "grad_norm": 0.0006691525923088193,
      "learning_rate": 0.15219118293884895,
      "loss": 0.2275,
      "num_input_tokens_seen": 17162048,
      "step": 19815
    },
    {
      "epoch": 9.344648750589345,
      "grad_norm": 0.0008228733786381781,
      "learning_rate": 0.15213228419435362,
      "loss": 0.21,
      "num_input_tokens_seen": 17166272,
      "step": 19820
    },
    {
      "epoch": 9.347006129184347,
      "grad_norm": 0.0007401698967441916,
      "learning_rate": 0.15207338512103327,
      "loss": 0.1788,
      "num_input_tokens_seen": 17170272,
      "step": 19825
    },
    {
      "epoch": 9.34936350777935,
      "grad_norm": 0.001659247325733304,
      "learning_rate": 0.1520144857279709,
      "loss": 0.2671,
      "num_input_tokens_seen": 17174800,
      "step": 19830
    },
    {
      "epoch": 9.351720886374352,
      "grad_norm": 0.0008296079467982054,
      "learning_rate": 0.1519555860242495,
      "loss": 0.2028,
      "num_input_tokens_seen": 17178688,
      "step": 19835
    },
    {
      "epoch": 9.354078264969354,
      "grad_norm": 0.0007022588397376239,
      "learning_rate": 0.15189668601895218,
      "loss": 0.2364,
      "num_input_tokens_seen": 17182720,
      "step": 19840
    },
    {
      "epoch": 9.356435643564357,
      "grad_norm": 0.0005818885983899236,
      "learning_rate": 0.151837785721162,
      "loss": 0.2196,
      "num_input_tokens_seen": 17186480,
      "step": 19845
    },
    {
      "epoch": 9.35879302215936,
      "grad_norm": 0.0005630734376609325,
      "learning_rate": 0.15177888513996218,
      "loss": 0.245,
      "num_input_tokens_seen": 17190896,
      "step": 19850
    },
    {
      "epoch": 9.361150400754362,
      "grad_norm": 0.0005624854238703847,
      "learning_rate": 0.15171998428443592,
      "loss": 0.2181,
      "num_input_tokens_seen": 17195200,
      "step": 19855
    },
    {
      "epoch": 9.363507779349364,
      "grad_norm": 0.0006372167845256627,
      "learning_rate": 0.1516610831636665,
      "loss": 0.2334,
      "num_input_tokens_seen": 17198880,
      "step": 19860
    },
    {
      "epoch": 9.365865157944366,
      "grad_norm": 0.000984302954748273,
      "learning_rate": 0.15160218178673715,
      "loss": 0.228,
      "num_input_tokens_seen": 17203632,
      "step": 19865
    },
    {
      "epoch": 9.368222536539369,
      "grad_norm": 0.0017497262451797724,
      "learning_rate": 0.15154328016273122,
      "loss": 0.2379,
      "num_input_tokens_seen": 17208496,
      "step": 19870
    },
    {
      "epoch": 9.370579915134371,
      "grad_norm": 0.0014233578694984317,
      "learning_rate": 0.1514843783007321,
      "loss": 0.2394,
      "num_input_tokens_seen": 17212576,
      "step": 19875
    },
    {
      "epoch": 9.372937293729374,
      "grad_norm": 0.0006300628301687539,
      "learning_rate": 0.15142547620982322,
      "loss": 0.2096,
      "num_input_tokens_seen": 17216288,
      "step": 19880
    },
    {
      "epoch": 9.375294672324376,
      "grad_norm": 0.0019232876366004348,
      "learning_rate": 0.15136657389908797,
      "loss": 0.2546,
      "num_input_tokens_seen": 17220560,
      "step": 19885
    },
    {
      "epoch": 9.377652050919378,
      "grad_norm": 0.0006782234995625913,
      "learning_rate": 0.15130767137760986,
      "loss": 0.2421,
      "num_input_tokens_seen": 17224368,
      "step": 19890
    },
    {
      "epoch": 9.38000942951438,
      "grad_norm": 0.0014399238862097263,
      "learning_rate": 0.15124876865447243,
      "loss": 0.23,
      "num_input_tokens_seen": 17228464,
      "step": 19895
    },
    {
      "epoch": 9.382366808109383,
      "grad_norm": 0.0007698136032558978,
      "learning_rate": 0.15118986573875912,
      "loss": 0.197,
      "num_input_tokens_seen": 17233552,
      "step": 19900
    },
    {
      "epoch": 9.384724186704386,
      "grad_norm": 0.0006928215734660625,
      "learning_rate": 0.15113096263955358,
      "loss": 0.2636,
      "num_input_tokens_seen": 17237536,
      "step": 19905
    },
    {
      "epoch": 9.387081565299386,
      "grad_norm": 0.0007053091540001333,
      "learning_rate": 0.1510720593659394,
      "loss": 0.2299,
      "num_input_tokens_seen": 17242048,
      "step": 19910
    },
    {
      "epoch": 9.389438943894389,
      "grad_norm": 0.0008625455666333437,
      "learning_rate": 0.15101315592700015,
      "loss": 0.2183,
      "num_input_tokens_seen": 17246000,
      "step": 19915
    },
    {
      "epoch": 9.391796322489391,
      "grad_norm": 0.001172647695057094,
      "learning_rate": 0.15095425233181956,
      "loss": 0.2186,
      "num_input_tokens_seen": 17250096,
      "step": 19920
    },
    {
      "epoch": 9.394153701084393,
      "grad_norm": 0.0006847079494036734,
      "learning_rate": 0.15089534858948128,
      "loss": 0.2301,
      "num_input_tokens_seen": 17255152,
      "step": 19925
    },
    {
      "epoch": 9.396511079679396,
      "grad_norm": 0.0006967468070797622,
      "learning_rate": 0.15083644470906898,
      "loss": 0.2259,
      "num_input_tokens_seen": 17260128,
      "step": 19930
    },
    {
      "epoch": 9.398868458274398,
      "grad_norm": 0.0014693060657009482,
      "learning_rate": 0.1507775406996664,
      "loss": 0.2126,
      "num_input_tokens_seen": 17264064,
      "step": 19935
    },
    {
      "epoch": 9.4012258368694,
      "grad_norm": 0.0007725813193246722,
      "learning_rate": 0.15071863657035725,
      "loss": 0.2332,
      "num_input_tokens_seen": 17268896,
      "step": 19940
    },
    {
      "epoch": 9.403583215464403,
      "grad_norm": 0.0006865648902021348,
      "learning_rate": 0.15065973233022534,
      "loss": 0.2045,
      "num_input_tokens_seen": 17273504,
      "step": 19945
    },
    {
      "epoch": 9.405940594059405,
      "grad_norm": 0.0013231628108769655,
      "learning_rate": 0.15060082798835442,
      "loss": 0.2179,
      "num_input_tokens_seen": 17278608,
      "step": 19950
    },
    {
      "epoch": 9.408297972654408,
      "grad_norm": 0.0006564148352481425,
      "learning_rate": 0.15054192355382823,
      "loss": 0.2085,
      "num_input_tokens_seen": 17282800,
      "step": 19955
    },
    {
      "epoch": 9.41065535124941,
      "grad_norm": 0.0015824259025976062,
      "learning_rate": 0.15048301903573066,
      "loss": 0.2227,
      "num_input_tokens_seen": 17286608,
      "step": 19960
    },
    {
      "epoch": 9.413012729844413,
      "grad_norm": 0.0009594332077540457,
      "learning_rate": 0.15042411444314546,
      "loss": 0.2341,
      "num_input_tokens_seen": 17290912,
      "step": 19965
    },
    {
      "epoch": 9.415370108439415,
      "grad_norm": 0.0011147784534841776,
      "learning_rate": 0.1503652097851565,
      "loss": 0.1642,
      "num_input_tokens_seen": 17294976,
      "step": 19970
    },
    {
      "epoch": 9.417727487034417,
      "grad_norm": 0.0006778439856134355,
      "learning_rate": 0.15030630507084758,
      "loss": 0.1856,
      "num_input_tokens_seen": 17299744,
      "step": 19975
    },
    {
      "epoch": 9.42008486562942,
      "grad_norm": 0.0009395316010341048,
      "learning_rate": 0.1502474003093026,
      "loss": 0.2392,
      "num_input_tokens_seen": 17303296,
      "step": 19980
    },
    {
      "epoch": 9.422442244224422,
      "grad_norm": 0.000885099871084094,
      "learning_rate": 0.15018849550960536,
      "loss": 0.1811,
      "num_input_tokens_seen": 17307472,
      "step": 19985
    },
    {
      "epoch": 9.424799622819425,
      "grad_norm": 0.0012235849862918258,
      "learning_rate": 0.15012959068083975,
      "loss": 0.2407,
      "num_input_tokens_seen": 17312656,
      "step": 19990
    },
    {
      "epoch": 9.427157001414427,
      "grad_norm": 0.0007197518134489655,
      "learning_rate": 0.1500706858320896,
      "loss": 0.1996,
      "num_input_tokens_seen": 17316400,
      "step": 19995
    },
    {
      "epoch": 9.42951438000943,
      "grad_norm": 0.0009691521408967674,
      "learning_rate": 0.15001178097243886,
      "loss": 0.2482,
      "num_input_tokens_seen": 17321280,
      "step": 20000
    },
    {
      "epoch": 9.42951438000943,
      "eval_loss": 0.21536685526371002,
      "eval_runtime": 21.909,
      "eval_samples_per_second": 43.042,
      "eval_steps_per_second": 21.544,
      "num_input_tokens_seen": 17321280,
      "step": 20000
    },
    {
      "epoch": 9.431871758604432,
      "grad_norm": 0.0015848366310819983,
      "learning_rate": 0.1499528761109713,
      "loss": 0.2042,
      "num_input_tokens_seen": 17325056,
      "step": 20005
    },
    {
      "epoch": 9.434229137199434,
      "grad_norm": 0.0013230458134785295,
      "learning_rate": 0.14989397125677087,
      "loss": 0.1813,
      "num_input_tokens_seen": 17329088,
      "step": 20010
    },
    {
      "epoch": 9.436586515794437,
      "grad_norm": 0.0006012945086695254,
      "learning_rate": 0.14983506641892141,
      "loss": 0.2424,
      "num_input_tokens_seen": 17333536,
      "step": 20015
    },
    {
      "epoch": 9.438943894389439,
      "grad_norm": 0.0008526949095539749,
      "learning_rate": 0.14977616160650672,
      "loss": 0.2236,
      "num_input_tokens_seen": 17338048,
      "step": 20020
    },
    {
      "epoch": 9.441301272984441,
      "grad_norm": 0.0006431685760617256,
      "learning_rate": 0.14971725682861076,
      "loss": 0.2106,
      "num_input_tokens_seen": 17341888,
      "step": 20025
    },
    {
      "epoch": 9.443658651579444,
      "grad_norm": 0.0013497823383659124,
      "learning_rate": 0.14965835209431738,
      "loss": 0.22,
      "num_input_tokens_seen": 17346544,
      "step": 20030
    },
    {
      "epoch": 9.446016030174446,
      "grad_norm": 0.0009231448639184237,
      "learning_rate": 0.14959944741271036,
      "loss": 0.1884,
      "num_input_tokens_seen": 17351072,
      "step": 20035
    },
    {
      "epoch": 9.448373408769449,
      "grad_norm": 0.0010897951433435082,
      "learning_rate": 0.14954054279287363,
      "loss": 0.2162,
      "num_input_tokens_seen": 17354960,
      "step": 20040
    },
    {
      "epoch": 9.450730787364451,
      "grad_norm": 0.000543992267921567,
      "learning_rate": 0.14948163824389094,
      "loss": 0.183,
      "num_input_tokens_seen": 17359376,
      "step": 20045
    },
    {
      "epoch": 9.453088165959453,
      "grad_norm": 0.0009608584805391729,
      "learning_rate": 0.14942273377484613,
      "loss": 0.2006,
      "num_input_tokens_seen": 17364336,
      "step": 20050
    },
    {
      "epoch": 9.455445544554456,
      "grad_norm": 0.0006766465958207846,
      "learning_rate": 0.1493638293948231,
      "loss": 0.1832,
      "num_input_tokens_seen": 17368992,
      "step": 20055
    },
    {
      "epoch": 9.457802923149458,
      "grad_norm": 0.0009858724661171436,
      "learning_rate": 0.14930492511290547,
      "loss": 0.2258,
      "num_input_tokens_seen": 17373712,
      "step": 20060
    },
    {
      "epoch": 9.46016030174446,
      "grad_norm": 0.0005642150645144284,
      "learning_rate": 0.14924602093817715,
      "loss": 0.2628,
      "num_input_tokens_seen": 17378304,
      "step": 20065
    },
    {
      "epoch": 9.462517680339463,
      "grad_norm": 0.0012104737106710672,
      "learning_rate": 0.14918711687972194,
      "loss": 0.1878,
      "num_input_tokens_seen": 17382928,
      "step": 20070
    },
    {
      "epoch": 9.464875058934465,
      "grad_norm": 0.000554158235900104,
      "learning_rate": 0.14912821294662346,
      "loss": 0.1932,
      "num_input_tokens_seen": 17387024,
      "step": 20075
    },
    {
      "epoch": 9.467232437529468,
      "grad_norm": 0.0010312390513718128,
      "learning_rate": 0.14906930914796554,
      "loss": 0.232,
      "num_input_tokens_seen": 17391056,
      "step": 20080
    },
    {
      "epoch": 9.46958981612447,
      "grad_norm": 0.0011156884720548987,
      "learning_rate": 0.14901040549283182,
      "loss": 0.1705,
      "num_input_tokens_seen": 17394944,
      "step": 20085
    },
    {
      "epoch": 9.471947194719473,
      "grad_norm": 0.0004572556645143777,
      "learning_rate": 0.148951501990306,
      "loss": 0.2236,
      "num_input_tokens_seen": 17398608,
      "step": 20090
    },
    {
      "epoch": 9.474304573314475,
      "grad_norm": 0.0005564548191614449,
      "learning_rate": 0.14889259864947177,
      "loss": 0.2522,
      "num_input_tokens_seen": 17402736,
      "step": 20095
    },
    {
      "epoch": 9.476661951909477,
      "grad_norm": 0.0013159830123186111,
      "learning_rate": 0.14883369547941272,
      "loss": 0.2486,
      "num_input_tokens_seen": 17406784,
      "step": 20100
    },
    {
      "epoch": 9.47901933050448,
      "grad_norm": 0.0006625318783335388,
      "learning_rate": 0.14877479248921247,
      "loss": 0.1941,
      "num_input_tokens_seen": 17411296,
      "step": 20105
    },
    {
      "epoch": 9.481376709099482,
      "grad_norm": 0.0005537057877518237,
      "learning_rate": 0.14871588968795468,
      "loss": 0.2208,
      "num_input_tokens_seen": 17415616,
      "step": 20110
    },
    {
      "epoch": 9.483734087694483,
      "grad_norm": 0.002039598533883691,
      "learning_rate": 0.1486569870847228,
      "loss": 0.249,
      "num_input_tokens_seen": 17420560,
      "step": 20115
    },
    {
      "epoch": 9.486091466289485,
      "grad_norm": 0.0006275964551605284,
      "learning_rate": 0.1485980846886004,
      "loss": 0.2037,
      "num_input_tokens_seen": 17424816,
      "step": 20120
    },
    {
      "epoch": 9.488448844884488,
      "grad_norm": 0.0006537820445373654,
      "learning_rate": 0.14853918250867096,
      "loss": 0.1896,
      "num_input_tokens_seen": 17429344,
      "step": 20125
    },
    {
      "epoch": 9.49080622347949,
      "grad_norm": 0.0008272115956060588,
      "learning_rate": 0.1484802805540179,
      "loss": 0.2213,
      "num_input_tokens_seen": 17434160,
      "step": 20130
    },
    {
      "epoch": 9.493163602074493,
      "grad_norm": 0.0013498219195753336,
      "learning_rate": 0.14842137883372472,
      "loss": 0.2823,
      "num_input_tokens_seen": 17438656,
      "step": 20135
    },
    {
      "epoch": 9.495520980669495,
      "grad_norm": 0.0005481025436893106,
      "learning_rate": 0.14836247735687474,
      "loss": 0.2403,
      "num_input_tokens_seen": 17443888,
      "step": 20140
    },
    {
      "epoch": 9.497878359264497,
      "grad_norm": 0.000894532073289156,
      "learning_rate": 0.14830357613255132,
      "loss": 0.2199,
      "num_input_tokens_seen": 17448336,
      "step": 20145
    },
    {
      "epoch": 9.5002357378595,
      "grad_norm": 0.0008404235704801977,
      "learning_rate": 0.1482446751698378,
      "loss": 0.1944,
      "num_input_tokens_seen": 17452928,
      "step": 20150
    },
    {
      "epoch": 9.502593116454502,
      "grad_norm": 0.0005628942744806409,
      "learning_rate": 0.14818577447781744,
      "loss": 0.2252,
      "num_input_tokens_seen": 17456848,
      "step": 20155
    },
    {
      "epoch": 9.504950495049505,
      "grad_norm": 0.0015349660534411669,
      "learning_rate": 0.14812687406557346,
      "loss": 0.2287,
      "num_input_tokens_seen": 17460656,
      "step": 20160
    },
    {
      "epoch": 9.507307873644507,
      "grad_norm": 0.0012589588295668364,
      "learning_rate": 0.14806797394218899,
      "loss": 0.2047,
      "num_input_tokens_seen": 17464528,
      "step": 20165
    },
    {
      "epoch": 9.50966525223951,
      "grad_norm": 0.0008098451653495431,
      "learning_rate": 0.1480090741167472,
      "loss": 0.2446,
      "num_input_tokens_seen": 17468544,
      "step": 20170
    },
    {
      "epoch": 9.512022630834512,
      "grad_norm": 0.0006556081352755427,
      "learning_rate": 0.1479501745983313,
      "loss": 0.208,
      "num_input_tokens_seen": 17472576,
      "step": 20175
    },
    {
      "epoch": 9.514380009429514,
      "grad_norm": 0.0006366188172250986,
      "learning_rate": 0.14789127539602415,
      "loss": 0.2349,
      "num_input_tokens_seen": 17476688,
      "step": 20180
    },
    {
      "epoch": 9.516737388024517,
      "grad_norm": 0.0008537110988982022,
      "learning_rate": 0.14783237651890885,
      "loss": 0.2177,
      "num_input_tokens_seen": 17481872,
      "step": 20185
    },
    {
      "epoch": 9.519094766619519,
      "grad_norm": 0.0016984982648864388,
      "learning_rate": 0.14777347797606838,
      "loss": 0.2476,
      "num_input_tokens_seen": 17486224,
      "step": 20190
    },
    {
      "epoch": 9.521452145214521,
      "grad_norm": 0.0008916830993257463,
      "learning_rate": 0.14771457977658553,
      "loss": 0.2133,
      "num_input_tokens_seen": 17491120,
      "step": 20195
    },
    {
      "epoch": 9.523809523809524,
      "grad_norm": 0.0019968922715634108,
      "learning_rate": 0.14765568192954326,
      "loss": 0.2261,
      "num_input_tokens_seen": 17495488,
      "step": 20200
    },
    {
      "epoch": 9.523809523809524,
      "eval_loss": 0.21515996754169464,
      "eval_runtime": 21.9539,
      "eval_samples_per_second": 42.954,
      "eval_steps_per_second": 21.5,
      "num_input_tokens_seen": 17495488,
      "step": 20200
    },
    {
      "epoch": 9.526166902404526,
      "grad_norm": 0.0008885944844223559,
      "learning_rate": 0.14759678444402421,
      "loss": 0.2301,
      "num_input_tokens_seen": 17499936,
      "step": 20205
    },
    {
      "epoch": 9.528524280999529,
      "grad_norm": 0.002289389492943883,
      "learning_rate": 0.14753788732911122,
      "loss": 0.2337,
      "num_input_tokens_seen": 17504512,
      "step": 20210
    },
    {
      "epoch": 9.530881659594531,
      "grad_norm": 0.0009423093288205564,
      "learning_rate": 0.147478990593887,
      "loss": 0.2233,
      "num_input_tokens_seen": 17508800,
      "step": 20215
    },
    {
      "epoch": 9.533239038189533,
      "grad_norm": 0.0011014282936230302,
      "learning_rate": 0.14742009424743405,
      "loss": 0.2212,
      "num_input_tokens_seen": 17512608,
      "step": 20220
    },
    {
      "epoch": 9.535596416784536,
      "grad_norm": 0.0014416513731703162,
      "learning_rate": 0.14736119829883504,
      "loss": 0.192,
      "num_input_tokens_seen": 17517376,
      "step": 20225
    },
    {
      "epoch": 9.537953795379538,
      "grad_norm": 0.0018658344633877277,
      "learning_rate": 0.14730230275717243,
      "loss": 0.2104,
      "num_input_tokens_seen": 17521360,
      "step": 20230
    },
    {
      "epoch": 9.54031117397454,
      "grad_norm": 0.000669886649120599,
      "learning_rate": 0.14724340763152854,
      "loss": 0.2502,
      "num_input_tokens_seen": 17525440,
      "step": 20235
    },
    {
      "epoch": 9.542668552569543,
      "grad_norm": 0.0015540856402367353,
      "learning_rate": 0.14718451293098594,
      "loss": 0.2608,
      "num_input_tokens_seen": 17530208,
      "step": 20240
    },
    {
      "epoch": 9.545025931164545,
      "grad_norm": 0.0008324977243319154,
      "learning_rate": 0.14712561866462676,
      "loss": 0.2026,
      "num_input_tokens_seen": 17534224,
      "step": 20245
    },
    {
      "epoch": 9.547383309759548,
      "grad_norm": 0.001377228181809187,
      "learning_rate": 0.1470667248415333,
      "loss": 0.223,
      "num_input_tokens_seen": 17538768,
      "step": 20250
    },
    {
      "epoch": 9.54974068835455,
      "grad_norm": 0.0006202780641615391,
      "learning_rate": 0.1470078314707878,
      "loss": 0.1932,
      "num_input_tokens_seen": 17543088,
      "step": 20255
    },
    {
      "epoch": 9.552098066949553,
      "grad_norm": 0.0014777590986341238,
      "learning_rate": 0.14694893856147223,
      "loss": 0.2044,
      "num_input_tokens_seen": 17547984,
      "step": 20260
    },
    {
      "epoch": 9.554455445544555,
      "grad_norm": 0.0009235324687324464,
      "learning_rate": 0.14689004612266868,
      "loss": 0.2171,
      "num_input_tokens_seen": 17551808,
      "step": 20265
    },
    {
      "epoch": 9.556812824139557,
      "grad_norm": 0.0007294192910194397,
      "learning_rate": 0.14683115416345913,
      "loss": 0.2175,
      "num_input_tokens_seen": 17555456,
      "step": 20270
    },
    {
      "epoch": 9.55917020273456,
      "grad_norm": 0.0005667241639457643,
      "learning_rate": 0.1467722626929254,
      "loss": 0.1927,
      "num_input_tokens_seen": 17559840,
      "step": 20275
    },
    {
      "epoch": 9.561527581329562,
      "grad_norm": 0.000680093711707741,
      "learning_rate": 0.14671337172014937,
      "loss": 0.2132,
      "num_input_tokens_seen": 17564016,
      "step": 20280
    },
    {
      "epoch": 9.563884959924565,
      "grad_norm": 0.0006578586180694401,
      "learning_rate": 0.14665448125421265,
      "loss": 0.2001,
      "num_input_tokens_seen": 17568320,
      "step": 20285
    },
    {
      "epoch": 9.566242338519567,
      "grad_norm": 0.0010942034423351288,
      "learning_rate": 0.146595591304197,
      "loss": 0.2135,
      "num_input_tokens_seen": 17572544,
      "step": 20290
    },
    {
      "epoch": 9.56859971711457,
      "grad_norm": 0.0013192775659263134,
      "learning_rate": 0.14653670187918397,
      "loss": 0.2201,
      "num_input_tokens_seen": 17576704,
      "step": 20295
    },
    {
      "epoch": 9.570957095709572,
      "grad_norm": 0.0010448156390339136,
      "learning_rate": 0.14647781298825502,
      "loss": 0.2155,
      "num_input_tokens_seen": 17581056,
      "step": 20300
    },
    {
      "epoch": 9.573314474304574,
      "grad_norm": 0.0011846715351566672,
      "learning_rate": 0.14641892464049153,
      "loss": 0.2374,
      "num_input_tokens_seen": 17585168,
      "step": 20305
    },
    {
      "epoch": 9.575671852899575,
      "grad_norm": 0.0007550598820671439,
      "learning_rate": 0.14636003684497495,
      "loss": 0.2525,
      "num_input_tokens_seen": 17589472,
      "step": 20310
    },
    {
      "epoch": 9.578029231494579,
      "grad_norm": 0.0009879018180072308,
      "learning_rate": 0.14630114961078636,
      "loss": 0.2191,
      "num_input_tokens_seen": 17593456,
      "step": 20315
    },
    {
      "epoch": 9.58038661008958,
      "grad_norm": 0.0016299979761242867,
      "learning_rate": 0.14624226294700704,
      "loss": 0.2025,
      "num_input_tokens_seen": 17598576,
      "step": 20320
    },
    {
      "epoch": 9.582743988684582,
      "grad_norm": 0.0014353878796100616,
      "learning_rate": 0.14618337686271793,
      "loss": 0.1863,
      "num_input_tokens_seen": 17602704,
      "step": 20325
    },
    {
      "epoch": 9.585101367279584,
      "grad_norm": 0.0018011172069236636,
      "learning_rate": 0.1461244913670001,
      "loss": 0.2806,
      "num_input_tokens_seen": 17607328,
      "step": 20330
    },
    {
      "epoch": 9.587458745874587,
      "grad_norm": 0.0008793168235570192,
      "learning_rate": 0.1460656064689344,
      "loss": 0.2207,
      "num_input_tokens_seen": 17611696,
      "step": 20335
    },
    {
      "epoch": 9.58981612446959,
      "grad_norm": 0.0010298278648406267,
      "learning_rate": 0.14600672217760163,
      "loss": 0.2078,
      "num_input_tokens_seen": 17616928,
      "step": 20340
    },
    {
      "epoch": 9.592173503064592,
      "grad_norm": 0.0009212171426042914,
      "learning_rate": 0.14594783850208248,
      "loss": 0.2157,
      "num_input_tokens_seen": 17621360,
      "step": 20345
    },
    {
      "epoch": 9.594530881659594,
      "grad_norm": 0.0017712191911414266,
      "learning_rate": 0.14588895545145758,
      "loss": 0.2021,
      "num_input_tokens_seen": 17625952,
      "step": 20350
    },
    {
      "epoch": 9.596888260254596,
      "grad_norm": 0.0013275428209453821,
      "learning_rate": 0.14583007303480738,
      "loss": 0.2012,
      "num_input_tokens_seen": 17630480,
      "step": 20355
    },
    {
      "epoch": 9.599245638849599,
      "grad_norm": 0.0010387555230408907,
      "learning_rate": 0.14577119126121235,
      "loss": 0.2301,
      "num_input_tokens_seen": 17634944,
      "step": 20360
    },
    {
      "epoch": 9.601603017444601,
      "grad_norm": 0.0009452144149690866,
      "learning_rate": 0.14571231013975272,
      "loss": 0.2229,
      "num_input_tokens_seen": 17638992,
      "step": 20365
    },
    {
      "epoch": 9.603960396039604,
      "grad_norm": 0.0016455514123663306,
      "learning_rate": 0.1456534296795088,
      "loss": 0.2078,
      "num_input_tokens_seen": 17643328,
      "step": 20370
    },
    {
      "epoch": 9.606317774634606,
      "grad_norm": 0.0008877373184077442,
      "learning_rate": 0.14559454988956066,
      "loss": 0.2094,
      "num_input_tokens_seen": 17648112,
      "step": 20375
    },
    {
      "epoch": 9.608675153229608,
      "grad_norm": 0.0016427877126261592,
      "learning_rate": 0.1455356707789882,
      "loss": 0.2501,
      "num_input_tokens_seen": 17653600,
      "step": 20380
    },
    {
      "epoch": 9.61103253182461,
      "grad_norm": 0.0008372567826882005,
      "learning_rate": 0.14547679235687147,
      "loss": 0.1835,
      "num_input_tokens_seen": 17658096,
      "step": 20385
    },
    {
      "epoch": 9.613389910419613,
      "grad_norm": 0.001427405164577067,
      "learning_rate": 0.14541791463229023,
      "loss": 0.2384,
      "num_input_tokens_seen": 17661776,
      "step": 20390
    },
    {
      "epoch": 9.615747289014616,
      "grad_norm": 0.0012673120945692062,
      "learning_rate": 0.14535903761432406,
      "loss": 0.1968,
      "num_input_tokens_seen": 17665664,
      "step": 20395
    },
    {
      "epoch": 9.618104667609618,
      "grad_norm": 0.001242350204847753,
      "learning_rate": 0.1453001613120527,
      "loss": 0.2319,
      "num_input_tokens_seen": 17670576,
      "step": 20400
    },
    {
      "epoch": 9.618104667609618,
      "eval_loss": 0.2168968915939331,
      "eval_runtime": 21.9028,
      "eval_samples_per_second": 43.054,
      "eval_steps_per_second": 21.55,
      "num_input_tokens_seen": 17670576,
      "step": 20400
    },
    {
      "epoch": 9.62046204620462,
      "grad_norm": 0.0018380923429504037,
      "learning_rate": 0.14524128573455547,
      "loss": 0.2208,
      "num_input_tokens_seen": 17674976,
      "step": 20405
    },
    {
      "epoch": 9.622819424799623,
      "grad_norm": 0.0005885528516955674,
      "learning_rate": 0.14518241089091177,
      "loss": 0.2693,
      "num_input_tokens_seen": 17679840,
      "step": 20410
    },
    {
      "epoch": 9.625176803394625,
      "grad_norm": 0.000979065545834601,
      "learning_rate": 0.1451235367902009,
      "loss": 0.216,
      "num_input_tokens_seen": 17684336,
      "step": 20415
    },
    {
      "epoch": 9.627534181989628,
      "grad_norm": 0.0008213087567128241,
      "learning_rate": 0.1450646634415019,
      "loss": 0.2207,
      "num_input_tokens_seen": 17688432,
      "step": 20420
    },
    {
      "epoch": 9.62989156058463,
      "grad_norm": 0.0007264121668413281,
      "learning_rate": 0.1450057908538938,
      "loss": 0.2226,
      "num_input_tokens_seen": 17692896,
      "step": 20425
    },
    {
      "epoch": 9.632248939179632,
      "grad_norm": 0.001003630575723946,
      "learning_rate": 0.14494691903645557,
      "loss": 0.2221,
      "num_input_tokens_seen": 17696560,
      "step": 20430
    },
    {
      "epoch": 9.634606317774635,
      "grad_norm": 0.0007365955971181393,
      "learning_rate": 0.14488804799826588,
      "loss": 0.2095,
      "num_input_tokens_seen": 17700432,
      "step": 20435
    },
    {
      "epoch": 9.636963696369637,
      "grad_norm": 0.001128949224948883,
      "learning_rate": 0.14482917774840348,
      "loss": 0.2217,
      "num_input_tokens_seen": 17704800,
      "step": 20440
    },
    {
      "epoch": 9.63932107496464,
      "grad_norm": 0.001799715799279511,
      "learning_rate": 0.14477030829594684,
      "loss": 0.2214,
      "num_input_tokens_seen": 17708368,
      "step": 20445
    },
    {
      "epoch": 9.641678453559642,
      "grad_norm": 0.0006580776534974575,
      "learning_rate": 0.14471143964997432,
      "loss": 0.2032,
      "num_input_tokens_seen": 17712976,
      "step": 20450
    },
    {
      "epoch": 9.644035832154644,
      "grad_norm": 0.0008678776794113219,
      "learning_rate": 0.14465257181956434,
      "loss": 0.2067,
      "num_input_tokens_seen": 17716992,
      "step": 20455
    },
    {
      "epoch": 9.646393210749647,
      "grad_norm": 0.0012524538906291127,
      "learning_rate": 0.1445937048137949,
      "loss": 0.2276,
      "num_input_tokens_seen": 17721200,
      "step": 20460
    },
    {
      "epoch": 9.64875058934465,
      "grad_norm": 0.0017552762292325497,
      "learning_rate": 0.14453483864174416,
      "loss": 0.2188,
      "num_input_tokens_seen": 17724432,
      "step": 20465
    },
    {
      "epoch": 9.651107967939652,
      "grad_norm": 0.000949060486163944,
      "learning_rate": 0.14447597331249,
      "loss": 0.2078,
      "num_input_tokens_seen": 17728928,
      "step": 20470
    },
    {
      "epoch": 9.653465346534654,
      "grad_norm": 0.0008475234499201179,
      "learning_rate": 0.1444171088351102,
      "loss": 0.2316,
      "num_input_tokens_seen": 17732768,
      "step": 20475
    },
    {
      "epoch": 9.655822725129656,
      "grad_norm": 0.0011733185965567827,
      "learning_rate": 0.14435824521868235,
      "loss": 0.2249,
      "num_input_tokens_seen": 17737584,
      "step": 20480
    },
    {
      "epoch": 9.658180103724659,
      "grad_norm": 0.0015648695407435298,
      "learning_rate": 0.14429938247228397,
      "loss": 0.1942,
      "num_input_tokens_seen": 17742608,
      "step": 20485
    },
    {
      "epoch": 9.660537482319661,
      "grad_norm": 0.0010379502782598138,
      "learning_rate": 0.14424052060499243,
      "loss": 0.2574,
      "num_input_tokens_seen": 17747696,
      "step": 20490
    },
    {
      "epoch": 9.662894860914664,
      "grad_norm": 0.0015979685122147202,
      "learning_rate": 0.14418165962588506,
      "loss": 0.2239,
      "num_input_tokens_seen": 17752144,
      "step": 20495
    },
    {
      "epoch": 9.665252239509666,
      "grad_norm": 0.001337651745416224,
      "learning_rate": 0.1441227995440388,
      "loss": 0.1875,
      "num_input_tokens_seen": 17756320,
      "step": 20500
    },
    {
      "epoch": 9.667609618104667,
      "grad_norm": 0.0015231596771627665,
      "learning_rate": 0.14406394036853082,
      "loss": 0.2081,
      "num_input_tokens_seen": 17760304,
      "step": 20505
    },
    {
      "epoch": 9.66996699669967,
      "grad_norm": 0.0007395526627078652,
      "learning_rate": 0.14400508210843774,
      "loss": 0.1752,
      "num_input_tokens_seen": 17765152,
      "step": 20510
    },
    {
      "epoch": 9.672324375294671,
      "grad_norm": 0.000899964477866888,
      "learning_rate": 0.1439462247728364,
      "loss": 0.2538,
      "num_input_tokens_seen": 17769280,
      "step": 20515
    },
    {
      "epoch": 9.674681753889674,
      "grad_norm": 0.0006994422874413431,
      "learning_rate": 0.14388736837080326,
      "loss": 0.2329,
      "num_input_tokens_seen": 17774432,
      "step": 20520
    },
    {
      "epoch": 9.677039132484676,
      "grad_norm": 0.0015399340773001313,
      "learning_rate": 0.14382851291141469,
      "loss": 0.2315,
      "num_input_tokens_seen": 17778400,
      "step": 20525
    },
    {
      "epoch": 9.679396511079679,
      "grad_norm": 0.0029570970218628645,
      "learning_rate": 0.14376965840374697,
      "loss": 0.2202,
      "num_input_tokens_seen": 17782720,
      "step": 20530
    },
    {
      "epoch": 9.681753889674681,
      "grad_norm": 0.0016727304318919778,
      "learning_rate": 0.14371080485687632,
      "loss": 0.2367,
      "num_input_tokens_seen": 17786832,
      "step": 20535
    },
    {
      "epoch": 9.684111268269683,
      "grad_norm": 0.0007829573005437851,
      "learning_rate": 0.1436519522798785,
      "loss": 0.22,
      "num_input_tokens_seen": 17790912,
      "step": 20540
    },
    {
      "epoch": 9.686468646864686,
      "grad_norm": 0.0008653285913169384,
      "learning_rate": 0.14359310068182948,
      "loss": 0.225,
      "num_input_tokens_seen": 17795424,
      "step": 20545
    },
    {
      "epoch": 9.688826025459688,
      "grad_norm": 0.0015246745897457004,
      "learning_rate": 0.14353425007180484,
      "loss": 0.1786,
      "num_input_tokens_seen": 17799744,
      "step": 20550
    },
    {
      "epoch": 9.69118340405469,
      "grad_norm": 0.0006649324204772711,
      "learning_rate": 0.14347540045888005,
      "loss": 0.2196,
      "num_input_tokens_seen": 17803872,
      "step": 20555
    },
    {
      "epoch": 9.693540782649693,
      "grad_norm": 0.0006599709740839899,
      "learning_rate": 0.14341655185213056,
      "loss": 0.2309,
      "num_input_tokens_seen": 17808880,
      "step": 20560
    },
    {
      "epoch": 9.695898161244696,
      "grad_norm": 0.0017684876220300794,
      "learning_rate": 0.14335770426063144,
      "loss": 0.2448,
      "num_input_tokens_seen": 17813328,
      "step": 20565
    },
    {
      "epoch": 9.698255539839698,
      "grad_norm": 0.0007855689036659896,
      "learning_rate": 0.1432988576934578,
      "loss": 0.1782,
      "num_input_tokens_seen": 17817760,
      "step": 20570
    },
    {
      "epoch": 9.7006129184347,
      "grad_norm": 0.000948451750446111,
      "learning_rate": 0.14324001215968457,
      "loss": 0.177,
      "num_input_tokens_seen": 17822080,
      "step": 20575
    },
    {
      "epoch": 9.702970297029703,
      "grad_norm": 0.0006067260983400047,
      "learning_rate": 0.14318116766838637,
      "loss": 0.2121,
      "num_input_tokens_seen": 17825984,
      "step": 20580
    },
    {
      "epoch": 9.705327675624705,
      "grad_norm": 0.0026487938594073057,
      "learning_rate": 0.14312232422863788,
      "loss": 0.2503,
      "num_input_tokens_seen": 17830240,
      "step": 20585
    },
    {
      "epoch": 9.707685054219708,
      "grad_norm": 0.0006089310627430677,
      "learning_rate": 0.14306348184951334,
      "loss": 0.2143,
      "num_input_tokens_seen": 17834912,
      "step": 20590
    },
    {
      "epoch": 9.71004243281471,
      "grad_norm": 0.000669643864966929,
      "learning_rate": 0.1430046405400871,
      "loss": 0.2116,
      "num_input_tokens_seen": 17838576,
      "step": 20595
    },
    {
      "epoch": 9.712399811409712,
      "grad_norm": 0.001280271215364337,
      "learning_rate": 0.14294580030943324,
      "loss": 0.205,
      "num_input_tokens_seen": 17843440,
      "step": 20600
    },
    {
      "epoch": 9.712399811409712,
      "eval_loss": 0.215559184551239,
      "eval_runtime": 21.9587,
      "eval_samples_per_second": 42.944,
      "eval_steps_per_second": 21.495,
      "num_input_tokens_seen": 17843440,
      "step": 20600
    },
    {
      "epoch": 9.714757190004715,
      "grad_norm": 0.000786292483098805,
      "learning_rate": 0.14288696116662553,
      "loss": 0.2386,
      "num_input_tokens_seen": 17848352,
      "step": 20605
    },
    {
      "epoch": 9.717114568599717,
      "grad_norm": 0.0011515662772580981,
      "learning_rate": 0.1428281231207378,
      "loss": 0.2102,
      "num_input_tokens_seen": 17852848,
      "step": 20610
    },
    {
      "epoch": 9.71947194719472,
      "grad_norm": 0.0011113951914012432,
      "learning_rate": 0.1427692861808437,
      "loss": 0.2242,
      "num_input_tokens_seen": 17856944,
      "step": 20615
    },
    {
      "epoch": 9.721829325789722,
      "grad_norm": 0.0015225758543238044,
      "learning_rate": 0.1427104503560165,
      "loss": 0.2231,
      "num_input_tokens_seen": 17861616,
      "step": 20620
    },
    {
      "epoch": 9.724186704384724,
      "grad_norm": 0.0008812613668851554,
      "learning_rate": 0.14265161565532947,
      "loss": 0.2053,
      "num_input_tokens_seen": 17865472,
      "step": 20625
    },
    {
      "epoch": 9.726544082979727,
      "grad_norm": 0.0010320247383788228,
      "learning_rate": 0.14259278208785564,
      "loss": 0.2165,
      "num_input_tokens_seen": 17870048,
      "step": 20630
    },
    {
      "epoch": 9.72890146157473,
      "grad_norm": 0.0008569251513108611,
      "learning_rate": 0.14253394966266789,
      "loss": 0.2451,
      "num_input_tokens_seen": 17874512,
      "step": 20635
    },
    {
      "epoch": 9.731258840169732,
      "grad_norm": 0.0013477386673912406,
      "learning_rate": 0.14247511838883894,
      "loss": 0.199,
      "num_input_tokens_seen": 17879136,
      "step": 20640
    },
    {
      "epoch": 9.733616218764734,
      "grad_norm": 0.0008317263564094901,
      "learning_rate": 0.14241628827544126,
      "loss": 0.2421,
      "num_input_tokens_seen": 17882528,
      "step": 20645
    },
    {
      "epoch": 9.735973597359736,
      "grad_norm": 0.0007979859365150332,
      "learning_rate": 0.14235745933154723,
      "loss": 0.184,
      "num_input_tokens_seen": 17886176,
      "step": 20650
    },
    {
      "epoch": 9.738330975954739,
      "grad_norm": 0.0023083535488694906,
      "learning_rate": 0.14229863156622907,
      "loss": 0.2672,
      "num_input_tokens_seen": 17890720,
      "step": 20655
    },
    {
      "epoch": 9.740688354549741,
      "grad_norm": 0.0011859270744025707,
      "learning_rate": 0.14223980498855868,
      "loss": 0.2582,
      "num_input_tokens_seen": 17894688,
      "step": 20660
    },
    {
      "epoch": 9.743045733144744,
      "grad_norm": 0.001322311582043767,
      "learning_rate": 0.14218097960760792,
      "loss": 0.2282,
      "num_input_tokens_seen": 17898480,
      "step": 20665
    },
    {
      "epoch": 9.745403111739746,
      "grad_norm": 0.001723092282190919,
      "learning_rate": 0.1421221554324483,
      "loss": 0.2162,
      "num_input_tokens_seen": 17902384,
      "step": 20670
    },
    {
      "epoch": 9.747760490334748,
      "grad_norm": 0.0013819780433550477,
      "learning_rate": 0.1420633324721513,
      "loss": 0.2319,
      "num_input_tokens_seen": 17906288,
      "step": 20675
    },
    {
      "epoch": 9.75011786892975,
      "grad_norm": 0.0023326657246798277,
      "learning_rate": 0.14200451073578824,
      "loss": 0.2244,
      "num_input_tokens_seen": 17910800,
      "step": 20680
    },
    {
      "epoch": 9.752475247524753,
      "grad_norm": 0.002872682409361005,
      "learning_rate": 0.14194569023243003,
      "loss": 0.2134,
      "num_input_tokens_seen": 17915488,
      "step": 20685
    },
    {
      "epoch": 9.754832626119756,
      "grad_norm": 0.0032045929692685604,
      "learning_rate": 0.14188687097114766,
      "loss": 0.2345,
      "num_input_tokens_seen": 17919136,
      "step": 20690
    },
    {
      "epoch": 9.757190004714758,
      "grad_norm": 0.0019054177682846785,
      "learning_rate": 0.14182805296101172,
      "loss": 0.2294,
      "num_input_tokens_seen": 17922928,
      "step": 20695
    },
    {
      "epoch": 9.75954738330976,
      "grad_norm": 0.003585758153349161,
      "learning_rate": 0.14176923621109272,
      "loss": 0.2103,
      "num_input_tokens_seen": 17927984,
      "step": 20700
    },
    {
      "epoch": 9.761904761904763,
      "grad_norm": 0.00202233181335032,
      "learning_rate": 0.14171042073046097,
      "loss": 0.2179,
      "num_input_tokens_seen": 17932144,
      "step": 20705
    },
    {
      "epoch": 9.764262140499763,
      "grad_norm": 0.0027521983720362186,
      "learning_rate": 0.14165160652818642,
      "loss": 0.2382,
      "num_input_tokens_seen": 17936032,
      "step": 20710
    },
    {
      "epoch": 9.766619519094768,
      "grad_norm": 0.0029339382890611887,
      "learning_rate": 0.14159279361333907,
      "loss": 0.2268,
      "num_input_tokens_seen": 17940464,
      "step": 20715
    },
    {
      "epoch": 9.768976897689768,
      "grad_norm": 0.0013346857158467174,
      "learning_rate": 0.14153398199498868,
      "loss": 0.2217,
      "num_input_tokens_seen": 17944352,
      "step": 20720
    },
    {
      "epoch": 9.77133427628477,
      "grad_norm": 0.0030369276646524668,
      "learning_rate": 0.14147517168220458,
      "loss": 0.201,
      "num_input_tokens_seen": 17948528,
      "step": 20725
    },
    {
      "epoch": 9.773691654879773,
      "grad_norm": 0.002730417065322399,
      "learning_rate": 0.14141636268405616,
      "loss": 0.2044,
      "num_input_tokens_seen": 17952496,
      "step": 20730
    },
    {
      "epoch": 9.776049033474775,
      "grad_norm": 0.0032366071827709675,
      "learning_rate": 0.14135755500961253,
      "loss": 0.2912,
      "num_input_tokens_seen": 17956880,
      "step": 20735
    },
    {
      "epoch": 9.778406412069778,
      "grad_norm": 0.004075288772583008,
      "learning_rate": 0.14129874866794245,
      "loss": 0.2262,
      "num_input_tokens_seen": 17961392,
      "step": 20740
    },
    {
      "epoch": 9.78076379066478,
      "grad_norm": 0.0016800656449049711,
      "learning_rate": 0.14123994366811476,
      "loss": 0.2186,
      "num_input_tokens_seen": 17965568,
      "step": 20745
    },
    {
      "epoch": 9.783121169259783,
      "grad_norm": 0.0035225781612098217,
      "learning_rate": 0.14118114001919774,
      "loss": 0.2123,
      "num_input_tokens_seen": 17969648,
      "step": 20750
    },
    {
      "epoch": 9.785478547854785,
      "grad_norm": 0.0032214259263128042,
      "learning_rate": 0.14112233773025978,
      "loss": 0.1869,
      "num_input_tokens_seen": 17974256,
      "step": 20755
    },
    {
      "epoch": 9.787835926449787,
      "grad_norm": 0.003996030893176794,
      "learning_rate": 0.14106353681036896,
      "loss": 0.2589,
      "num_input_tokens_seen": 17978192,
      "step": 20760
    },
    {
      "epoch": 9.79019330504479,
      "grad_norm": 0.003911331295967102,
      "learning_rate": 0.14100473726859303,
      "loss": 0.2196,
      "num_input_tokens_seen": 17981904,
      "step": 20765
    },
    {
      "epoch": 9.792550683639792,
      "grad_norm": 0.010588770732283592,
      "learning_rate": 0.14094593911399964,
      "loss": 0.2317,
      "num_input_tokens_seen": 17986512,
      "step": 20770
    },
    {
      "epoch": 9.794908062234795,
      "grad_norm": 0.008972296491265297,
      "learning_rate": 0.14088714235565625,
      "loss": 0.2462,
      "num_input_tokens_seen": 17990464,
      "step": 20775
    },
    {
      "epoch": 9.797265440829797,
      "grad_norm": 0.009860249236226082,
      "learning_rate": 0.14082834700263,
      "loss": 0.2429,
      "num_input_tokens_seen": 17995040,
      "step": 20780
    },
    {
      "epoch": 9.7996228194248,
      "grad_norm": 0.00562833808362484,
      "learning_rate": 0.14076955306398795,
      "loss": 0.233,
      "num_input_tokens_seen": 17999104,
      "step": 20785
    },
    {
      "epoch": 9.801980198019802,
      "grad_norm": 0.004770326893776655,
      "learning_rate": 0.14071076054879675,
      "loss": 0.2429,
      "num_input_tokens_seen": 18003776,
      "step": 20790
    },
    {
      "epoch": 9.804337576614804,
      "grad_norm": 0.010428115725517273,
      "learning_rate": 0.14065196946612302,
      "loss": 0.2426,
      "num_input_tokens_seen": 18008064,
      "step": 20795
    },
    {
      "epoch": 9.806694955209807,
      "grad_norm": 0.0032120042014867067,
      "learning_rate": 0.1405931798250331,
      "loss": 0.2642,
      "num_input_tokens_seen": 18012496,
      "step": 20800
    },
    {
      "epoch": 9.806694955209807,
      "eval_loss": 0.2300536334514618,
      "eval_runtime": 21.8914,
      "eval_samples_per_second": 43.076,
      "eval_steps_per_second": 21.561,
      "num_input_tokens_seen": 18012496,
      "step": 20800
    },
    {
      "epoch": 9.809052333804809,
      "grad_norm": 0.0030970994848757982,
      "learning_rate": 0.14053439163459308,
      "loss": 0.2422,
      "num_input_tokens_seen": 18016016,
      "step": 20805
    },
    {
      "epoch": 9.811409712399811,
      "grad_norm": 0.005022671073675156,
      "learning_rate": 0.14047560490386876,
      "loss": 0.2307,
      "num_input_tokens_seen": 18019792,
      "step": 20810
    },
    {
      "epoch": 9.813767090994814,
      "grad_norm": 0.0026529813185334206,
      "learning_rate": 0.14041681964192593,
      "loss": 0.1726,
      "num_input_tokens_seen": 18023792,
      "step": 20815
    },
    {
      "epoch": 9.816124469589816,
      "grad_norm": 0.003863097634166479,
      "learning_rate": 0.14035803585782988,
      "loss": 0.3141,
      "num_input_tokens_seen": 18027696,
      "step": 20820
    },
    {
      "epoch": 9.818481848184819,
      "grad_norm": 0.0015404460718855262,
      "learning_rate": 0.14029925356064593,
      "loss": 0.2269,
      "num_input_tokens_seen": 18031760,
      "step": 20825
    },
    {
      "epoch": 9.820839226779821,
      "grad_norm": 0.003424615366384387,
      "learning_rate": 0.1402404727594389,
      "loss": 0.2129,
      "num_input_tokens_seen": 18036416,
      "step": 20830
    },
    {
      "epoch": 9.823196605374823,
      "grad_norm": 0.002035647165030241,
      "learning_rate": 0.1401816934632737,
      "loss": 0.2355,
      "num_input_tokens_seen": 18041248,
      "step": 20835
    },
    {
      "epoch": 9.825553983969826,
      "grad_norm": 0.002866423688828945,
      "learning_rate": 0.1401229156812147,
      "loss": 0.2393,
      "num_input_tokens_seen": 18045680,
      "step": 20840
    },
    {
      "epoch": 9.827911362564828,
      "grad_norm": 0.0016310347709804773,
      "learning_rate": 0.14006413942232626,
      "loss": 0.2337,
      "num_input_tokens_seen": 18049680,
      "step": 20845
    },
    {
      "epoch": 9.83026874115983,
      "grad_norm": 0.002456455025821924,
      "learning_rate": 0.14000536469567235,
      "loss": 0.2399,
      "num_input_tokens_seen": 18054688,
      "step": 20850
    },
    {
      "epoch": 9.832626119754833,
      "grad_norm": 0.0029478874057531357,
      "learning_rate": 0.13994659151031685,
      "loss": 0.242,
      "num_input_tokens_seen": 18059600,
      "step": 20855
    },
    {
      "epoch": 9.834983498349835,
      "grad_norm": 0.001765622990205884,
      "learning_rate": 0.13988781987532323,
      "loss": 0.2043,
      "num_input_tokens_seen": 18064272,
      "step": 20860
    },
    {
      "epoch": 9.837340876944838,
      "grad_norm": 0.0019855352584272623,
      "learning_rate": 0.1398290497997549,
      "loss": 0.2481,
      "num_input_tokens_seen": 18068448,
      "step": 20865
    },
    {
      "epoch": 9.83969825553984,
      "grad_norm": 0.0021878390107303858,
      "learning_rate": 0.13977028129267488,
      "loss": 0.2512,
      "num_input_tokens_seen": 18072768,
      "step": 20870
    },
    {
      "epoch": 9.842055634134843,
      "grad_norm": 0.0013581026578322053,
      "learning_rate": 0.13971151436314605,
      "loss": 0.2349,
      "num_input_tokens_seen": 18077264,
      "step": 20875
    },
    {
      "epoch": 9.844413012729845,
      "grad_norm": 0.0028387701604515314,
      "learning_rate": 0.13965274902023103,
      "loss": 0.2098,
      "num_input_tokens_seen": 18082560,
      "step": 20880
    },
    {
      "epoch": 9.846770391324847,
      "grad_norm": 0.0010935210157185793,
      "learning_rate": 0.13959398527299208,
      "loss": 0.2224,
      "num_input_tokens_seen": 18086432,
      "step": 20885
    },
    {
      "epoch": 9.84912776991985,
      "grad_norm": 0.00331748160533607,
      "learning_rate": 0.13953522313049138,
      "loss": 0.2837,
      "num_input_tokens_seen": 18090832,
      "step": 20890
    },
    {
      "epoch": 9.851485148514852,
      "grad_norm": 0.002927246503531933,
      "learning_rate": 0.13947646260179083,
      "loss": 0.2372,
      "num_input_tokens_seen": 18094240,
      "step": 20895
    },
    {
      "epoch": 9.853842527109855,
      "grad_norm": 0.002717055846005678,
      "learning_rate": 0.13941770369595194,
      "loss": 0.217,
      "num_input_tokens_seen": 18099104,
      "step": 20900
    },
    {
      "epoch": 9.856199905704855,
      "grad_norm": 0.0019298747647553682,
      "learning_rate": 0.1393589464220362,
      "loss": 0.2513,
      "num_input_tokens_seen": 18103296,
      "step": 20905
    },
    {
      "epoch": 9.85855728429986,
      "grad_norm": 0.0016360790468752384,
      "learning_rate": 0.13930019078910455,
      "loss": 0.2239,
      "num_input_tokens_seen": 18107856,
      "step": 20910
    },
    {
      "epoch": 9.86091466289486,
      "grad_norm": 0.0031283670105040073,
      "learning_rate": 0.139241436806218,
      "loss": 0.2309,
      "num_input_tokens_seen": 18112064,
      "step": 20915
    },
    {
      "epoch": 9.863272041489862,
      "grad_norm": 0.0018226697575300932,
      "learning_rate": 0.13918268448243712,
      "loss": 0.1835,
      "num_input_tokens_seen": 18116464,
      "step": 20920
    },
    {
      "epoch": 9.865629420084865,
      "grad_norm": 0.008400446735322475,
      "learning_rate": 0.13912393382682217,
      "loss": 0.2324,
      "num_input_tokens_seen": 18120896,
      "step": 20925
    },
    {
      "epoch": 9.867986798679867,
      "grad_norm": 0.0018560690805315971,
      "learning_rate": 0.1390651848484333,
      "loss": 0.2568,
      "num_input_tokens_seen": 18125712,
      "step": 20930
    },
    {
      "epoch": 9.87034417727487,
      "grad_norm": 0.004047660622745752,
      "learning_rate": 0.1390064375563304,
      "loss": 0.2559,
      "num_input_tokens_seen": 18130592,
      "step": 20935
    },
    {
      "epoch": 9.872701555869872,
      "grad_norm": 0.0024155036080628633,
      "learning_rate": 0.13894769195957293,
      "loss": 0.2097,
      "num_input_tokens_seen": 18135248,
      "step": 20940
    },
    {
      "epoch": 9.875058934464874,
      "grad_norm": 0.004974984098225832,
      "learning_rate": 0.13888894806722032,
      "loss": 0.2157,
      "num_input_tokens_seen": 18139216,
      "step": 20945
    },
    {
      "epoch": 9.877416313059877,
      "grad_norm": 0.001171640818938613,
      "learning_rate": 0.1388302058883315,
      "loss": 0.237,
      "num_input_tokens_seen": 18144016,
      "step": 20950
    },
    {
      "epoch": 9.87977369165488,
      "grad_norm": 0.0011932163033634424,
      "learning_rate": 0.13877146543196528,
      "loss": 0.2176,
      "num_input_tokens_seen": 18147984,
      "step": 20955
    },
    {
      "epoch": 9.882131070249882,
      "grad_norm": 0.0011462057009339333,
      "learning_rate": 0.13871272670718027,
      "loss": 0.1857,
      "num_input_tokens_seen": 18152080,
      "step": 20960
    },
    {
      "epoch": 9.884488448844884,
      "grad_norm": 0.00197031581774354,
      "learning_rate": 0.13865398972303455,
      "loss": 0.2425,
      "num_input_tokens_seen": 18156128,
      "step": 20965
    },
    {
      "epoch": 9.886845827439886,
      "grad_norm": 0.0017002859385684133,
      "learning_rate": 0.13859525448858623,
      "loss": 0.2347,
      "num_input_tokens_seen": 18160224,
      "step": 20970
    },
    {
      "epoch": 9.889203206034889,
      "grad_norm": 0.001827147090807557,
      "learning_rate": 0.13853652101289304,
      "loss": 0.1967,
      "num_input_tokens_seen": 18164128,
      "step": 20975
    },
    {
      "epoch": 9.891560584629891,
      "grad_norm": 0.0016634489875286818,
      "learning_rate": 0.13847778930501234,
      "loss": 0.1633,
      "num_input_tokens_seen": 18168800,
      "step": 20980
    },
    {
      "epoch": 9.893917963224894,
      "grad_norm": 0.0012185194063931704,
      "learning_rate": 0.1384190593740013,
      "loss": 0.2056,
      "num_input_tokens_seen": 18173200,
      "step": 20985
    },
    {
      "epoch": 9.896275341819896,
      "grad_norm": 0.00219575478695333,
      "learning_rate": 0.13836033122891686,
      "loss": 0.264,
      "num_input_tokens_seen": 18177344,
      "step": 20990
    },
    {
      "epoch": 9.898632720414899,
      "grad_norm": 0.00421916414052248,
      "learning_rate": 0.1383016048788156,
      "loss": 0.2442,
      "num_input_tokens_seen": 18181968,
      "step": 20995
    },
    {
      "epoch": 9.900990099009901,
      "grad_norm": 0.0024581674952059984,
      "learning_rate": 0.13824288033275392,
      "loss": 0.2083,
      "num_input_tokens_seen": 18186480,
      "step": 21000
    },
    {
      "epoch": 9.900990099009901,
      "eval_loss": 0.22014720737934113,
      "eval_runtime": 21.9322,
      "eval_samples_per_second": 42.996,
      "eval_steps_per_second": 21.521,
      "num_input_tokens_seen": 18186480,
      "step": 21000
    },
    {
      "epoch": 9.903347477604903,
      "grad_norm": 0.0018667903495952487,
      "learning_rate": 0.1381841575997878,
      "loss": 0.1631,
      "num_input_tokens_seen": 18189808,
      "step": 21005
    },
    {
      "epoch": 9.905704856199906,
      "grad_norm": 0.0020262333564460278,
      "learning_rate": 0.13812543668897306,
      "loss": 0.2335,
      "num_input_tokens_seen": 18195152,
      "step": 21010
    },
    {
      "epoch": 9.908062234794908,
      "grad_norm": 0.0013739721616730094,
      "learning_rate": 0.13806671760936526,
      "loss": 0.2201,
      "num_input_tokens_seen": 18199424,
      "step": 21015
    },
    {
      "epoch": 9.91041961338991,
      "grad_norm": 0.0012936403509229422,
      "learning_rate": 0.13800800037001956,
      "loss": 0.1722,
      "num_input_tokens_seen": 18204592,
      "step": 21020
    },
    {
      "epoch": 9.912776991984913,
      "grad_norm": 0.002181220566853881,
      "learning_rate": 0.13794928497999087,
      "loss": 0.225,
      "num_input_tokens_seen": 18209456,
      "step": 21025
    },
    {
      "epoch": 9.915134370579915,
      "grad_norm": 0.0009585630614310503,
      "learning_rate": 0.1378905714483339,
      "loss": 0.2322,
      "num_input_tokens_seen": 18213952,
      "step": 21030
    },
    {
      "epoch": 9.917491749174918,
      "grad_norm": 0.0009372510248795152,
      "learning_rate": 0.13783185978410295,
      "loss": 0.2135,
      "num_input_tokens_seen": 18217600,
      "step": 21035
    },
    {
      "epoch": 9.91984912776992,
      "grad_norm": 0.00174662412609905,
      "learning_rate": 0.13777314999635218,
      "loss": 0.206,
      "num_input_tokens_seen": 18221904,
      "step": 21040
    },
    {
      "epoch": 9.922206506364923,
      "grad_norm": 0.0018752949545159936,
      "learning_rate": 0.1377144420941353,
      "loss": 0.2124,
      "num_input_tokens_seen": 18226464,
      "step": 21045
    },
    {
      "epoch": 9.924563884959925,
      "grad_norm": 0.0009554074495099485,
      "learning_rate": 0.13765573608650586,
      "loss": 0.2166,
      "num_input_tokens_seen": 18230192,
      "step": 21050
    },
    {
      "epoch": 9.926921263554927,
      "grad_norm": 0.001096214517019689,
      "learning_rate": 0.13759703198251702,
      "loss": 0.2553,
      "num_input_tokens_seen": 18234624,
      "step": 21055
    },
    {
      "epoch": 9.92927864214993,
      "grad_norm": 0.0020234626717865467,
      "learning_rate": 0.13753832979122174,
      "loss": 0.2229,
      "num_input_tokens_seen": 18238992,
      "step": 21060
    },
    {
      "epoch": 9.931636020744932,
      "grad_norm": 0.0009081608150154352,
      "learning_rate": 0.13747962952167264,
      "loss": 0.2205,
      "num_input_tokens_seen": 18242864,
      "step": 21065
    },
    {
      "epoch": 9.933993399339935,
      "grad_norm": 0.0016321418806910515,
      "learning_rate": 0.13742093118292192,
      "loss": 0.2402,
      "num_input_tokens_seen": 18247504,
      "step": 21070
    },
    {
      "epoch": 9.936350777934937,
      "grad_norm": 0.001026210025884211,
      "learning_rate": 0.13736223478402174,
      "loss": 0.2217,
      "num_input_tokens_seen": 18251024,
      "step": 21075
    },
    {
      "epoch": 9.93870815652994,
      "grad_norm": 0.002431320259347558,
      "learning_rate": 0.1373035403340238,
      "loss": 0.2029,
      "num_input_tokens_seen": 18256160,
      "step": 21080
    },
    {
      "epoch": 9.941065535124942,
      "grad_norm": 0.0019542360678315163,
      "learning_rate": 0.13724484784197943,
      "loss": 0.2308,
      "num_input_tokens_seen": 18260336,
      "step": 21085
    },
    {
      "epoch": 9.943422913719944,
      "grad_norm": 0.002780736656859517,
      "learning_rate": 0.13718615731693987,
      "loss": 0.2161,
      "num_input_tokens_seen": 18265536,
      "step": 21090
    },
    {
      "epoch": 9.945780292314947,
      "grad_norm": 0.0017699002055451274,
      "learning_rate": 0.13712746876795587,
      "loss": 0.219,
      "num_input_tokens_seen": 18269280,
      "step": 21095
    },
    {
      "epoch": 9.948137670909949,
      "grad_norm": 0.0017739393515512347,
      "learning_rate": 0.13706878220407792,
      "loss": 0.2349,
      "num_input_tokens_seen": 18273552,
      "step": 21100
    },
    {
      "epoch": 9.950495049504951,
      "grad_norm": 0.0027136639691889286,
      "learning_rate": 0.13701009763435631,
      "loss": 0.2109,
      "num_input_tokens_seen": 18277632,
      "step": 21105
    },
    {
      "epoch": 9.952852428099952,
      "grad_norm": 0.002867842325940728,
      "learning_rate": 0.13695141506784084,
      "loss": 0.2384,
      "num_input_tokens_seen": 18282576,
      "step": 21110
    },
    {
      "epoch": 9.955209806694956,
      "grad_norm": 0.0011049900203943253,
      "learning_rate": 0.13689273451358114,
      "loss": 0.2365,
      "num_input_tokens_seen": 18286064,
      "step": 21115
    },
    {
      "epoch": 9.957567185289957,
      "grad_norm": 0.002707589417695999,
      "learning_rate": 0.13683405598062653,
      "loss": 0.236,
      "num_input_tokens_seen": 18291392,
      "step": 21120
    },
    {
      "epoch": 9.95992456388496,
      "grad_norm": 0.0031974036246538162,
      "learning_rate": 0.1367753794780259,
      "loss": 0.2399,
      "num_input_tokens_seen": 18295888,
      "step": 21125
    },
    {
      "epoch": 9.962281942479962,
      "grad_norm": 0.0021392125636339188,
      "learning_rate": 0.13671670501482802,
      "loss": 0.2267,
      "num_input_tokens_seen": 18299856,
      "step": 21130
    },
    {
      "epoch": 9.964639321074964,
      "grad_norm": 0.0012960751773789525,
      "learning_rate": 0.1366580326000811,
      "loss": 0.2074,
      "num_input_tokens_seen": 18303584,
      "step": 21135
    },
    {
      "epoch": 9.966996699669966,
      "grad_norm": 0.0014119481202214956,
      "learning_rate": 0.1365993622428332,
      "loss": 0.1963,
      "num_input_tokens_seen": 18307792,
      "step": 21140
    },
    {
      "epoch": 9.969354078264969,
      "grad_norm": 0.001165733439847827,
      "learning_rate": 0.13654069395213211,
      "loss": 0.2008,
      "num_input_tokens_seen": 18312416,
      "step": 21145
    },
    {
      "epoch": 9.971711456859971,
      "grad_norm": 0.001463819993659854,
      "learning_rate": 0.13648202773702509,
      "loss": 0.288,
      "num_input_tokens_seen": 18316560,
      "step": 21150
    },
    {
      "epoch": 9.974068835454974,
      "grad_norm": 0.0026052154134958982,
      "learning_rate": 0.13642336360655927,
      "loss": 0.2013,
      "num_input_tokens_seen": 18321568,
      "step": 21155
    },
    {
      "epoch": 9.976426214049976,
      "grad_norm": 0.0010159129742532969,
      "learning_rate": 0.13636470156978145,
      "loss": 0.2346,
      "num_input_tokens_seen": 18326016,
      "step": 21160
    },
    {
      "epoch": 9.978783592644978,
      "grad_norm": 0.001300600590184331,
      "learning_rate": 0.13630604163573798,
      "loss": 0.1881,
      "num_input_tokens_seen": 18330560,
      "step": 21165
    },
    {
      "epoch": 9.98114097123998,
      "grad_norm": 0.001627613208256662,
      "learning_rate": 0.13624738381347495,
      "loss": 0.2026,
      "num_input_tokens_seen": 18335040,
      "step": 21170
    },
    {
      "epoch": 9.983498349834983,
      "grad_norm": 0.0023862365633249283,
      "learning_rate": 0.1361887281120382,
      "loss": 0.2305,
      "num_input_tokens_seen": 18339024,
      "step": 21175
    },
    {
      "epoch": 9.985855728429986,
      "grad_norm": 0.0027408418245613575,
      "learning_rate": 0.13613007454047307,
      "loss": 0.221,
      "num_input_tokens_seen": 18343520,
      "step": 21180
    },
    {
      "epoch": 9.988213107024988,
      "grad_norm": 0.001805687672458589,
      "learning_rate": 0.13607142310782486,
      "loss": 0.2216,
      "num_input_tokens_seen": 18348304,
      "step": 21185
    },
    {
      "epoch": 9.99057048561999,
      "grad_norm": 0.0013148213038221002,
      "learning_rate": 0.13601277382313814,
      "loss": 0.2183,
      "num_input_tokens_seen": 18352272,
      "step": 21190
    },
    {
      "epoch": 9.992927864214993,
      "grad_norm": 0.001472087693400681,
      "learning_rate": 0.1359541266954575,
      "loss": 0.212,
      "num_input_tokens_seen": 18356640,
      "step": 21195
    },
    {
      "epoch": 9.995285242809995,
      "grad_norm": 0.0012460732832551003,
      "learning_rate": 0.13589548173382707,
      "loss": 0.2329,
      "num_input_tokens_seen": 18360368,
      "step": 21200
    },
    {
      "epoch": 9.995285242809995,
      "eval_loss": 0.21801726520061493,
      "eval_runtime": 21.9151,
      "eval_samples_per_second": 43.03,
      "eval_steps_per_second": 21.538,
      "num_input_tokens_seen": 18360368,
      "step": 21200
    },
    {
      "epoch": 9.997642621404998,
      "grad_norm": 0.001578094088472426,
      "learning_rate": 0.1358368389472906,
      "loss": 0.2541,
      "num_input_tokens_seen": 18364368,
      "step": 21205
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.0013486967654898763,
      "learning_rate": 0.13577819834489155,
      "loss": 0.2144,
      "num_input_tokens_seen": 18368192,
      "step": 21210
    },
    {
      "epoch": 10.002357378595002,
      "grad_norm": 0.002214340493083,
      "learning_rate": 0.135719559935673,
      "loss": 0.201,
      "num_input_tokens_seen": 18372992,
      "step": 21215
    },
    {
      "epoch": 10.004714757190005,
      "grad_norm": 0.0021097101271152496,
      "learning_rate": 0.13566092372867775,
      "loss": 0.2184,
      "num_input_tokens_seen": 18377920,
      "step": 21220
    },
    {
      "epoch": 10.007072135785007,
      "grad_norm": 0.002251502126455307,
      "learning_rate": 0.13560228973294833,
      "loss": 0.2173,
      "num_input_tokens_seen": 18383072,
      "step": 21225
    },
    {
      "epoch": 10.00942951438001,
      "grad_norm": 0.001967727905139327,
      "learning_rate": 0.13554365795752668,
      "loss": 0.2283,
      "num_input_tokens_seen": 18387824,
      "step": 21230
    },
    {
      "epoch": 10.011786892975012,
      "grad_norm": 0.0018676373874768615,
      "learning_rate": 0.1354850284114547,
      "loss": 0.2291,
      "num_input_tokens_seen": 18391760,
      "step": 21235
    },
    {
      "epoch": 10.014144271570014,
      "grad_norm": 0.0013665041187778115,
      "learning_rate": 0.13542640110377374,
      "loss": 0.2586,
      "num_input_tokens_seen": 18396720,
      "step": 21240
    },
    {
      "epoch": 10.016501650165017,
      "grad_norm": 0.0016202329425141215,
      "learning_rate": 0.13536777604352487,
      "loss": 0.2196,
      "num_input_tokens_seen": 18400704,
      "step": 21245
    },
    {
      "epoch": 10.01885902876002,
      "grad_norm": 0.0039044334553182125,
      "learning_rate": 0.13530915323974887,
      "loss": 0.2246,
      "num_input_tokens_seen": 18404624,
      "step": 21250
    },
    {
      "epoch": 10.021216407355022,
      "grad_norm": 0.001436038757674396,
      "learning_rate": 0.13525053270148596,
      "loss": 0.2223,
      "num_input_tokens_seen": 18409104,
      "step": 21255
    },
    {
      "epoch": 10.023573785950024,
      "grad_norm": 0.0016989236464723945,
      "learning_rate": 0.13519191443777628,
      "loss": 0.2223,
      "num_input_tokens_seen": 18413168,
      "step": 21260
    },
    {
      "epoch": 10.025931164545026,
      "grad_norm": 0.0012037827400490642,
      "learning_rate": 0.13513329845765953,
      "loss": 0.2197,
      "num_input_tokens_seen": 18417456,
      "step": 21265
    },
    {
      "epoch": 10.028288543140029,
      "grad_norm": 0.001256100251339376,
      "learning_rate": 0.13507468477017495,
      "loss": 0.2262,
      "num_input_tokens_seen": 18422256,
      "step": 21270
    },
    {
      "epoch": 10.030645921735031,
      "grad_norm": 0.0017386305844411254,
      "learning_rate": 0.13501607338436153,
      "loss": 0.1869,
      "num_input_tokens_seen": 18428192,
      "step": 21275
    },
    {
      "epoch": 10.033003300330034,
      "grad_norm": 0.001125015551224351,
      "learning_rate": 0.13495746430925798,
      "loss": 0.2079,
      "num_input_tokens_seen": 18432384,
      "step": 21280
    },
    {
      "epoch": 10.035360678925036,
      "grad_norm": 0.0031460796017199755,
      "learning_rate": 0.13489885755390238,
      "loss": 0.2575,
      "num_input_tokens_seen": 18436784,
      "step": 21285
    },
    {
      "epoch": 10.037718057520038,
      "grad_norm": 0.0011611274676397443,
      "learning_rate": 0.13484025312733275,
      "loss": 0.2817,
      "num_input_tokens_seen": 18441312,
      "step": 21290
    },
    {
      "epoch": 10.04007543611504,
      "grad_norm": 0.0016808726359158754,
      "learning_rate": 0.13478165103858658,
      "loss": 0.1929,
      "num_input_tokens_seen": 18445728,
      "step": 21295
    },
    {
      "epoch": 10.042432814710043,
      "grad_norm": 0.001612075138837099,
      "learning_rate": 0.13472305129670106,
      "loss": 0.1939,
      "num_input_tokens_seen": 18450096,
      "step": 21300
    },
    {
      "epoch": 10.044790193305046,
      "grad_norm": 0.002591305412352085,
      "learning_rate": 0.13466445391071305,
      "loss": 0.2557,
      "num_input_tokens_seen": 18454560,
      "step": 21305
    },
    {
      "epoch": 10.047147571900048,
      "grad_norm": 0.001299537718296051,
      "learning_rate": 0.13460585888965895,
      "loss": 0.2096,
      "num_input_tokens_seen": 18458656,
      "step": 21310
    },
    {
      "epoch": 10.049504950495049,
      "grad_norm": 0.0017594894161447883,
      "learning_rate": 0.13454726624257482,
      "loss": 0.254,
      "num_input_tokens_seen": 18463072,
      "step": 21315
    },
    {
      "epoch": 10.051862329090051,
      "grad_norm": 0.0026462057139724493,
      "learning_rate": 0.1344886759784965,
      "loss": 0.2314,
      "num_input_tokens_seen": 18467424,
      "step": 21320
    },
    {
      "epoch": 10.054219707685053,
      "grad_norm": 0.0018532495014369488,
      "learning_rate": 0.13443008810645923,
      "loss": 0.2125,
      "num_input_tokens_seen": 18471952,
      "step": 21325
    },
    {
      "epoch": 10.056577086280056,
      "grad_norm": 0.0010463576763868332,
      "learning_rate": 0.13437150263549807,
      "loss": 0.2408,
      "num_input_tokens_seen": 18476416,
      "step": 21330
    },
    {
      "epoch": 10.058934464875058,
      "grad_norm": 0.0013170493766665459,
      "learning_rate": 0.13431291957464755,
      "loss": 0.224,
      "num_input_tokens_seen": 18480768,
      "step": 21335
    },
    {
      "epoch": 10.06129184347006,
      "grad_norm": 0.0013173696352168918,
      "learning_rate": 0.13425433893294197,
      "loss": 0.215,
      "num_input_tokens_seen": 18485472,
      "step": 21340
    },
    {
      "epoch": 10.063649222065063,
      "grad_norm": 0.0019737286493182182,
      "learning_rate": 0.13419576071941525,
      "loss": 0.2372,
      "num_input_tokens_seen": 18490336,
      "step": 21345
    },
    {
      "epoch": 10.066006600660065,
      "grad_norm": 0.0014504195423796773,
      "learning_rate": 0.1341371849431008,
      "loss": 0.1848,
      "num_input_tokens_seen": 18494864,
      "step": 21350
    },
    {
      "epoch": 10.068363979255068,
      "grad_norm": 0.0016785485204309225,
      "learning_rate": 0.13407861161303178,
      "loss": 0.25,
      "num_input_tokens_seen": 18498960,
      "step": 21355
    },
    {
      "epoch": 10.07072135785007,
      "grad_norm": 0.001222933758981526,
      "learning_rate": 0.13402004073824098,
      "loss": 0.2358,
      "num_input_tokens_seen": 18503136,
      "step": 21360
    },
    {
      "epoch": 10.073078736445073,
      "grad_norm": 0.0011035986244678497,
      "learning_rate": 0.13396147232776062,
      "loss": 0.223,
      "num_input_tokens_seen": 18507808,
      "step": 21365
    },
    {
      "epoch": 10.075436115040075,
      "grad_norm": 0.0011290839174762368,
      "learning_rate": 0.13390290639062288,
      "loss": 0.2029,
      "num_input_tokens_seen": 18512496,
      "step": 21370
    },
    {
      "epoch": 10.077793493635077,
      "grad_norm": 0.0017865909030660987,
      "learning_rate": 0.13384434293585917,
      "loss": 0.2003,
      "num_input_tokens_seen": 18516976,
      "step": 21375
    },
    {
      "epoch": 10.08015087223008,
      "grad_norm": 0.0018875985406339169,
      "learning_rate": 0.13378578197250088,
      "loss": 0.2374,
      "num_input_tokens_seen": 18521216,
      "step": 21380
    },
    {
      "epoch": 10.082508250825082,
      "grad_norm": 0.0018964592600241303,
      "learning_rate": 0.13372722350957872,
      "loss": 0.2264,
      "num_input_tokens_seen": 18526768,
      "step": 21385
    },
    {
      "epoch": 10.084865629420085,
      "grad_norm": 0.0014962619170546532,
      "learning_rate": 0.13366866755612322,
      "loss": 0.2432,
      "num_input_tokens_seen": 18530624,
      "step": 21390
    },
    {
      "epoch": 10.087223008015087,
      "grad_norm": 0.0019652624614536762,
      "learning_rate": 0.13361011412116436,
      "loss": 0.229,
      "num_input_tokens_seen": 18535696,
      "step": 21395
    },
    {
      "epoch": 10.08958038661009,
      "grad_norm": 0.0009250770090147853,
      "learning_rate": 0.13355156321373196,
      "loss": 0.2151,
      "num_input_tokens_seen": 18539664,
      "step": 21400
    },
    {
      "epoch": 10.08958038661009,
      "eval_loss": 0.21888743340969086,
      "eval_runtime": 21.9452,
      "eval_samples_per_second": 42.971,
      "eval_steps_per_second": 21.508,
      "num_input_tokens_seen": 18539664,
      "step": 21400
    },
    {
      "epoch": 10.091937765205092,
      "grad_norm": 0.002055337419733405,
      "learning_rate": 0.13349301484285514,
      "loss": 0.2287,
      "num_input_tokens_seen": 18544992,
      "step": 21405
    },
    {
      "epoch": 10.094295143800094,
      "grad_norm": 0.0011586814653128386,
      "learning_rate": 0.13343446901756295,
      "loss": 0.2051,
      "num_input_tokens_seen": 18549200,
      "step": 21410
    },
    {
      "epoch": 10.096652522395097,
      "grad_norm": 0.0016689538024365902,
      "learning_rate": 0.13337592574688376,
      "loss": 0.2206,
      "num_input_tokens_seen": 18553632,
      "step": 21415
    },
    {
      "epoch": 10.099009900990099,
      "grad_norm": 0.0018155189463868737,
      "learning_rate": 0.13331738503984572,
      "loss": 0.2285,
      "num_input_tokens_seen": 18557632,
      "step": 21420
    },
    {
      "epoch": 10.101367279585101,
      "grad_norm": 0.001697544939815998,
      "learning_rate": 0.1332588469054766,
      "loss": 0.2359,
      "num_input_tokens_seen": 18562704,
      "step": 21425
    },
    {
      "epoch": 10.103724658180104,
      "grad_norm": 0.0009774485370144248,
      "learning_rate": 0.1332003113528036,
      "loss": 0.2105,
      "num_input_tokens_seen": 18567216,
      "step": 21430
    },
    {
      "epoch": 10.106082036775106,
      "grad_norm": 0.001356321619823575,
      "learning_rate": 0.13314177839085373,
      "loss": 0.2309,
      "num_input_tokens_seen": 18571856,
      "step": 21435
    },
    {
      "epoch": 10.108439415370109,
      "grad_norm": 0.0016018390888348222,
      "learning_rate": 0.13308324802865354,
      "loss": 0.217,
      "num_input_tokens_seen": 18575936,
      "step": 21440
    },
    {
      "epoch": 10.110796793965111,
      "grad_norm": 0.0014719140017405152,
      "learning_rate": 0.13302472027522905,
      "loss": 0.2335,
      "num_input_tokens_seen": 18580064,
      "step": 21445
    },
    {
      "epoch": 10.113154172560114,
      "grad_norm": 0.0016623333794996142,
      "learning_rate": 0.13296619513960606,
      "loss": 0.2308,
      "num_input_tokens_seen": 18584336,
      "step": 21450
    },
    {
      "epoch": 10.115511551155116,
      "grad_norm": 0.0014535210793837905,
      "learning_rate": 0.1329076726308098,
      "loss": 0.2247,
      "num_input_tokens_seen": 18588656,
      "step": 21455
    },
    {
      "epoch": 10.117868929750118,
      "grad_norm": 0.0013843660708516836,
      "learning_rate": 0.13284915275786519,
      "loss": 0.2248,
      "num_input_tokens_seen": 18592096,
      "step": 21460
    },
    {
      "epoch": 10.12022630834512,
      "grad_norm": 0.0017334980657324195,
      "learning_rate": 0.1327906355297968,
      "loss": 0.2198,
      "num_input_tokens_seen": 18596944,
      "step": 21465
    },
    {
      "epoch": 10.122583686940123,
      "grad_norm": 0.0021600709296762943,
      "learning_rate": 0.13273212095562867,
      "loss": 0.2063,
      "num_input_tokens_seen": 18601856,
      "step": 21470
    },
    {
      "epoch": 10.124941065535126,
      "grad_norm": 0.0016057930188253522,
      "learning_rate": 0.13267360904438444,
      "loss": 0.2525,
      "num_input_tokens_seen": 18607088,
      "step": 21475
    },
    {
      "epoch": 10.127298444130128,
      "grad_norm": 0.001730194315314293,
      "learning_rate": 0.1326150998050875,
      "loss": 0.2278,
      "num_input_tokens_seen": 18611888,
      "step": 21480
    },
    {
      "epoch": 10.12965582272513,
      "grad_norm": 0.0009589291294105351,
      "learning_rate": 0.1325565932467606,
      "loss": 0.1965,
      "num_input_tokens_seen": 18615904,
      "step": 21485
    },
    {
      "epoch": 10.132013201320133,
      "grad_norm": 0.001745629939250648,
      "learning_rate": 0.13249808937842628,
      "loss": 0.2192,
      "num_input_tokens_seen": 18620224,
      "step": 21490
    },
    {
      "epoch": 10.134370579915135,
      "grad_norm": 0.00217812554910779,
      "learning_rate": 0.1324395882091065,
      "loss": 0.249,
      "num_input_tokens_seen": 18624160,
      "step": 21495
    },
    {
      "epoch": 10.136727958510138,
      "grad_norm": 0.0013365367194637656,
      "learning_rate": 0.13238108974782284,
      "loss": 0.2014,
      "num_input_tokens_seen": 18628496,
      "step": 21500
    },
    {
      "epoch": 10.13908533710514,
      "grad_norm": 0.0018434942467138171,
      "learning_rate": 0.13232259400359664,
      "loss": 0.2006,
      "num_input_tokens_seen": 18632304,
      "step": 21505
    },
    {
      "epoch": 10.14144271570014,
      "grad_norm": 0.0016380795277655125,
      "learning_rate": 0.13226410098544852,
      "loss": 0.1927,
      "num_input_tokens_seen": 18637008,
      "step": 21510
    },
    {
      "epoch": 10.143800094295143,
      "grad_norm": 0.0020819443743675947,
      "learning_rate": 0.13220561070239892,
      "loss": 0.2414,
      "num_input_tokens_seen": 18641392,
      "step": 21515
    },
    {
      "epoch": 10.146157472890145,
      "grad_norm": 0.0030648489482700825,
      "learning_rate": 0.13214712316346783,
      "loss": 0.184,
      "num_input_tokens_seen": 18646368,
      "step": 21520
    },
    {
      "epoch": 10.148514851485148,
      "grad_norm": 0.001453018863685429,
      "learning_rate": 0.13208863837767465,
      "loss": 0.2525,
      "num_input_tokens_seen": 18650240,
      "step": 21525
    },
    {
      "epoch": 10.15087223008015,
      "grad_norm": 0.0017490035388618708,
      "learning_rate": 0.13203015635403856,
      "loss": 0.2153,
      "num_input_tokens_seen": 18655136,
      "step": 21530
    },
    {
      "epoch": 10.153229608675153,
      "grad_norm": 0.001426578382961452,
      "learning_rate": 0.13197167710157817,
      "loss": 0.1791,
      "num_input_tokens_seen": 18659632,
      "step": 21535
    },
    {
      "epoch": 10.155586987270155,
      "grad_norm": 0.0015461939619854093,
      "learning_rate": 0.13191320062931167,
      "loss": 0.2193,
      "num_input_tokens_seen": 18664080,
      "step": 21540
    },
    {
      "epoch": 10.157944365865157,
      "grad_norm": 0.0012613626895472407,
      "learning_rate": 0.13185472694625702,
      "loss": 0.1665,
      "num_input_tokens_seen": 18668272,
      "step": 21545
    },
    {
      "epoch": 10.16030174446016,
      "grad_norm": 0.0011577388504520059,
      "learning_rate": 0.13179625606143142,
      "loss": 0.2317,
      "num_input_tokens_seen": 18672352,
      "step": 21550
    },
    {
      "epoch": 10.162659123055162,
      "grad_norm": 0.002230259357020259,
      "learning_rate": 0.13173778798385188,
      "loss": 0.3123,
      "num_input_tokens_seen": 18678400,
      "step": 21555
    },
    {
      "epoch": 10.165016501650165,
      "grad_norm": 0.0012496886774897575,
      "learning_rate": 0.13167932272253505,
      "loss": 0.2203,
      "num_input_tokens_seen": 18682608,
      "step": 21560
    },
    {
      "epoch": 10.167373880245167,
      "grad_norm": 0.0010703924344852567,
      "learning_rate": 0.1316208602864968,
      "loss": 0.252,
      "num_input_tokens_seen": 18686832,
      "step": 21565
    },
    {
      "epoch": 10.16973125884017,
      "grad_norm": 0.0014066774165257812,
      "learning_rate": 0.13156240068475292,
      "loss": 0.2309,
      "num_input_tokens_seen": 18691808,
      "step": 21570
    },
    {
      "epoch": 10.172088637435172,
      "grad_norm": 0.0010762856109067798,
      "learning_rate": 0.1315039439263185,
      "loss": 0.2189,
      "num_input_tokens_seen": 18697088,
      "step": 21575
    },
    {
      "epoch": 10.174446016030174,
      "grad_norm": 0.0017218475695699453,
      "learning_rate": 0.13144549002020833,
      "loss": 0.2181,
      "num_input_tokens_seen": 18701440,
      "step": 21580
    },
    {
      "epoch": 10.176803394625177,
      "grad_norm": 0.0010263760341331363,
      "learning_rate": 0.13138703897543688,
      "loss": 0.2491,
      "num_input_tokens_seen": 18705472,
      "step": 21585
    },
    {
      "epoch": 10.179160773220179,
      "grad_norm": 0.0015922652091830969,
      "learning_rate": 0.1313285908010178,
      "loss": 0.1831,
      "num_input_tokens_seen": 18709616,
      "step": 21590
    },
    {
      "epoch": 10.181518151815181,
      "grad_norm": 0.0009499292937107384,
      "learning_rate": 0.13127014550596475,
      "loss": 0.1716,
      "num_input_tokens_seen": 18714176,
      "step": 21595
    },
    {
      "epoch": 10.183875530410184,
      "grad_norm": 0.0011007396969944239,
      "learning_rate": 0.1312117030992906,
      "loss": 0.2113,
      "num_input_tokens_seen": 18718016,
      "step": 21600
    },
    {
      "epoch": 10.183875530410184,
      "eval_loss": 0.22318710386753082,
      "eval_runtime": 21.9081,
      "eval_samples_per_second": 43.043,
      "eval_steps_per_second": 21.544,
      "num_input_tokens_seen": 18718016,
      "step": 21600
    },
    {
      "epoch": 10.186232909005186,
      "grad_norm": 0.0015942996833473444,
      "learning_rate": 0.13115326359000795,
      "loss": 0.2556,
      "num_input_tokens_seen": 18723008,
      "step": 21605
    },
    {
      "epoch": 10.188590287600189,
      "grad_norm": 0.0012011429062113166,
      "learning_rate": 0.13109482698712896,
      "loss": 0.2379,
      "num_input_tokens_seen": 18727184,
      "step": 21610
    },
    {
      "epoch": 10.190947666195191,
      "grad_norm": 0.0008231604588218033,
      "learning_rate": 0.1310363932996651,
      "loss": 0.2403,
      "num_input_tokens_seen": 18731072,
      "step": 21615
    },
    {
      "epoch": 10.193305044790193,
      "grad_norm": 0.0011203772155568004,
      "learning_rate": 0.13097796253662775,
      "loss": 0.2397,
      "num_input_tokens_seen": 18735424,
      "step": 21620
    },
    {
      "epoch": 10.195662423385196,
      "grad_norm": 0.002201799303293228,
      "learning_rate": 0.1309195347070277,
      "loss": 0.2135,
      "num_input_tokens_seen": 18739856,
      "step": 21625
    },
    {
      "epoch": 10.198019801980198,
      "grad_norm": 0.001779241138137877,
      "learning_rate": 0.13086110981987506,
      "loss": 0.2139,
      "num_input_tokens_seen": 18744720,
      "step": 21630
    },
    {
      "epoch": 10.2003771805752,
      "grad_norm": 0.0012009958736598492,
      "learning_rate": 0.13080268788417987,
      "loss": 0.2103,
      "num_input_tokens_seen": 18748656,
      "step": 21635
    },
    {
      "epoch": 10.202734559170203,
      "grad_norm": 0.001297451090067625,
      "learning_rate": 0.1307442689089515,
      "loss": 0.2371,
      "num_input_tokens_seen": 18751968,
      "step": 21640
    },
    {
      "epoch": 10.205091937765205,
      "grad_norm": 0.0014946481678634882,
      "learning_rate": 0.13068585290319873,
      "loss": 0.1917,
      "num_input_tokens_seen": 18756032,
      "step": 21645
    },
    {
      "epoch": 10.207449316360208,
      "grad_norm": 0.0010517353657633066,
      "learning_rate": 0.13062743987593026,
      "loss": 0.2133,
      "num_input_tokens_seen": 18759856,
      "step": 21650
    },
    {
      "epoch": 10.20980669495521,
      "grad_norm": 0.0008070350158959627,
      "learning_rate": 0.13056902983615395,
      "loss": 0.2321,
      "num_input_tokens_seen": 18764064,
      "step": 21655
    },
    {
      "epoch": 10.212164073550213,
      "grad_norm": 0.0022804324980825186,
      "learning_rate": 0.13051062279287742,
      "loss": 0.2576,
      "num_input_tokens_seen": 18768512,
      "step": 21660
    },
    {
      "epoch": 10.214521452145215,
      "grad_norm": 0.0014580650022253394,
      "learning_rate": 0.13045221875510782,
      "loss": 0.2102,
      "num_input_tokens_seen": 18772544,
      "step": 21665
    },
    {
      "epoch": 10.216878830740217,
      "grad_norm": 0.0029510511085391045,
      "learning_rate": 0.13039381773185174,
      "loss": 0.228,
      "num_input_tokens_seen": 18776656,
      "step": 21670
    },
    {
      "epoch": 10.21923620933522,
      "grad_norm": 0.0015310606686398387,
      "learning_rate": 0.1303354197321153,
      "loss": 0.2261,
      "num_input_tokens_seen": 18781200,
      "step": 21675
    },
    {
      "epoch": 10.221593587930222,
      "grad_norm": 0.0018566305516287684,
      "learning_rate": 0.13027702476490433,
      "loss": 0.2261,
      "num_input_tokens_seen": 18785024,
      "step": 21680
    },
    {
      "epoch": 10.223950966525225,
      "grad_norm": 0.001569773769006133,
      "learning_rate": 0.1302186328392239,
      "loss": 0.2163,
      "num_input_tokens_seen": 18789616,
      "step": 21685
    },
    {
      "epoch": 10.226308345120227,
      "grad_norm": 0.0009505623602308333,
      "learning_rate": 0.130160243964079,
      "loss": 0.2079,
      "num_input_tokens_seen": 18793872,
      "step": 21690
    },
    {
      "epoch": 10.22866572371523,
      "grad_norm": 0.0015276246704161167,
      "learning_rate": 0.13010185814847372,
      "loss": 0.2091,
      "num_input_tokens_seen": 18797920,
      "step": 21695
    },
    {
      "epoch": 10.231023102310232,
      "grad_norm": 0.001353686093352735,
      "learning_rate": 0.13004347540141192,
      "loss": 0.185,
      "num_input_tokens_seen": 18802784,
      "step": 21700
    },
    {
      "epoch": 10.233380480905234,
      "grad_norm": 0.0010096713667735457,
      "learning_rate": 0.12998509573189712,
      "loss": 0.2349,
      "num_input_tokens_seen": 18807200,
      "step": 21705
    },
    {
      "epoch": 10.235737859500237,
      "grad_norm": 0.0016245967708528042,
      "learning_rate": 0.12992671914893203,
      "loss": 0.208,
      "num_input_tokens_seen": 18812544,
      "step": 21710
    },
    {
      "epoch": 10.238095238095237,
      "grad_norm": 0.0017846112605184317,
      "learning_rate": 0.12986834566151909,
      "loss": 0.262,
      "num_input_tokens_seen": 18816752,
      "step": 21715
    },
    {
      "epoch": 10.24045261669024,
      "grad_norm": 0.002208509249612689,
      "learning_rate": 0.12980997527866028,
      "loss": 0.2166,
      "num_input_tokens_seen": 18821456,
      "step": 21720
    },
    {
      "epoch": 10.242809995285242,
      "grad_norm": 0.0019410925451666117,
      "learning_rate": 0.12975160800935692,
      "loss": 0.2495,
      "num_input_tokens_seen": 18826032,
      "step": 21725
    },
    {
      "epoch": 10.245167373880244,
      "grad_norm": 0.0011525312438607216,
      "learning_rate": 0.12969324386261016,
      "loss": 0.1902,
      "num_input_tokens_seen": 18830304,
      "step": 21730
    },
    {
      "epoch": 10.247524752475247,
      "grad_norm": 0.0012868059566244483,
      "learning_rate": 0.12963488284742034,
      "loss": 0.2204,
      "num_input_tokens_seen": 18834512,
      "step": 21735
    },
    {
      "epoch": 10.24988213107025,
      "grad_norm": 0.0017551748314872384,
      "learning_rate": 0.12957652497278752,
      "loss": 0.2513,
      "num_input_tokens_seen": 18838416,
      "step": 21740
    },
    {
      "epoch": 10.252239509665252,
      "grad_norm": 0.0009911122033372521,
      "learning_rate": 0.12951817024771117,
      "loss": 0.1872,
      "num_input_tokens_seen": 18842912,
      "step": 21745
    },
    {
      "epoch": 10.254596888260254,
      "grad_norm": 0.0016291238134726882,
      "learning_rate": 0.12945981868119041,
      "loss": 0.2241,
      "num_input_tokens_seen": 18846944,
      "step": 21750
    },
    {
      "epoch": 10.256954266855256,
      "grad_norm": 0.0016594991320744157,
      "learning_rate": 0.12940147028222376,
      "loss": 0.2299,
      "num_input_tokens_seen": 18851040,
      "step": 21755
    },
    {
      "epoch": 10.259311645450259,
      "grad_norm": 0.0010734983952715993,
      "learning_rate": 0.12934312505980916,
      "loss": 0.2028,
      "num_input_tokens_seen": 18855200,
      "step": 21760
    },
    {
      "epoch": 10.261669024045261,
      "grad_norm": 0.0016369061777368188,
      "learning_rate": 0.1292847830229443,
      "loss": 0.1998,
      "num_input_tokens_seen": 18859408,
      "step": 21765
    },
    {
      "epoch": 10.264026402640264,
      "grad_norm": 0.0010639375541359186,
      "learning_rate": 0.12922644418062626,
      "loss": 0.2318,
      "num_input_tokens_seen": 18863216,
      "step": 21770
    },
    {
      "epoch": 10.266383781235266,
      "grad_norm": 0.0014716271543875337,
      "learning_rate": 0.1291681085418515,
      "loss": 0.228,
      "num_input_tokens_seen": 18867184,
      "step": 21775
    },
    {
      "epoch": 10.268741159830268,
      "grad_norm": 0.002565921051427722,
      "learning_rate": 0.12910977611561628,
      "loss": 0.2596,
      "num_input_tokens_seen": 18871856,
      "step": 21780
    },
    {
      "epoch": 10.27109853842527,
      "grad_norm": 0.0009796597296372056,
      "learning_rate": 0.1290514469109161,
      "loss": 0.2189,
      "num_input_tokens_seen": 18876384,
      "step": 21785
    },
    {
      "epoch": 10.273455917020273,
      "grad_norm": 0.0014451759634539485,
      "learning_rate": 0.128993120936746,
      "loss": 0.2018,
      "num_input_tokens_seen": 18880672,
      "step": 21790
    },
    {
      "epoch": 10.275813295615276,
      "grad_norm": 0.0010475526796653867,
      "learning_rate": 0.12893479820210071,
      "loss": 0.2241,
      "num_input_tokens_seen": 18884240,
      "step": 21795
    },
    {
      "epoch": 10.278170674210278,
      "grad_norm": 0.0010318325366824865,
      "learning_rate": 0.1288764787159742,
      "loss": 0.2365,
      "num_input_tokens_seen": 18888560,
      "step": 21800
    },
    {
      "epoch": 10.278170674210278,
      "eval_loss": 0.21806401014328003,
      "eval_runtime": 21.94,
      "eval_samples_per_second": 42.981,
      "eval_steps_per_second": 21.513,
      "num_input_tokens_seen": 18888560,
      "step": 21800
    },
    {
      "epoch": 10.28052805280528,
      "grad_norm": 0.0012473678216338158,
      "learning_rate": 0.1288181624873601,
      "loss": 0.205,
      "num_input_tokens_seen": 18893248,
      "step": 21805
    },
    {
      "epoch": 10.282885431400283,
      "grad_norm": 0.0009067459614016116,
      "learning_rate": 0.12875984952525163,
      "loss": 0.1993,
      "num_input_tokens_seen": 18898320,
      "step": 21810
    },
    {
      "epoch": 10.285242809995285,
      "grad_norm": 0.0009492195094935596,
      "learning_rate": 0.12870153983864122,
      "loss": 0.2185,
      "num_input_tokens_seen": 18902720,
      "step": 21815
    },
    {
      "epoch": 10.287600188590288,
      "grad_norm": 0.0007980326190590858,
      "learning_rate": 0.12864323343652104,
      "loss": 0.1979,
      "num_input_tokens_seen": 18906912,
      "step": 21820
    },
    {
      "epoch": 10.28995756718529,
      "grad_norm": 0.0014669899828732014,
      "learning_rate": 0.12858493032788268,
      "loss": 0.2947,
      "num_input_tokens_seen": 18911728,
      "step": 21825
    },
    {
      "epoch": 10.292314945780292,
      "grad_norm": 0.001951332320459187,
      "learning_rate": 0.12852663052171714,
      "loss": 0.1875,
      "num_input_tokens_seen": 18916000,
      "step": 21830
    },
    {
      "epoch": 10.294672324375295,
      "grad_norm": 0.0026865974068641663,
      "learning_rate": 0.12846833402701507,
      "loss": 0.2132,
      "num_input_tokens_seen": 18920320,
      "step": 21835
    },
    {
      "epoch": 10.297029702970297,
      "grad_norm": 0.0012008611811324954,
      "learning_rate": 0.12841004085276642,
      "loss": 0.2429,
      "num_input_tokens_seen": 18924272,
      "step": 21840
    },
    {
      "epoch": 10.2993870815653,
      "grad_norm": 0.0012169939000159502,
      "learning_rate": 0.12835175100796076,
      "loss": 0.2582,
      "num_input_tokens_seen": 18929360,
      "step": 21845
    },
    {
      "epoch": 10.301744460160302,
      "grad_norm": 0.00099290837533772,
      "learning_rate": 0.12829346450158724,
      "loss": 0.2283,
      "num_input_tokens_seen": 18934096,
      "step": 21850
    },
    {
      "epoch": 10.304101838755304,
      "grad_norm": 0.0013641746481880546,
      "learning_rate": 0.12823518134263423,
      "loss": 0.2261,
      "num_input_tokens_seen": 18938336,
      "step": 21855
    },
    {
      "epoch": 10.306459217350307,
      "grad_norm": 0.0021834061481058598,
      "learning_rate": 0.12817690154008973,
      "loss": 0.2287,
      "num_input_tokens_seen": 18942704,
      "step": 21860
    },
    {
      "epoch": 10.30881659594531,
      "grad_norm": 0.0013836021535098553,
      "learning_rate": 0.12811862510294134,
      "loss": 0.2121,
      "num_input_tokens_seen": 18946672,
      "step": 21865
    },
    {
      "epoch": 10.311173974540312,
      "grad_norm": 0.0007522696978412569,
      "learning_rate": 0.12806035204017585,
      "loss": 0.212,
      "num_input_tokens_seen": 18950800,
      "step": 21870
    },
    {
      "epoch": 10.313531353135314,
      "grad_norm": 0.0018207726534456015,
      "learning_rate": 0.12800208236077987,
      "loss": 0.272,
      "num_input_tokens_seen": 18955888,
      "step": 21875
    },
    {
      "epoch": 10.315888731730317,
      "grad_norm": 0.001096277846954763,
      "learning_rate": 0.12794381607373917,
      "loss": 0.2357,
      "num_input_tokens_seen": 18959792,
      "step": 21880
    },
    {
      "epoch": 10.318246110325319,
      "grad_norm": 0.00134360883384943,
      "learning_rate": 0.12788555318803924,
      "loss": 0.2199,
      "num_input_tokens_seen": 18964224,
      "step": 21885
    },
    {
      "epoch": 10.320603488920321,
      "grad_norm": 0.0019122592639178038,
      "learning_rate": 0.1278272937126649,
      "loss": 0.2062,
      "num_input_tokens_seen": 18968368,
      "step": 21890
    },
    {
      "epoch": 10.322960867515324,
      "grad_norm": 0.002998204668983817,
      "learning_rate": 0.1277690376566005,
      "loss": 0.2548,
      "num_input_tokens_seen": 18973152,
      "step": 21895
    },
    {
      "epoch": 10.325318246110326,
      "grad_norm": 0.002046105219051242,
      "learning_rate": 0.12771078502882985,
      "loss": 0.2357,
      "num_input_tokens_seen": 18976816,
      "step": 21900
    },
    {
      "epoch": 10.327675624705329,
      "grad_norm": 0.0011666463688015938,
      "learning_rate": 0.12765253583833633,
      "loss": 0.2324,
      "num_input_tokens_seen": 18980352,
      "step": 21905
    },
    {
      "epoch": 10.33003300330033,
      "grad_norm": 0.0013692479114979506,
      "learning_rate": 0.12759429009410256,
      "loss": 0.2182,
      "num_input_tokens_seen": 18984288,
      "step": 21910
    },
    {
      "epoch": 10.332390381895332,
      "grad_norm": 0.0012697133934125304,
      "learning_rate": 0.12753604780511085,
      "loss": 0.2178,
      "num_input_tokens_seen": 18988544,
      "step": 21915
    },
    {
      "epoch": 10.334747760490334,
      "grad_norm": 0.0011422073002904654,
      "learning_rate": 0.12747780898034283,
      "loss": 0.2119,
      "num_input_tokens_seen": 18992448,
      "step": 21920
    },
    {
      "epoch": 10.337105139085336,
      "grad_norm": 0.0009961493778973818,
      "learning_rate": 0.12741957362877973,
      "loss": 0.2043,
      "num_input_tokens_seen": 18997696,
      "step": 21925
    },
    {
      "epoch": 10.339462517680339,
      "grad_norm": 0.0008563697338104248,
      "learning_rate": 0.12736134175940214,
      "loss": 0.1927,
      "num_input_tokens_seen": 19001856,
      "step": 21930
    },
    {
      "epoch": 10.341819896275341,
      "grad_norm": 0.0036909363698214293,
      "learning_rate": 0.12730311338119016,
      "loss": 0.292,
      "num_input_tokens_seen": 19006672,
      "step": 21935
    },
    {
      "epoch": 10.344177274870344,
      "grad_norm": 0.0011630273656919599,
      "learning_rate": 0.12724488850312327,
      "loss": 0.2306,
      "num_input_tokens_seen": 19010432,
      "step": 21940
    },
    {
      "epoch": 10.346534653465346,
      "grad_norm": 0.0011693460401147604,
      "learning_rate": 0.1271866671341806,
      "loss": 0.2169,
      "num_input_tokens_seen": 19014720,
      "step": 21945
    },
    {
      "epoch": 10.348892032060348,
      "grad_norm": 0.001077774097211659,
      "learning_rate": 0.12712844928334047,
      "loss": 0.2276,
      "num_input_tokens_seen": 19018624,
      "step": 21950
    },
    {
      "epoch": 10.35124941065535,
      "grad_norm": 0.002058951184153557,
      "learning_rate": 0.12707023495958095,
      "loss": 0.2351,
      "num_input_tokens_seen": 19022880,
      "step": 21955
    },
    {
      "epoch": 10.353606789250353,
      "grad_norm": 0.002122128615155816,
      "learning_rate": 0.12701202417187932,
      "loss": 0.2265,
      "num_input_tokens_seen": 19026288,
      "step": 21960
    },
    {
      "epoch": 10.355964167845356,
      "grad_norm": 0.0019919213373214006,
      "learning_rate": 0.12695381692921243,
      "loss": 0.2278,
      "num_input_tokens_seen": 19030720,
      "step": 21965
    },
    {
      "epoch": 10.358321546440358,
      "grad_norm": 0.0008595167309977114,
      "learning_rate": 0.12689561324055665,
      "loss": 0.231,
      "num_input_tokens_seen": 19035856,
      "step": 21970
    },
    {
      "epoch": 10.36067892503536,
      "grad_norm": 0.0017992517678067088,
      "learning_rate": 0.12683741311488758,
      "loss": 0.2458,
      "num_input_tokens_seen": 19040464,
      "step": 21975
    },
    {
      "epoch": 10.363036303630363,
      "grad_norm": 0.001202175160869956,
      "learning_rate": 0.1267792165611805,
      "loss": 0.2181,
      "num_input_tokens_seen": 19044592,
      "step": 21980
    },
    {
      "epoch": 10.365393682225365,
      "grad_norm": 0.0018862357828766108,
      "learning_rate": 0.1267210235884101,
      "loss": 0.1967,
      "num_input_tokens_seen": 19048640,
      "step": 21985
    },
    {
      "epoch": 10.367751060820368,
      "grad_norm": 0.001172007410787046,
      "learning_rate": 0.12666283420555033,
      "loss": 0.2086,
      "num_input_tokens_seen": 19053008,
      "step": 21990
    },
    {
      "epoch": 10.37010843941537,
      "grad_norm": 0.005234818905591965,
      "learning_rate": 0.12660464842157487,
      "loss": 0.2236,
      "num_input_tokens_seen": 19057808,
      "step": 21995
    },
    {
      "epoch": 10.372465818010372,
      "grad_norm": 0.001749674091115594,
      "learning_rate": 0.1265464662454566,
      "loss": 0.2137,
      "num_input_tokens_seen": 19061328,
      "step": 22000
    },
    {
      "epoch": 10.372465818010372,
      "eval_loss": 0.21637344360351562,
      "eval_runtime": 21.9709,
      "eval_samples_per_second": 42.92,
      "eval_steps_per_second": 21.483,
      "num_input_tokens_seen": 19061328,
      "step": 22000
    },
    {
      "epoch": 10.374823196605375,
      "grad_norm": 0.001183142652735114,
      "learning_rate": 0.12648828768616793,
      "loss": 0.2039,
      "num_input_tokens_seen": 19065376,
      "step": 22005
    },
    {
      "epoch": 10.377180575200377,
      "grad_norm": 0.0014305043732747436,
      "learning_rate": 0.12643011275268085,
      "loss": 0.2252,
      "num_input_tokens_seen": 19069360,
      "step": 22010
    },
    {
      "epoch": 10.37953795379538,
      "grad_norm": 0.0009327513398602605,
      "learning_rate": 0.1263719414539665,
      "loss": 0.1765,
      "num_input_tokens_seen": 19073296,
      "step": 22015
    },
    {
      "epoch": 10.381895332390382,
      "grad_norm": 0.0016643195413053036,
      "learning_rate": 0.1263137737989957,
      "loss": 0.1986,
      "num_input_tokens_seen": 19077008,
      "step": 22020
    },
    {
      "epoch": 10.384252710985384,
      "grad_norm": 0.0025107041001319885,
      "learning_rate": 0.1262556097967387,
      "loss": 0.2359,
      "num_input_tokens_seen": 19081856,
      "step": 22025
    },
    {
      "epoch": 10.386610089580387,
      "grad_norm": 0.001126704621128738,
      "learning_rate": 0.126197449456165,
      "loss": 0.2059,
      "num_input_tokens_seen": 19086576,
      "step": 22030
    },
    {
      "epoch": 10.38896746817539,
      "grad_norm": 0.0016975468024611473,
      "learning_rate": 0.12613929278624378,
      "loss": 0.2447,
      "num_input_tokens_seen": 19091408,
      "step": 22035
    },
    {
      "epoch": 10.391324846770392,
      "grad_norm": 0.0032961175311356783,
      "learning_rate": 0.12608113979594343,
      "loss": 0.2263,
      "num_input_tokens_seen": 19096608,
      "step": 22040
    },
    {
      "epoch": 10.393682225365394,
      "grad_norm": 0.001177679980173707,
      "learning_rate": 0.1260229904942319,
      "loss": 0.2087,
      "num_input_tokens_seen": 19100768,
      "step": 22045
    },
    {
      "epoch": 10.396039603960396,
      "grad_norm": 0.002918731886893511,
      "learning_rate": 0.12596484489007662,
      "loss": 0.2051,
      "num_input_tokens_seen": 19105440,
      "step": 22050
    },
    {
      "epoch": 10.398396982555399,
      "grad_norm": 0.0013701600255444646,
      "learning_rate": 0.1259067029924442,
      "loss": 0.2047,
      "num_input_tokens_seen": 19109488,
      "step": 22055
    },
    {
      "epoch": 10.400754361150401,
      "grad_norm": 0.002191714243963361,
      "learning_rate": 0.12584856481030096,
      "loss": 0.2083,
      "num_input_tokens_seen": 19113808,
      "step": 22060
    },
    {
      "epoch": 10.403111739745404,
      "grad_norm": 0.0009776485385373235,
      "learning_rate": 0.12579043035261261,
      "loss": 0.2069,
      "num_input_tokens_seen": 19117488,
      "step": 22065
    },
    {
      "epoch": 10.405469118340406,
      "grad_norm": 0.0017057891236618161,
      "learning_rate": 0.1257322996283441,
      "loss": 0.2275,
      "num_input_tokens_seen": 19122448,
      "step": 22070
    },
    {
      "epoch": 10.407826496935408,
      "grad_norm": 0.0008475471404381096,
      "learning_rate": 0.12567417264645994,
      "loss": 0.2028,
      "num_input_tokens_seen": 19126384,
      "step": 22075
    },
    {
      "epoch": 10.41018387553041,
      "grad_norm": 0.0010021085618063807,
      "learning_rate": 0.12561604941592408,
      "loss": 0.2502,
      "num_input_tokens_seen": 19130944,
      "step": 22080
    },
    {
      "epoch": 10.412541254125413,
      "grad_norm": 0.0011967484606429935,
      "learning_rate": 0.12555792994569978,
      "loss": 0.2131,
      "num_input_tokens_seen": 19135232,
      "step": 22085
    },
    {
      "epoch": 10.414898632720416,
      "grad_norm": 0.0009032103698700666,
      "learning_rate": 0.1254998142447499,
      "loss": 0.2237,
      "num_input_tokens_seen": 19139664,
      "step": 22090
    },
    {
      "epoch": 10.417256011315418,
      "grad_norm": 0.0019213058985769749,
      "learning_rate": 0.1254417023220365,
      "loss": 0.2302,
      "num_input_tokens_seen": 19144336,
      "step": 22095
    },
    {
      "epoch": 10.41961338991042,
      "grad_norm": 0.001135032158344984,
      "learning_rate": 0.12538359418652126,
      "loss": 0.1976,
      "num_input_tokens_seen": 19150320,
      "step": 22100
    },
    {
      "epoch": 10.421970768505423,
      "grad_norm": 0.0012120010796934366,
      "learning_rate": 0.12532548984716513,
      "loss": 0.2029,
      "num_input_tokens_seen": 19154768,
      "step": 22105
    },
    {
      "epoch": 10.424328147100425,
      "grad_norm": 0.001986401155591011,
      "learning_rate": 0.12526738931292855,
      "loss": 0.2255,
      "num_input_tokens_seen": 19158976,
      "step": 22110
    },
    {
      "epoch": 10.426685525695426,
      "grad_norm": 0.0014710555551573634,
      "learning_rate": 0.1252092925927714,
      "loss": 0.2616,
      "num_input_tokens_seen": 19162960,
      "step": 22115
    },
    {
      "epoch": 10.429042904290428,
      "grad_norm": 0.001634966116398573,
      "learning_rate": 0.12515119969565278,
      "loss": 0.2467,
      "num_input_tokens_seen": 19167296,
      "step": 22120
    },
    {
      "epoch": 10.43140028288543,
      "grad_norm": 0.0015396481612697244,
      "learning_rate": 0.12509311063053144,
      "loss": 0.2358,
      "num_input_tokens_seen": 19171104,
      "step": 22125
    },
    {
      "epoch": 10.433757661480433,
      "grad_norm": 0.0010717500699684024,
      "learning_rate": 0.1250350254063655,
      "loss": 0.2317,
      "num_input_tokens_seen": 19174880,
      "step": 22130
    },
    {
      "epoch": 10.436115040075435,
      "grad_norm": 0.001125298091210425,
      "learning_rate": 0.1249769440321123,
      "loss": 0.2419,
      "num_input_tokens_seen": 19179296,
      "step": 22135
    },
    {
      "epoch": 10.438472418670438,
      "grad_norm": 0.001966638257727027,
      "learning_rate": 0.12491886651672884,
      "loss": 0.2202,
      "num_input_tokens_seen": 19183504,
      "step": 22140
    },
    {
      "epoch": 10.44082979726544,
      "grad_norm": 0.0013706288300454617,
      "learning_rate": 0.12486079286917139,
      "loss": 0.2318,
      "num_input_tokens_seen": 19188096,
      "step": 22145
    },
    {
      "epoch": 10.443187175860443,
      "grad_norm": 0.0011789592681452632,
      "learning_rate": 0.12480272309839553,
      "loss": 0.2247,
      "num_input_tokens_seen": 19192464,
      "step": 22150
    },
    {
      "epoch": 10.445544554455445,
      "grad_norm": 0.0007968675927259028,
      "learning_rate": 0.12474465721335648,
      "loss": 0.2241,
      "num_input_tokens_seen": 19196576,
      "step": 22155
    },
    {
      "epoch": 10.447901933050447,
      "grad_norm": 0.0009436819818802178,
      "learning_rate": 0.12468659522300861,
      "loss": 0.2351,
      "num_input_tokens_seen": 19200784,
      "step": 22160
    },
    {
      "epoch": 10.45025931164545,
      "grad_norm": 0.0018445131136104465,
      "learning_rate": 0.12462853713630584,
      "loss": 0.2343,
      "num_input_tokens_seen": 19204864,
      "step": 22165
    },
    {
      "epoch": 10.452616690240452,
      "grad_norm": 0.0012664683163166046,
      "learning_rate": 0.12457048296220156,
      "loss": 0.232,
      "num_input_tokens_seen": 19209232,
      "step": 22170
    },
    {
      "epoch": 10.454974068835455,
      "grad_norm": 0.002282599685713649,
      "learning_rate": 0.12451243270964832,
      "loss": 0.2263,
      "num_input_tokens_seen": 19213552,
      "step": 22175
    },
    {
      "epoch": 10.457331447430457,
      "grad_norm": 0.001316384761594236,
      "learning_rate": 0.12445438638759827,
      "loss": 0.2257,
      "num_input_tokens_seen": 19218224,
      "step": 22180
    },
    {
      "epoch": 10.45968882602546,
      "grad_norm": 0.0010058984626084566,
      "learning_rate": 0.1243963440050029,
      "loss": 0.2061,
      "num_input_tokens_seen": 19222880,
      "step": 22185
    },
    {
      "epoch": 10.462046204620462,
      "grad_norm": 0.002403038088232279,
      "learning_rate": 0.12433830557081298,
      "loss": 0.2533,
      "num_input_tokens_seen": 19227040,
      "step": 22190
    },
    {
      "epoch": 10.464403583215464,
      "grad_norm": 0.004003562498837709,
      "learning_rate": 0.12428027109397889,
      "loss": 0.2293,
      "num_input_tokens_seen": 19231008,
      "step": 22195
    },
    {
      "epoch": 10.466760961810467,
      "grad_norm": 0.0011721281334757805,
      "learning_rate": 0.12422224058345015,
      "loss": 0.2118,
      "num_input_tokens_seen": 19236176,
      "step": 22200
    },
    {
      "epoch": 10.466760961810467,
      "eval_loss": 0.2165602296590805,
      "eval_runtime": 21.8687,
      "eval_samples_per_second": 43.121,
      "eval_steps_per_second": 21.583,
      "num_input_tokens_seen": 19236176,
      "step": 22200
    },
    {
      "epoch": 10.469118340405469,
      "grad_norm": 0.002238075016066432,
      "learning_rate": 0.12416421404817583,
      "loss": 0.2135,
      "num_input_tokens_seen": 19239840,
      "step": 22205
    },
    {
      "epoch": 10.471475719000471,
      "grad_norm": 0.001323640695773065,
      "learning_rate": 0.12410619149710447,
      "loss": 0.2421,
      "num_input_tokens_seen": 19244080,
      "step": 22210
    },
    {
      "epoch": 10.473833097595474,
      "grad_norm": 0.000849499658215791,
      "learning_rate": 0.12404817293918374,
      "loss": 0.2487,
      "num_input_tokens_seen": 19247856,
      "step": 22215
    },
    {
      "epoch": 10.476190476190476,
      "grad_norm": 0.0012140254257246852,
      "learning_rate": 0.12399015838336086,
      "loss": 0.2171,
      "num_input_tokens_seen": 19251760,
      "step": 22220
    },
    {
      "epoch": 10.478547854785479,
      "grad_norm": 0.0019237084779888391,
      "learning_rate": 0.12393214783858246,
      "loss": 0.2062,
      "num_input_tokens_seen": 19256480,
      "step": 22225
    },
    {
      "epoch": 10.480905233380481,
      "grad_norm": 0.0017789190169423819,
      "learning_rate": 0.1238741413137944,
      "loss": 0.2512,
      "num_input_tokens_seen": 19260720,
      "step": 22230
    },
    {
      "epoch": 10.483262611975483,
      "grad_norm": 0.0009467643685638905,
      "learning_rate": 0.12381613881794212,
      "loss": 0.2171,
      "num_input_tokens_seen": 19264672,
      "step": 22235
    },
    {
      "epoch": 10.485619990570486,
      "grad_norm": 0.002289701486006379,
      "learning_rate": 0.12375814035997022,
      "loss": 0.2002,
      "num_input_tokens_seen": 19269312,
      "step": 22240
    },
    {
      "epoch": 10.487977369165488,
      "grad_norm": 0.001704637543298304,
      "learning_rate": 0.12370014594882285,
      "loss": 0.2272,
      "num_input_tokens_seen": 19273920,
      "step": 22245
    },
    {
      "epoch": 10.49033474776049,
      "grad_norm": 0.0008350195712409914,
      "learning_rate": 0.12364215559344356,
      "loss": 0.2048,
      "num_input_tokens_seen": 19278112,
      "step": 22250
    },
    {
      "epoch": 10.492692126355493,
      "grad_norm": 0.0017205007607117295,
      "learning_rate": 0.12358416930277506,
      "loss": 0.2258,
      "num_input_tokens_seen": 19282784,
      "step": 22255
    },
    {
      "epoch": 10.495049504950495,
      "grad_norm": 0.0012135091237723827,
      "learning_rate": 0.1235261870857596,
      "loss": 0.2063,
      "num_input_tokens_seen": 19287840,
      "step": 22260
    },
    {
      "epoch": 10.497406883545498,
      "grad_norm": 0.0016083796508610249,
      "learning_rate": 0.12346820895133884,
      "loss": 0.2142,
      "num_input_tokens_seen": 19291664,
      "step": 22265
    },
    {
      "epoch": 10.4997642621405,
      "grad_norm": 0.0015308193396776915,
      "learning_rate": 0.12341023490845361,
      "loss": 0.2064,
      "num_input_tokens_seen": 19295872,
      "step": 22270
    },
    {
      "epoch": 10.502121640735503,
      "grad_norm": 0.0011116726091131568,
      "learning_rate": 0.12335226496604437,
      "loss": 0.2,
      "num_input_tokens_seen": 19300320,
      "step": 22275
    },
    {
      "epoch": 10.504479019330505,
      "grad_norm": 0.0010352479293942451,
      "learning_rate": 0.12329429913305069,
      "loss": 0.2634,
      "num_input_tokens_seen": 19303872,
      "step": 22280
    },
    {
      "epoch": 10.506836397925507,
      "grad_norm": 0.0016570861916989088,
      "learning_rate": 0.12323633741841171,
      "loss": 0.2394,
      "num_input_tokens_seen": 19308656,
      "step": 22285
    },
    {
      "epoch": 10.50919377652051,
      "grad_norm": 0.002036277437582612,
      "learning_rate": 0.12317837983106583,
      "loss": 0.219,
      "num_input_tokens_seen": 19312928,
      "step": 22290
    },
    {
      "epoch": 10.511551155115512,
      "grad_norm": 0.0009706165874376893,
      "learning_rate": 0.12312042637995087,
      "loss": 0.21,
      "num_input_tokens_seen": 19316912,
      "step": 22295
    },
    {
      "epoch": 10.513908533710515,
      "grad_norm": 0.001657850225456059,
      "learning_rate": 0.12306247707400389,
      "loss": 0.2245,
      "num_input_tokens_seen": 19321040,
      "step": 22300
    },
    {
      "epoch": 10.516265912305517,
      "grad_norm": 0.0011152183869853616,
      "learning_rate": 0.12300453192216154,
      "loss": 0.2439,
      "num_input_tokens_seen": 19324960,
      "step": 22305
    },
    {
      "epoch": 10.518623290900518,
      "grad_norm": 0.001008142833597958,
      "learning_rate": 0.12294659093335956,
      "loss": 0.2269,
      "num_input_tokens_seen": 19329648,
      "step": 22310
    },
    {
      "epoch": 10.520980669495522,
      "grad_norm": 0.0024475669488310814,
      "learning_rate": 0.12288865411653327,
      "loss": 0.2352,
      "num_input_tokens_seen": 19333296,
      "step": 22315
    },
    {
      "epoch": 10.523338048090523,
      "grad_norm": 0.0009934682166203856,
      "learning_rate": 0.12283072148061717,
      "loss": 0.2375,
      "num_input_tokens_seen": 19337952,
      "step": 22320
    },
    {
      "epoch": 10.525695426685525,
      "grad_norm": 0.0014270105166360736,
      "learning_rate": 0.12277279303454529,
      "loss": 0.2279,
      "num_input_tokens_seen": 19341616,
      "step": 22325
    },
    {
      "epoch": 10.528052805280527,
      "grad_norm": 0.0014033850748091936,
      "learning_rate": 0.12271486878725091,
      "loss": 0.2267,
      "num_input_tokens_seen": 19345568,
      "step": 22330
    },
    {
      "epoch": 10.53041018387553,
      "grad_norm": 0.0014108485775068402,
      "learning_rate": 0.12265694874766658,
      "loss": 0.2254,
      "num_input_tokens_seen": 19349568,
      "step": 22335
    },
    {
      "epoch": 10.532767562470532,
      "grad_norm": 0.0009394697262905538,
      "learning_rate": 0.12259903292472435,
      "loss": 0.2194,
      "num_input_tokens_seen": 19352816,
      "step": 22340
    },
    {
      "epoch": 10.535124941065535,
      "grad_norm": 0.0010639302199706435,
      "learning_rate": 0.12254112132735567,
      "loss": 0.2169,
      "num_input_tokens_seen": 19357392,
      "step": 22345
    },
    {
      "epoch": 10.537482319660537,
      "grad_norm": 0.0010175523348152637,
      "learning_rate": 0.12248321396449108,
      "loss": 0.2341,
      "num_input_tokens_seen": 19361376,
      "step": 22350
    },
    {
      "epoch": 10.53983969825554,
      "grad_norm": 0.001321848132647574,
      "learning_rate": 0.12242531084506075,
      "loss": 0.2332,
      "num_input_tokens_seen": 19365680,
      "step": 22355
    },
    {
      "epoch": 10.542197076850542,
      "grad_norm": 0.0010497123003005981,
      "learning_rate": 0.122367411977994,
      "loss": 0.2318,
      "num_input_tokens_seen": 19370256,
      "step": 22360
    },
    {
      "epoch": 10.544554455445544,
      "grad_norm": 0.00101200130302459,
      "learning_rate": 0.12230951737221954,
      "loss": 0.2122,
      "num_input_tokens_seen": 19373936,
      "step": 22365
    },
    {
      "epoch": 10.546911834040547,
      "grad_norm": 0.0010317484848201275,
      "learning_rate": 0.12225162703666555,
      "loss": 0.2304,
      "num_input_tokens_seen": 19378096,
      "step": 22370
    },
    {
      "epoch": 10.549269212635549,
      "grad_norm": 0.002733059925958514,
      "learning_rate": 0.1221937409802593,
      "loss": 0.2341,
      "num_input_tokens_seen": 19383152,
      "step": 22375
    },
    {
      "epoch": 10.551626591230551,
      "grad_norm": 0.0022361630108207464,
      "learning_rate": 0.12213585921192768,
      "loss": 0.2058,
      "num_input_tokens_seen": 19388176,
      "step": 22380
    },
    {
      "epoch": 10.553983969825554,
      "grad_norm": 0.0015509643126279116,
      "learning_rate": 0.1220779817405967,
      "loss": 0.2464,
      "num_input_tokens_seen": 19392720,
      "step": 22385
    },
    {
      "epoch": 10.556341348420556,
      "grad_norm": 0.00231257826089859,
      "learning_rate": 0.12202010857519181,
      "loss": 0.2198,
      "num_input_tokens_seen": 19396944,
      "step": 22390
    },
    {
      "epoch": 10.558698727015559,
      "grad_norm": 0.0010198347736150026,
      "learning_rate": 0.12196223972463785,
      "loss": 0.2367,
      "num_input_tokens_seen": 19400576,
      "step": 22395
    },
    {
      "epoch": 10.561056105610561,
      "grad_norm": 0.000954817864112556,
      "learning_rate": 0.12190437519785885,
      "loss": 0.208,
      "num_input_tokens_seen": 19404288,
      "step": 22400
    },
    {
      "epoch": 10.561056105610561,
      "eval_loss": 0.21777023375034332,
      "eval_runtime": 21.9132,
      "eval_samples_per_second": 43.034,
      "eval_steps_per_second": 21.54,
      "num_input_tokens_seen": 19404288,
      "step": 22400
    },
    {
      "epoch": 10.563413484205563,
      "grad_norm": 0.0006534851272590458,
      "learning_rate": 0.12184651500377823,
      "loss": 0.2141,
      "num_input_tokens_seen": 19407504,
      "step": 22405
    },
    {
      "epoch": 10.565770862800566,
      "grad_norm": 0.000744357705116272,
      "learning_rate": 0.12178865915131885,
      "loss": 0.226,
      "num_input_tokens_seen": 19411504,
      "step": 22410
    },
    {
      "epoch": 10.568128241395568,
      "grad_norm": 0.001136588747613132,
      "learning_rate": 0.1217308076494027,
      "loss": 0.2159,
      "num_input_tokens_seen": 19415856,
      "step": 22415
    },
    {
      "epoch": 10.57048561999057,
      "grad_norm": 0.0008811786537989974,
      "learning_rate": 0.12167296050695134,
      "loss": 0.2348,
      "num_input_tokens_seen": 19420304,
      "step": 22420
    },
    {
      "epoch": 10.572842998585573,
      "grad_norm": 0.0009200049098581076,
      "learning_rate": 0.12161511773288536,
      "loss": 0.2361,
      "num_input_tokens_seen": 19424640,
      "step": 22425
    },
    {
      "epoch": 10.575200377180575,
      "grad_norm": 0.0010602250695228577,
      "learning_rate": 0.121557279336125,
      "loss": 0.2243,
      "num_input_tokens_seen": 19429072,
      "step": 22430
    },
    {
      "epoch": 10.577557755775578,
      "grad_norm": 0.0009873583912849426,
      "learning_rate": 0.12149944532558957,
      "loss": 0.2196,
      "num_input_tokens_seen": 19433232,
      "step": 22435
    },
    {
      "epoch": 10.57991513437058,
      "grad_norm": 0.0016008374514058232,
      "learning_rate": 0.12144161571019785,
      "loss": 0.2128,
      "num_input_tokens_seen": 19438240,
      "step": 22440
    },
    {
      "epoch": 10.582272512965583,
      "grad_norm": 0.0009181021014228463,
      "learning_rate": 0.12138379049886781,
      "loss": 0.2155,
      "num_input_tokens_seen": 19442832,
      "step": 22445
    },
    {
      "epoch": 10.584629891560585,
      "grad_norm": 0.0011158337583765388,
      "learning_rate": 0.12132596970051697,
      "loss": 0.2444,
      "num_input_tokens_seen": 19447152,
      "step": 22450
    },
    {
      "epoch": 10.586987270155587,
      "grad_norm": 0.002858510008081794,
      "learning_rate": 0.12126815332406189,
      "loss": 0.2096,
      "num_input_tokens_seen": 19451872,
      "step": 22455
    },
    {
      "epoch": 10.58934464875059,
      "grad_norm": 0.0010608130833134055,
      "learning_rate": 0.12121034137841868,
      "loss": 0.2306,
      "num_input_tokens_seen": 19456448,
      "step": 22460
    },
    {
      "epoch": 10.591702027345592,
      "grad_norm": 0.0011453211773186922,
      "learning_rate": 0.12115253387250258,
      "loss": 0.1844,
      "num_input_tokens_seen": 19460896,
      "step": 22465
    },
    {
      "epoch": 10.594059405940595,
      "grad_norm": 0.001133226789534092,
      "learning_rate": 0.12109473081522831,
      "loss": 0.1907,
      "num_input_tokens_seen": 19465088,
      "step": 22470
    },
    {
      "epoch": 10.596416784535597,
      "grad_norm": 0.0020264792256057262,
      "learning_rate": 0.12103693221550982,
      "loss": 0.1822,
      "num_input_tokens_seen": 19469232,
      "step": 22475
    },
    {
      "epoch": 10.5987741631306,
      "grad_norm": 0.002018054248765111,
      "learning_rate": 0.12097913808226027,
      "loss": 0.2356,
      "num_input_tokens_seen": 19473424,
      "step": 22480
    },
    {
      "epoch": 10.601131541725602,
      "grad_norm": 0.0011062873527407646,
      "learning_rate": 0.12092134842439234,
      "loss": 0.246,
      "num_input_tokens_seen": 19477728,
      "step": 22485
    },
    {
      "epoch": 10.603488920320604,
      "grad_norm": 0.0013154511107131839,
      "learning_rate": 0.12086356325081798,
      "loss": 0.262,
      "num_input_tokens_seen": 19481888,
      "step": 22490
    },
    {
      "epoch": 10.605846298915607,
      "grad_norm": 0.0033291231375187635,
      "learning_rate": 0.12080578257044824,
      "loss": 0.2015,
      "num_input_tokens_seen": 19485872,
      "step": 22495
    },
    {
      "epoch": 10.608203677510609,
      "grad_norm": 0.0007461765198968351,
      "learning_rate": 0.12074800639219378,
      "loss": 0.2107,
      "num_input_tokens_seen": 19489952,
      "step": 22500
    },
    {
      "epoch": 10.61056105610561,
      "grad_norm": 0.0017839580541476607,
      "learning_rate": 0.12069023472496428,
      "loss": 0.2118,
      "num_input_tokens_seen": 19493568,
      "step": 22505
    },
    {
      "epoch": 10.612918434700614,
      "grad_norm": 0.0012284445110708475,
      "learning_rate": 0.12063246757766893,
      "loss": 0.2465,
      "num_input_tokens_seen": 19497216,
      "step": 22510
    },
    {
      "epoch": 10.615275813295614,
      "grad_norm": 0.001055700471624732,
      "learning_rate": 0.12057470495921618,
      "loss": 0.213,
      "num_input_tokens_seen": 19501984,
      "step": 22515
    },
    {
      "epoch": 10.617633191890617,
      "grad_norm": 0.0009513085824437439,
      "learning_rate": 0.12051694687851364,
      "loss": 0.2249,
      "num_input_tokens_seen": 19506528,
      "step": 22520
    },
    {
      "epoch": 10.61999057048562,
      "grad_norm": 0.0009714479092508554,
      "learning_rate": 0.12045919334446839,
      "loss": 0.209,
      "num_input_tokens_seen": 19510784,
      "step": 22525
    },
    {
      "epoch": 10.622347949080622,
      "grad_norm": 0.001339233247563243,
      "learning_rate": 0.12040144436598683,
      "loss": 0.2142,
      "num_input_tokens_seen": 19515392,
      "step": 22530
    },
    {
      "epoch": 10.624705327675624,
      "grad_norm": 0.0016931908903643489,
      "learning_rate": 0.12034369995197444,
      "loss": 0.224,
      "num_input_tokens_seen": 19519424,
      "step": 22535
    },
    {
      "epoch": 10.627062706270626,
      "grad_norm": 0.0008039309759624302,
      "learning_rate": 0.12028596011133627,
      "loss": 0.2201,
      "num_input_tokens_seen": 19523424,
      "step": 22540
    },
    {
      "epoch": 10.629420084865629,
      "grad_norm": 0.0011120367562398314,
      "learning_rate": 0.12022822485297643,
      "loss": 0.2333,
      "num_input_tokens_seen": 19527904,
      "step": 22545
    },
    {
      "epoch": 10.631777463460631,
      "grad_norm": 0.0018162421183660626,
      "learning_rate": 0.12017049418579843,
      "loss": 0.2011,
      "num_input_tokens_seen": 19532192,
      "step": 22550
    },
    {
      "epoch": 10.634134842055634,
      "grad_norm": 0.0018988415831699967,
      "learning_rate": 0.12011276811870514,
      "loss": 0.2266,
      "num_input_tokens_seen": 19536992,
      "step": 22555
    },
    {
      "epoch": 10.636492220650636,
      "grad_norm": 0.002750081242993474,
      "learning_rate": 0.12005504666059852,
      "loss": 0.1877,
      "num_input_tokens_seen": 19541168,
      "step": 22560
    },
    {
      "epoch": 10.638849599245638,
      "grad_norm": 0.0010525877587497234,
      "learning_rate": 0.11999732982038003,
      "loss": 0.2223,
      "num_input_tokens_seen": 19545024,
      "step": 22565
    },
    {
      "epoch": 10.64120697784064,
      "grad_norm": 0.0027161044999957085,
      "learning_rate": 0.11993961760695038,
      "loss": 0.2283,
      "num_input_tokens_seen": 19549120,
      "step": 22570
    },
    {
      "epoch": 10.643564356435643,
      "grad_norm": 0.0010227242019027472,
      "learning_rate": 0.11988191002920942,
      "loss": 0.2197,
      "num_input_tokens_seen": 19553520,
      "step": 22575
    },
    {
      "epoch": 10.645921735030646,
      "grad_norm": 0.0016675328370183706,
      "learning_rate": 0.11982420709605641,
      "loss": 0.2349,
      "num_input_tokens_seen": 19558096,
      "step": 22580
    },
    {
      "epoch": 10.648279113625648,
      "grad_norm": 0.0018678001360967755,
      "learning_rate": 0.11976650881638991,
      "loss": 0.2082,
      "num_input_tokens_seen": 19562224,
      "step": 22585
    },
    {
      "epoch": 10.65063649222065,
      "grad_norm": 0.0014401058433577418,
      "learning_rate": 0.11970881519910764,
      "loss": 0.2128,
      "num_input_tokens_seen": 19566000,
      "step": 22590
    },
    {
      "epoch": 10.652993870815653,
      "grad_norm": 0.0012255280744284391,
      "learning_rate": 0.1196511262531068,
      "loss": 0.2505,
      "num_input_tokens_seen": 19569680,
      "step": 22595
    },
    {
      "epoch": 10.655351249410655,
      "grad_norm": 0.001619109301827848,
      "learning_rate": 0.11959344198728361,
      "loss": 0.2201,
      "num_input_tokens_seen": 19574224,
      "step": 22600
    },
    {
      "epoch": 10.655351249410655,
      "eval_loss": 0.2144869565963745,
      "eval_runtime": 21.9154,
      "eval_samples_per_second": 43.029,
      "eval_steps_per_second": 21.537,
      "num_input_tokens_seen": 19574224,
      "step": 22600
    },
    {
      "epoch": 10.657708628005658,
      "grad_norm": 0.0012130734976381063,
      "learning_rate": 0.11953576241053378,
      "loss": 0.2371,
      "num_input_tokens_seen": 19578000,
      "step": 22605
    },
    {
      "epoch": 10.66006600660066,
      "grad_norm": 0.0009198078187182546,
      "learning_rate": 0.11947808753175228,
      "loss": 0.2005,
      "num_input_tokens_seen": 19581552,
      "step": 22610
    },
    {
      "epoch": 10.662423385195662,
      "grad_norm": 0.0015829779440537095,
      "learning_rate": 0.1194204173598332,
      "loss": 0.1999,
      "num_input_tokens_seen": 19585184,
      "step": 22615
    },
    {
      "epoch": 10.664780763790665,
      "grad_norm": 0.0023783536162227392,
      "learning_rate": 0.11936275190367007,
      "loss": 0.2211,
      "num_input_tokens_seen": 19589840,
      "step": 22620
    },
    {
      "epoch": 10.667138142385667,
      "grad_norm": 0.0006507565267384052,
      "learning_rate": 0.11930509117215563,
      "loss": 0.2127,
      "num_input_tokens_seen": 19593744,
      "step": 22625
    },
    {
      "epoch": 10.66949552098067,
      "grad_norm": 0.0011839231010526419,
      "learning_rate": 0.11924743517418179,
      "loss": 0.1765,
      "num_input_tokens_seen": 19598112,
      "step": 22630
    },
    {
      "epoch": 10.671852899575672,
      "grad_norm": 0.0015322677791118622,
      "learning_rate": 0.11918978391864,
      "loss": 0.2543,
      "num_input_tokens_seen": 19602320,
      "step": 22635
    },
    {
      "epoch": 10.674210278170674,
      "grad_norm": 0.0008136096294037998,
      "learning_rate": 0.11913213741442065,
      "loss": 0.2067,
      "num_input_tokens_seen": 19606416,
      "step": 22640
    },
    {
      "epoch": 10.676567656765677,
      "grad_norm": 0.0014673310797661543,
      "learning_rate": 0.11907449567041364,
      "loss": 0.2607,
      "num_input_tokens_seen": 19610896,
      "step": 22645
    },
    {
      "epoch": 10.67892503536068,
      "grad_norm": 0.00149513257201761,
      "learning_rate": 0.11901685869550803,
      "loss": 0.2148,
      "num_input_tokens_seen": 19615904,
      "step": 22650
    },
    {
      "epoch": 10.681282413955682,
      "grad_norm": 0.0008098832331597805,
      "learning_rate": 0.1189592264985922,
      "loss": 0.2228,
      "num_input_tokens_seen": 19620096,
      "step": 22655
    },
    {
      "epoch": 10.683639792550684,
      "grad_norm": 0.0010021724738180637,
      "learning_rate": 0.11890159908855373,
      "loss": 0.2287,
      "num_input_tokens_seen": 19623712,
      "step": 22660
    },
    {
      "epoch": 10.685997171145686,
      "grad_norm": 0.0018505444750189781,
      "learning_rate": 0.11884397647427941,
      "loss": 0.2046,
      "num_input_tokens_seen": 19628432,
      "step": 22665
    },
    {
      "epoch": 10.688354549740689,
      "grad_norm": 0.0008550783968530595,
      "learning_rate": 0.11878635866465546,
      "loss": 0.2025,
      "num_input_tokens_seen": 19631952,
      "step": 22670
    },
    {
      "epoch": 10.690711928335691,
      "grad_norm": 0.0009578528697602451,
      "learning_rate": 0.11872874566856734,
      "loss": 0.2203,
      "num_input_tokens_seen": 19636432,
      "step": 22675
    },
    {
      "epoch": 10.693069306930694,
      "grad_norm": 0.0013436555163934827,
      "learning_rate": 0.11867113749489955,
      "loss": 0.2039,
      "num_input_tokens_seen": 19641120,
      "step": 22680
    },
    {
      "epoch": 10.695426685525696,
      "grad_norm": 0.0012823735596612096,
      "learning_rate": 0.11861353415253607,
      "loss": 0.2605,
      "num_input_tokens_seen": 19645216,
      "step": 22685
    },
    {
      "epoch": 10.697784064120698,
      "grad_norm": 0.0011053390335291624,
      "learning_rate": 0.11855593565036011,
      "loss": 0.2253,
      "num_input_tokens_seen": 19649232,
      "step": 22690
    },
    {
      "epoch": 10.700141442715701,
      "grad_norm": 0.0015341602265834808,
      "learning_rate": 0.11849834199725394,
      "loss": 0.2464,
      "num_input_tokens_seen": 19653344,
      "step": 22695
    },
    {
      "epoch": 10.702498821310703,
      "grad_norm": 0.0008902903646230698,
      "learning_rate": 0.1184407532020994,
      "loss": 0.2098,
      "num_input_tokens_seen": 19657904,
      "step": 22700
    },
    {
      "epoch": 10.704856199905706,
      "grad_norm": 0.0007464297232218087,
      "learning_rate": 0.11838316927377723,
      "loss": 0.2294,
      "num_input_tokens_seen": 19662368,
      "step": 22705
    },
    {
      "epoch": 10.707213578500706,
      "grad_norm": 0.0011085174046456814,
      "learning_rate": 0.11832559022116766,
      "loss": 0.2135,
      "num_input_tokens_seen": 19666608,
      "step": 22710
    },
    {
      "epoch": 10.70957095709571,
      "grad_norm": 0.0010259922128170729,
      "learning_rate": 0.11826801605315022,
      "loss": 0.1963,
      "num_input_tokens_seen": 19671088,
      "step": 22715
    },
    {
      "epoch": 10.711928335690711,
      "grad_norm": 0.0010548167629167438,
      "learning_rate": 0.1182104467786034,
      "loss": 0.213,
      "num_input_tokens_seen": 19675568,
      "step": 22720
    },
    {
      "epoch": 10.714285714285714,
      "grad_norm": 0.0016320622526109219,
      "learning_rate": 0.1181528824064052,
      "loss": 0.1927,
      "num_input_tokens_seen": 19680336,
      "step": 22725
    },
    {
      "epoch": 10.716643092880716,
      "grad_norm": 0.0008820195798762143,
      "learning_rate": 0.11809532294543279,
      "loss": 0.2039,
      "num_input_tokens_seen": 19683920,
      "step": 22730
    },
    {
      "epoch": 10.719000471475718,
      "grad_norm": 0.0011103474535048008,
      "learning_rate": 0.11803776840456245,
      "loss": 0.1979,
      "num_input_tokens_seen": 19689360,
      "step": 22735
    },
    {
      "epoch": 10.72135785007072,
      "grad_norm": 0.001783317537046969,
      "learning_rate": 0.11798021879266997,
      "loss": 0.2301,
      "num_input_tokens_seen": 19693200,
      "step": 22740
    },
    {
      "epoch": 10.723715228665723,
      "grad_norm": 0.0015118020819500089,
      "learning_rate": 0.11792267411863006,
      "loss": 0.1766,
      "num_input_tokens_seen": 19697216,
      "step": 22745
    },
    {
      "epoch": 10.726072607260726,
      "grad_norm": 0.0018448153277859092,
      "learning_rate": 0.1178651343913169,
      "loss": 0.2207,
      "num_input_tokens_seen": 19701952,
      "step": 22750
    },
    {
      "epoch": 10.728429985855728,
      "grad_norm": 0.0021835092920809984,
      "learning_rate": 0.11780759961960392,
      "loss": 0.2509,
      "num_input_tokens_seen": 19706176,
      "step": 22755
    },
    {
      "epoch": 10.73078736445073,
      "grad_norm": 0.0009327690349891782,
      "learning_rate": 0.1177500698123636,
      "loss": 0.2119,
      "num_input_tokens_seen": 19710400,
      "step": 22760
    },
    {
      "epoch": 10.733144743045733,
      "grad_norm": 0.001328403246589005,
      "learning_rate": 0.11769254497846778,
      "loss": 0.2537,
      "num_input_tokens_seen": 19714832,
      "step": 22765
    },
    {
      "epoch": 10.735502121640735,
      "grad_norm": 0.0008197848801501095,
      "learning_rate": 0.11763502512678758,
      "loss": 0.2328,
      "num_input_tokens_seen": 19718592,
      "step": 22770
    },
    {
      "epoch": 10.737859500235738,
      "grad_norm": 0.0009917898569256067,
      "learning_rate": 0.11757751026619315,
      "loss": 0.2197,
      "num_input_tokens_seen": 19722464,
      "step": 22775
    },
    {
      "epoch": 10.74021687883074,
      "grad_norm": 0.0021469490602612495,
      "learning_rate": 0.11752000040555416,
      "loss": 0.2285,
      "num_input_tokens_seen": 19727200,
      "step": 22780
    },
    {
      "epoch": 10.742574257425742,
      "grad_norm": 0.0010109434369951487,
      "learning_rate": 0.11746249555373921,
      "loss": 0.2037,
      "num_input_tokens_seen": 19732000,
      "step": 22785
    },
    {
      "epoch": 10.744931636020745,
      "grad_norm": 0.0012190317502245307,
      "learning_rate": 0.11740499571961638,
      "loss": 0.1956,
      "num_input_tokens_seen": 19735808,
      "step": 22790
    },
    {
      "epoch": 10.747289014615747,
      "grad_norm": 0.0009148928220383823,
      "learning_rate": 0.11734750091205279,
      "loss": 0.2184,
      "num_input_tokens_seen": 19740624,
      "step": 22795
    },
    {
      "epoch": 10.74964639321075,
      "grad_norm": 0.0012424859451130033,
      "learning_rate": 0.11729001113991493,
      "loss": 0.2331,
      "num_input_tokens_seen": 19744496,
      "step": 22800
    },
    {
      "epoch": 10.74964639321075,
      "eval_loss": 0.21467708051204681,
      "eval_runtime": 21.9515,
      "eval_samples_per_second": 42.958,
      "eval_steps_per_second": 21.502,
      "num_input_tokens_seen": 19744496,
      "step": 22800
    },
    {
      "epoch": 10.752003771805752,
      "grad_norm": 0.0007923890952952206,
      "learning_rate": 0.11723252641206837,
      "loss": 0.2158,
      "num_input_tokens_seen": 19748960,
      "step": 22805
    },
    {
      "epoch": 10.754361150400754,
      "grad_norm": 0.0008607939817011356,
      "learning_rate": 0.11717504673737808,
      "loss": 0.2293,
      "num_input_tokens_seen": 19752928,
      "step": 22810
    },
    {
      "epoch": 10.756718528995757,
      "grad_norm": 0.0013919376069679856,
      "learning_rate": 0.11711757212470802,
      "loss": 0.2137,
      "num_input_tokens_seen": 19756784,
      "step": 22815
    },
    {
      "epoch": 10.75907590759076,
      "grad_norm": 0.0011429657461121678,
      "learning_rate": 0.11706010258292165,
      "loss": 0.2,
      "num_input_tokens_seen": 19761248,
      "step": 22820
    },
    {
      "epoch": 10.761433286185762,
      "grad_norm": 0.000871662050485611,
      "learning_rate": 0.11700263812088131,
      "loss": 0.1905,
      "num_input_tokens_seen": 19764736,
      "step": 22825
    },
    {
      "epoch": 10.763790664780764,
      "grad_norm": 0.0010070183780044317,
      "learning_rate": 0.11694517874744892,
      "loss": 0.1878,
      "num_input_tokens_seen": 19768704,
      "step": 22830
    },
    {
      "epoch": 10.766148043375766,
      "grad_norm": 0.002095734467729926,
      "learning_rate": 0.11688772447148532,
      "loss": 0.2114,
      "num_input_tokens_seen": 19773488,
      "step": 22835
    },
    {
      "epoch": 10.768505421970769,
      "grad_norm": 0.0010858295718207955,
      "learning_rate": 0.11683027530185074,
      "loss": 0.1957,
      "num_input_tokens_seen": 19777360,
      "step": 22840
    },
    {
      "epoch": 10.770862800565771,
      "grad_norm": 0.0021071031223982573,
      "learning_rate": 0.11677283124740451,
      "loss": 0.2217,
      "num_input_tokens_seen": 19781904,
      "step": 22845
    },
    {
      "epoch": 10.773220179160774,
      "grad_norm": 0.0010490230051800609,
      "learning_rate": 0.11671539231700531,
      "loss": 0.2093,
      "num_input_tokens_seen": 19785472,
      "step": 22850
    },
    {
      "epoch": 10.775577557755776,
      "grad_norm": 0.0016758970450609922,
      "learning_rate": 0.11665795851951084,
      "loss": 0.1718,
      "num_input_tokens_seen": 19790160,
      "step": 22855
    },
    {
      "epoch": 10.777934936350778,
      "grad_norm": 0.0012555245775729418,
      "learning_rate": 0.11660052986377825,
      "loss": 0.1764,
      "num_input_tokens_seen": 19794448,
      "step": 22860
    },
    {
      "epoch": 10.78029231494578,
      "grad_norm": 0.0008648642105981708,
      "learning_rate": 0.1165431063586636,
      "loss": 0.2643,
      "num_input_tokens_seen": 19798560,
      "step": 22865
    },
    {
      "epoch": 10.782649693540783,
      "grad_norm": 0.002352221170440316,
      "learning_rate": 0.11648568801302245,
      "loss": 0.2543,
      "num_input_tokens_seen": 19803056,
      "step": 22870
    },
    {
      "epoch": 10.785007072135786,
      "grad_norm": 0.001754462136887014,
      "learning_rate": 0.11642827483570937,
      "loss": 0.1859,
      "num_input_tokens_seen": 19807056,
      "step": 22875
    },
    {
      "epoch": 10.787364450730788,
      "grad_norm": 0.0010804458288475871,
      "learning_rate": 0.11637086683557815,
      "loss": 0.2188,
      "num_input_tokens_seen": 19811808,
      "step": 22880
    },
    {
      "epoch": 10.78972182932579,
      "grad_norm": 0.0011585332686081529,
      "learning_rate": 0.11631346402148188,
      "loss": 0.2301,
      "num_input_tokens_seen": 19816032,
      "step": 22885
    },
    {
      "epoch": 10.792079207920793,
      "grad_norm": 0.001051040948368609,
      "learning_rate": 0.11625606640227285,
      "loss": 0.2388,
      "num_input_tokens_seen": 19820112,
      "step": 22890
    },
    {
      "epoch": 10.794436586515795,
      "grad_norm": 0.0015964859630912542,
      "learning_rate": 0.11619867398680238,
      "loss": 0.2238,
      "num_input_tokens_seen": 19824656,
      "step": 22895
    },
    {
      "epoch": 10.796793965110798,
      "grad_norm": 0.001063556526787579,
      "learning_rate": 0.11614128678392119,
      "loss": 0.2235,
      "num_input_tokens_seen": 19828832,
      "step": 22900
    },
    {
      "epoch": 10.799151343705798,
      "grad_norm": 0.0009501239401288331,
      "learning_rate": 0.11608390480247906,
      "loss": 0.2216,
      "num_input_tokens_seen": 19833072,
      "step": 22905
    },
    {
      "epoch": 10.801508722300802,
      "grad_norm": 0.0031957945320755243,
      "learning_rate": 0.11602652805132499,
      "loss": 0.2272,
      "num_input_tokens_seen": 19836976,
      "step": 22910
    },
    {
      "epoch": 10.803866100895803,
      "grad_norm": 0.0017655331175774336,
      "learning_rate": 0.11596915653930731,
      "loss": 0.2241,
      "num_input_tokens_seen": 19841248,
      "step": 22915
    },
    {
      "epoch": 10.806223479490805,
      "grad_norm": 0.0008380957879126072,
      "learning_rate": 0.11591179027527328,
      "loss": 0.2009,
      "num_input_tokens_seen": 19845728,
      "step": 22920
    },
    {
      "epoch": 10.808580858085808,
      "grad_norm": 0.0007886758539825678,
      "learning_rate": 0.11585442926806956,
      "loss": 0.2118,
      "num_input_tokens_seen": 19850640,
      "step": 22925
    },
    {
      "epoch": 10.81093823668081,
      "grad_norm": 0.002840352011844516,
      "learning_rate": 0.11579707352654202,
      "loss": 0.2117,
      "num_input_tokens_seen": 19857152,
      "step": 22930
    },
    {
      "epoch": 10.813295615275813,
      "grad_norm": 0.0007142882095649838,
      "learning_rate": 0.11573972305953548,
      "loss": 0.2228,
      "num_input_tokens_seen": 19860704,
      "step": 22935
    },
    {
      "epoch": 10.815652993870815,
      "grad_norm": 0.0013685787562280893,
      "learning_rate": 0.11568237787589426,
      "loss": 0.1931,
      "num_input_tokens_seen": 19865360,
      "step": 22940
    },
    {
      "epoch": 10.818010372465817,
      "grad_norm": 0.0014829769497737288,
      "learning_rate": 0.11562503798446161,
      "loss": 0.2534,
      "num_input_tokens_seen": 19870048,
      "step": 22945
    },
    {
      "epoch": 10.82036775106082,
      "grad_norm": 0.0007740053115412593,
      "learning_rate": 0.11556770339408005,
      "loss": 0.1968,
      "num_input_tokens_seen": 19874336,
      "step": 22950
    },
    {
      "epoch": 10.822725129655822,
      "grad_norm": 0.0013344844337552786,
      "learning_rate": 0.1155103741135914,
      "loss": 0.2129,
      "num_input_tokens_seen": 19877984,
      "step": 22955
    },
    {
      "epoch": 10.825082508250825,
      "grad_norm": 0.0037793894298374653,
      "learning_rate": 0.1154530501518364,
      "loss": 0.2715,
      "num_input_tokens_seen": 19881744,
      "step": 22960
    },
    {
      "epoch": 10.827439886845827,
      "grad_norm": 0.0007798086735419929,
      "learning_rate": 0.11539573151765523,
      "loss": 0.21,
      "num_input_tokens_seen": 19885920,
      "step": 22965
    },
    {
      "epoch": 10.82979726544083,
      "grad_norm": 0.0008358924533240497,
      "learning_rate": 0.11533841821988719,
      "loss": 0.1901,
      "num_input_tokens_seen": 19889456,
      "step": 22970
    },
    {
      "epoch": 10.832154644035832,
      "grad_norm": 0.0007854973082430661,
      "learning_rate": 0.11528111026737059,
      "loss": 0.1904,
      "num_input_tokens_seen": 19893696,
      "step": 22975
    },
    {
      "epoch": 10.834512022630834,
      "grad_norm": 0.0007651916239410639,
      "learning_rate": 0.11522380766894312,
      "loss": 0.2183,
      "num_input_tokens_seen": 19898064,
      "step": 22980
    },
    {
      "epoch": 10.836869401225837,
      "grad_norm": 0.0009095256100408733,
      "learning_rate": 0.11516651043344152,
      "loss": 0.2205,
      "num_input_tokens_seen": 19902480,
      "step": 22985
    },
    {
      "epoch": 10.839226779820839,
      "grad_norm": 0.0009304385748691857,
      "learning_rate": 0.11510921856970172,
      "loss": 0.2044,
      "num_input_tokens_seen": 19907008,
      "step": 22990
    },
    {
      "epoch": 10.841584158415841,
      "grad_norm": 0.0012064108159393072,
      "learning_rate": 0.11505193208655895,
      "loss": 0.2069,
      "num_input_tokens_seen": 19911696,
      "step": 22995
    },
    {
      "epoch": 10.843941537010844,
      "grad_norm": 0.0007563559338450432,
      "learning_rate": 0.11499465099284738,
      "loss": 0.2107,
      "num_input_tokens_seen": 19915984,
      "step": 23000
    },
    {
      "epoch": 10.843941537010844,
      "eval_loss": 0.2145698070526123,
      "eval_runtime": 21.8986,
      "eval_samples_per_second": 43.062,
      "eval_steps_per_second": 21.554,
      "num_input_tokens_seen": 19915984,
      "step": 23000
    },
    {
      "epoch": 10.846298915605846,
      "grad_norm": 0.0013370446395128965,
      "learning_rate": 0.1149373752974006,
      "loss": 0.1957,
      "num_input_tokens_seen": 19920032,
      "step": 23005
    },
    {
      "epoch": 10.848656294200849,
      "grad_norm": 0.0019378806464374065,
      "learning_rate": 0.11488010500905109,
      "loss": 0.2459,
      "num_input_tokens_seen": 19925024,
      "step": 23010
    },
    {
      "epoch": 10.851013672795851,
      "grad_norm": 0.0009166030213236809,
      "learning_rate": 0.11482284013663077,
      "loss": 0.2605,
      "num_input_tokens_seen": 19929424,
      "step": 23015
    },
    {
      "epoch": 10.853371051390853,
      "grad_norm": 0.001316463341936469,
      "learning_rate": 0.11476558068897061,
      "loss": 0.2484,
      "num_input_tokens_seen": 19934048,
      "step": 23020
    },
    {
      "epoch": 10.855728429985856,
      "grad_norm": 0.0007123580435290933,
      "learning_rate": 0.11470832667490061,
      "loss": 0.2279,
      "num_input_tokens_seen": 19937968,
      "step": 23025
    },
    {
      "epoch": 10.858085808580858,
      "grad_norm": 0.001273282221518457,
      "learning_rate": 0.11465107810325013,
      "loss": 0.2317,
      "num_input_tokens_seen": 19942416,
      "step": 23030
    },
    {
      "epoch": 10.86044318717586,
      "grad_norm": 0.0008236832218244672,
      "learning_rate": 0.11459383498284771,
      "loss": 0.2384,
      "num_input_tokens_seen": 19946688,
      "step": 23035
    },
    {
      "epoch": 10.862800565770863,
      "grad_norm": 0.0007184587884694338,
      "learning_rate": 0.11453659732252082,
      "loss": 0.2166,
      "num_input_tokens_seen": 19950592,
      "step": 23040
    },
    {
      "epoch": 10.865157944365865,
      "grad_norm": 0.0021585498470813036,
      "learning_rate": 0.11447936513109633,
      "loss": 0.2505,
      "num_input_tokens_seen": 19955296,
      "step": 23045
    },
    {
      "epoch": 10.867515322960868,
      "grad_norm": 0.0012633205624297261,
      "learning_rate": 0.11442213841740011,
      "loss": 0.2297,
      "num_input_tokens_seen": 19959328,
      "step": 23050
    },
    {
      "epoch": 10.86987270155587,
      "grad_norm": 0.000999503885395825,
      "learning_rate": 0.1143649171902572,
      "loss": 0.2343,
      "num_input_tokens_seen": 19963296,
      "step": 23055
    },
    {
      "epoch": 10.872230080150873,
      "grad_norm": 0.0007588480366393924,
      "learning_rate": 0.11430770145849194,
      "loss": 0.2019,
      "num_input_tokens_seen": 19967520,
      "step": 23060
    },
    {
      "epoch": 10.874587458745875,
      "grad_norm": 0.0006340722902677953,
      "learning_rate": 0.11425049123092756,
      "loss": 0.195,
      "num_input_tokens_seen": 19972416,
      "step": 23065
    },
    {
      "epoch": 10.876944837340877,
      "grad_norm": 0.0006020409637130797,
      "learning_rate": 0.11419328651638674,
      "loss": 0.2181,
      "num_input_tokens_seen": 19977296,
      "step": 23070
    },
    {
      "epoch": 10.87930221593588,
      "grad_norm": 0.0011248351074755192,
      "learning_rate": 0.11413608732369115,
      "loss": 0.1815,
      "num_input_tokens_seen": 19981968,
      "step": 23075
    },
    {
      "epoch": 10.881659594530882,
      "grad_norm": 0.0012005064636468887,
      "learning_rate": 0.11407889366166153,
      "loss": 0.2181,
      "num_input_tokens_seen": 19986864,
      "step": 23080
    },
    {
      "epoch": 10.884016973125885,
      "grad_norm": 0.001299140159972012,
      "learning_rate": 0.11402170553911797,
      "loss": 0.2117,
      "num_input_tokens_seen": 19990992,
      "step": 23085
    },
    {
      "epoch": 10.886374351720887,
      "grad_norm": 0.0006875817780382931,
      "learning_rate": 0.11396452296487955,
      "loss": 0.2289,
      "num_input_tokens_seen": 19995280,
      "step": 23090
    },
    {
      "epoch": 10.88873173031589,
      "grad_norm": 0.0007676462410017848,
      "learning_rate": 0.11390734594776449,
      "loss": 0.2241,
      "num_input_tokens_seen": 19999376,
      "step": 23095
    },
    {
      "epoch": 10.891089108910892,
      "grad_norm": 0.0010512993903830647,
      "learning_rate": 0.11385017449659031,
      "loss": 0.2047,
      "num_input_tokens_seen": 20003072,
      "step": 23100
    },
    {
      "epoch": 10.893446487505894,
      "grad_norm": 0.0011791749857366085,
      "learning_rate": 0.11379300862017344,
      "loss": 0.23,
      "num_input_tokens_seen": 20007936,
      "step": 23105
    },
    {
      "epoch": 10.895803866100895,
      "grad_norm": 0.0011103141587227583,
      "learning_rate": 0.11373584832732966,
      "loss": 0.2031,
      "num_input_tokens_seen": 20012464,
      "step": 23110
    },
    {
      "epoch": 10.898161244695899,
      "grad_norm": 0.0008360130595974624,
      "learning_rate": 0.11367869362687386,
      "loss": 0.2238,
      "num_input_tokens_seen": 20016064,
      "step": 23115
    },
    {
      "epoch": 10.9005186232909,
      "grad_norm": 0.0010881806956604123,
      "learning_rate": 0.11362154452761988,
      "loss": 0.161,
      "num_input_tokens_seen": 20020288,
      "step": 23120
    },
    {
      "epoch": 10.902876001885902,
      "grad_norm": 0.0016385485650971532,
      "learning_rate": 0.11356440103838095,
      "loss": 0.2146,
      "num_input_tokens_seen": 20024512,
      "step": 23125
    },
    {
      "epoch": 10.905233380480905,
      "grad_norm": 0.0028529844712466,
      "learning_rate": 0.11350726316796922,
      "loss": 0.2454,
      "num_input_tokens_seen": 20028768,
      "step": 23130
    },
    {
      "epoch": 10.907590759075907,
      "grad_norm": 0.0012602547649294138,
      "learning_rate": 0.11345013092519607,
      "loss": 0.1702,
      "num_input_tokens_seen": 20033248,
      "step": 23135
    },
    {
      "epoch": 10.90994813767091,
      "grad_norm": 0.0017458338988944888,
      "learning_rate": 0.11339300431887213,
      "loss": 0.2451,
      "num_input_tokens_seen": 20037920,
      "step": 23140
    },
    {
      "epoch": 10.912305516265912,
      "grad_norm": 0.0006550296675413847,
      "learning_rate": 0.11333588335780687,
      "loss": 0.2476,
      "num_input_tokens_seen": 20044048,
      "step": 23145
    },
    {
      "epoch": 10.914662894860914,
      "grad_norm": 0.0010822601616382599,
      "learning_rate": 0.11327876805080916,
      "loss": 0.1903,
      "num_input_tokens_seen": 20048096,
      "step": 23150
    },
    {
      "epoch": 10.917020273455917,
      "grad_norm": 0.0012574298307299614,
      "learning_rate": 0.11322165840668696,
      "loss": 0.2167,
      "num_input_tokens_seen": 20052576,
      "step": 23155
    },
    {
      "epoch": 10.919377652050919,
      "grad_norm": 0.001156458049081266,
      "learning_rate": 0.11316455443424717,
      "loss": 0.2,
      "num_input_tokens_seen": 20056496,
      "step": 23160
    },
    {
      "epoch": 10.921735030645921,
      "grad_norm": 0.0017996854148805141,
      "learning_rate": 0.11310745614229603,
      "loss": 0.2538,
      "num_input_tokens_seen": 20060992,
      "step": 23165
    },
    {
      "epoch": 10.924092409240924,
      "grad_norm": 0.0007214352954179049,
      "learning_rate": 0.1130503635396387,
      "loss": 0.2464,
      "num_input_tokens_seen": 20065104,
      "step": 23170
    },
    {
      "epoch": 10.926449787835926,
      "grad_norm": 0.0012655975297093391,
      "learning_rate": 0.11299327663507966,
      "loss": 0.1832,
      "num_input_tokens_seen": 20069200,
      "step": 23175
    },
    {
      "epoch": 10.928807166430929,
      "grad_norm": 0.0008490518084727228,
      "learning_rate": 0.11293619543742246,
      "loss": 0.223,
      "num_input_tokens_seen": 20073232,
      "step": 23180
    },
    {
      "epoch": 10.931164545025931,
      "grad_norm": 0.0014907806180417538,
      "learning_rate": 0.11287911995546965,
      "loss": 0.2093,
      "num_input_tokens_seen": 20077392,
      "step": 23185
    },
    {
      "epoch": 10.933521923620933,
      "grad_norm": 0.0016722565051168203,
      "learning_rate": 0.11282205019802308,
      "loss": 0.2216,
      "num_input_tokens_seen": 20082448,
      "step": 23190
    },
    {
      "epoch": 10.935879302215936,
      "grad_norm": 0.0009421691065654159,
      "learning_rate": 0.11276498617388354,
      "loss": 0.1866,
      "num_input_tokens_seen": 20086544,
      "step": 23195
    },
    {
      "epoch": 10.938236680810938,
      "grad_norm": 0.00103525270242244,
      "learning_rate": 0.11270792789185109,
      "loss": 0.2288,
      "num_input_tokens_seen": 20090944,
      "step": 23200
    },
    {
      "epoch": 10.938236680810938,
      "eval_loss": 0.21472252905368805,
      "eval_runtime": 21.9349,
      "eval_samples_per_second": 42.991,
      "eval_steps_per_second": 21.518,
      "num_input_tokens_seen": 20090944,
      "step": 23200
    },
    {
      "epoch": 10.94059405940594,
      "grad_norm": 0.0008417396456934512,
      "learning_rate": 0.11265087536072482,
      "loss": 0.1847,
      "num_input_tokens_seen": 20095760,
      "step": 23205
    },
    {
      "epoch": 10.942951438000943,
      "grad_norm": 0.000897797173820436,
      "learning_rate": 0.11259382858930288,
      "loss": 0.2104,
      "num_input_tokens_seen": 20100384,
      "step": 23210
    },
    {
      "epoch": 10.945308816595945,
      "grad_norm": 0.0014295177534222603,
      "learning_rate": 0.11253678758638262,
      "loss": 0.2312,
      "num_input_tokens_seen": 20104880,
      "step": 23215
    },
    {
      "epoch": 10.947666195190948,
      "grad_norm": 0.0010773816611617804,
      "learning_rate": 0.11247975236076059,
      "loss": 0.1856,
      "num_input_tokens_seen": 20109648,
      "step": 23220
    },
    {
      "epoch": 10.95002357378595,
      "grad_norm": 0.0008797547779977322,
      "learning_rate": 0.11242272292123218,
      "loss": 0.2675,
      "num_input_tokens_seen": 20113456,
      "step": 23225
    },
    {
      "epoch": 10.952380952380953,
      "grad_norm": 0.0007080081850290298,
      "learning_rate": 0.11236569927659217,
      "loss": 0.2226,
      "num_input_tokens_seen": 20118160,
      "step": 23230
    },
    {
      "epoch": 10.954738330975955,
      "grad_norm": 0.0018796961521729827,
      "learning_rate": 0.11230868143563429,
      "loss": 0.2277,
      "num_input_tokens_seen": 20122384,
      "step": 23235
    },
    {
      "epoch": 10.957095709570957,
      "grad_norm": 0.0016922520007938147,
      "learning_rate": 0.11225166940715131,
      "loss": 0.2239,
      "num_input_tokens_seen": 20125760,
      "step": 23240
    },
    {
      "epoch": 10.95945308816596,
      "grad_norm": 0.0015863478183746338,
      "learning_rate": 0.11219466319993537,
      "loss": 0.2124,
      "num_input_tokens_seen": 20130048,
      "step": 23245
    },
    {
      "epoch": 10.961810466760962,
      "grad_norm": 0.0010020764311775565,
      "learning_rate": 0.11213766282277739,
      "loss": 0.2152,
      "num_input_tokens_seen": 20134480,
      "step": 23250
    },
    {
      "epoch": 10.964167845355965,
      "grad_norm": 0.001910592196509242,
      "learning_rate": 0.11208066828446761,
      "loss": 0.2261,
      "num_input_tokens_seen": 20139264,
      "step": 23255
    },
    {
      "epoch": 10.966525223950967,
      "grad_norm": 0.0014729124959558249,
      "learning_rate": 0.11202367959379537,
      "loss": 0.2272,
      "num_input_tokens_seen": 20143312,
      "step": 23260
    },
    {
      "epoch": 10.96888260254597,
      "grad_norm": 0.0012148782843723893,
      "learning_rate": 0.11196669675954894,
      "loss": 0.2116,
      "num_input_tokens_seen": 20147840,
      "step": 23265
    },
    {
      "epoch": 10.971239981140972,
      "grad_norm": 0.0009844157611951232,
      "learning_rate": 0.1119097197905158,
      "loss": 0.1816,
      "num_input_tokens_seen": 20151856,
      "step": 23270
    },
    {
      "epoch": 10.973597359735974,
      "grad_norm": 0.0010323496535420418,
      "learning_rate": 0.11185274869548259,
      "loss": 0.2437,
      "num_input_tokens_seen": 20155424,
      "step": 23275
    },
    {
      "epoch": 10.975954738330977,
      "grad_norm": 0.0008316069142892957,
      "learning_rate": 0.11179578348323486,
      "loss": 0.186,
      "num_input_tokens_seen": 20159472,
      "step": 23280
    },
    {
      "epoch": 10.978312116925979,
      "grad_norm": 0.0007583939586766064,
      "learning_rate": 0.1117388241625575,
      "loss": 0.2146,
      "num_input_tokens_seen": 20163488,
      "step": 23285
    },
    {
      "epoch": 10.980669495520981,
      "grad_norm": 0.0019978827331215143,
      "learning_rate": 0.11168187074223421,
      "loss": 0.2566,
      "num_input_tokens_seen": 20168320,
      "step": 23290
    },
    {
      "epoch": 10.983026874115984,
      "grad_norm": 0.001236087060533464,
      "learning_rate": 0.11162492323104796,
      "loss": 0.1884,
      "num_input_tokens_seen": 20172992,
      "step": 23295
    },
    {
      "epoch": 10.985384252710986,
      "grad_norm": 0.0016348578501492739,
      "learning_rate": 0.11156798163778091,
      "loss": 0.2398,
      "num_input_tokens_seen": 20177136,
      "step": 23300
    },
    {
      "epoch": 10.987741631305987,
      "grad_norm": 0.0020291681867092848,
      "learning_rate": 0.11151104597121399,
      "loss": 0.1902,
      "num_input_tokens_seen": 20181984,
      "step": 23305
    },
    {
      "epoch": 10.990099009900991,
      "grad_norm": 0.0017043852712959051,
      "learning_rate": 0.11145411624012742,
      "loss": 0.2123,
      "num_input_tokens_seen": 20185840,
      "step": 23310
    },
    {
      "epoch": 10.992456388495992,
      "grad_norm": 0.0011680348543450236,
      "learning_rate": 0.11139719245330063,
      "loss": 0.1776,
      "num_input_tokens_seen": 20191200,
      "step": 23315
    },
    {
      "epoch": 10.994813767090994,
      "grad_norm": 0.001845953636802733,
      "learning_rate": 0.11134027461951179,
      "loss": 0.2404,
      "num_input_tokens_seen": 20195264,
      "step": 23320
    },
    {
      "epoch": 10.997171145685996,
      "grad_norm": 0.0021170955151319504,
      "learning_rate": 0.11128336274753849,
      "loss": 0.2096,
      "num_input_tokens_seen": 20199808,
      "step": 23325
    },
    {
      "epoch": 10.999528524280999,
      "grad_norm": 0.00138994085136801,
      "learning_rate": 0.11122645684615715,
      "loss": 0.2267,
      "num_input_tokens_seen": 20203920,
      "step": 23330
    },
    {
      "epoch": 11.001885902876001,
      "grad_norm": 0.000994442729279399,
      "learning_rate": 0.11116955692414345,
      "loss": 0.1809,
      "num_input_tokens_seen": 20207728,
      "step": 23335
    },
    {
      "epoch": 11.004243281471004,
      "grad_norm": 0.0008911281474865973,
      "learning_rate": 0.11111266299027203,
      "loss": 0.2033,
      "num_input_tokens_seen": 20212592,
      "step": 23340
    },
    {
      "epoch": 11.006600660066006,
      "grad_norm": 0.0010320983128622174,
      "learning_rate": 0.11105577505331668,
      "loss": 0.2248,
      "num_input_tokens_seen": 20217056,
      "step": 23345
    },
    {
      "epoch": 11.008958038661008,
      "grad_norm": 0.0019610875751823187,
      "learning_rate": 0.11099889312205018,
      "loss": 0.2319,
      "num_input_tokens_seen": 20220848,
      "step": 23350
    },
    {
      "epoch": 11.01131541725601,
      "grad_norm": 0.0010735843097791076,
      "learning_rate": 0.11094201720524455,
      "loss": 0.211,
      "num_input_tokens_seen": 20225360,
      "step": 23355
    },
    {
      "epoch": 11.013672795851013,
      "grad_norm": 0.0024423038121312857,
      "learning_rate": 0.11088514731167064,
      "loss": 0.2235,
      "num_input_tokens_seen": 20230480,
      "step": 23360
    },
    {
      "epoch": 11.016030174446016,
      "grad_norm": 0.00119048822671175,
      "learning_rate": 0.11082828345009862,
      "loss": 0.222,
      "num_input_tokens_seen": 20234112,
      "step": 23365
    },
    {
      "epoch": 11.018387553041018,
      "grad_norm": 0.003125028917565942,
      "learning_rate": 0.11077142562929748,
      "loss": 0.2144,
      "num_input_tokens_seen": 20238320,
      "step": 23370
    },
    {
      "epoch": 11.02074493163602,
      "grad_norm": 0.0030460748821496964,
      "learning_rate": 0.11071457385803554,
      "loss": 0.2161,
      "num_input_tokens_seen": 20242592,
      "step": 23375
    },
    {
      "epoch": 11.023102310231023,
      "grad_norm": 0.002021255437284708,
      "learning_rate": 0.11065772814508001,
      "loss": 0.2318,
      "num_input_tokens_seen": 20246912,
      "step": 23380
    },
    {
      "epoch": 11.025459688826025,
      "grad_norm": 0.0014620937872678041,
      "learning_rate": 0.11060088849919715,
      "loss": 0.2086,
      "num_input_tokens_seen": 20251216,
      "step": 23385
    },
    {
      "epoch": 11.027817067421028,
      "grad_norm": 0.0012292565079405904,
      "learning_rate": 0.11054405492915244,
      "loss": 0.1968,
      "num_input_tokens_seen": 20256000,
      "step": 23390
    },
    {
      "epoch": 11.03017444601603,
      "grad_norm": 0.002381534082815051,
      "learning_rate": 0.11048722744371031,
      "loss": 0.226,
      "num_input_tokens_seen": 20259760,
      "step": 23395
    },
    {
      "epoch": 11.032531824611032,
      "grad_norm": 0.0018527020001783967,
      "learning_rate": 0.1104304060516342,
      "loss": 0.2446,
      "num_input_tokens_seen": 20264992,
      "step": 23400
    },
    {
      "epoch": 11.032531824611032,
      "eval_loss": 0.21772626042366028,
      "eval_runtime": 21.9538,
      "eval_samples_per_second": 42.954,
      "eval_steps_per_second": 21.5,
      "num_input_tokens_seen": 20264992,
      "step": 23400
    },
    {
      "epoch": 11.034889203206035,
      "grad_norm": 0.0011606266489252448,
      "learning_rate": 0.11037359076168682,
      "loss": 0.2531,
      "num_input_tokens_seen": 20268720,
      "step": 23405
    },
    {
      "epoch": 11.037246581801037,
      "grad_norm": 0.0011548015754669905,
      "learning_rate": 0.11031678158262966,
      "loss": 0.1999,
      "num_input_tokens_seen": 20273072,
      "step": 23410
    },
    {
      "epoch": 11.03960396039604,
      "grad_norm": 0.000966964231338352,
      "learning_rate": 0.11025997852322349,
      "loss": 0.2065,
      "num_input_tokens_seen": 20276864,
      "step": 23415
    },
    {
      "epoch": 11.041961338991042,
      "grad_norm": 0.0014709875686094165,
      "learning_rate": 0.11020318159222807,
      "loss": 0.2297,
      "num_input_tokens_seen": 20281168,
      "step": 23420
    },
    {
      "epoch": 11.044318717586044,
      "grad_norm": 0.0031000811140984297,
      "learning_rate": 0.1101463907984021,
      "loss": 0.1868,
      "num_input_tokens_seen": 20285664,
      "step": 23425
    },
    {
      "epoch": 11.046676096181047,
      "grad_norm": 0.0011572479270398617,
      "learning_rate": 0.11008960615050352,
      "loss": 0.2329,
      "num_input_tokens_seen": 20289680,
      "step": 23430
    },
    {
      "epoch": 11.04903347477605,
      "grad_norm": 0.00136491097509861,
      "learning_rate": 0.11003282765728925,
      "loss": 0.2089,
      "num_input_tokens_seen": 20293824,
      "step": 23435
    },
    {
      "epoch": 11.051390853371052,
      "grad_norm": 0.0012604534858837724,
      "learning_rate": 0.10997605532751518,
      "loss": 0.1942,
      "num_input_tokens_seen": 20298160,
      "step": 23440
    },
    {
      "epoch": 11.053748231966054,
      "grad_norm": 0.000972271547652781,
      "learning_rate": 0.1099192891699364,
      "loss": 0.2104,
      "num_input_tokens_seen": 20302096,
      "step": 23445
    },
    {
      "epoch": 11.056105610561056,
      "grad_norm": 0.0013990378938615322,
      "learning_rate": 0.10986252919330687,
      "loss": 0.2185,
      "num_input_tokens_seen": 20306432,
      "step": 23450
    },
    {
      "epoch": 11.058462989156059,
      "grad_norm": 0.0010516188340261579,
      "learning_rate": 0.10980577540637973,
      "loss": 0.1974,
      "num_input_tokens_seen": 20310736,
      "step": 23455
    },
    {
      "epoch": 11.060820367751061,
      "grad_norm": 0.00156517478171736,
      "learning_rate": 0.10974902781790719,
      "loss": 0.2622,
      "num_input_tokens_seen": 20315904,
      "step": 23460
    },
    {
      "epoch": 11.063177746346064,
      "grad_norm": 0.0012299699010327458,
      "learning_rate": 0.10969228643664032,
      "loss": 0.2208,
      "num_input_tokens_seen": 20319680,
      "step": 23465
    },
    {
      "epoch": 11.065535124941066,
      "grad_norm": 0.0019358410499989986,
      "learning_rate": 0.10963555127132942,
      "loss": 0.2119,
      "num_input_tokens_seen": 20323760,
      "step": 23470
    },
    {
      "epoch": 11.067892503536068,
      "grad_norm": 0.002813896397128701,
      "learning_rate": 0.10957882233072382,
      "loss": 0.2255,
      "num_input_tokens_seen": 20328608,
      "step": 23475
    },
    {
      "epoch": 11.07024988213107,
      "grad_norm": 0.002009752206504345,
      "learning_rate": 0.10952209962357176,
      "loss": 0.223,
      "num_input_tokens_seen": 20333280,
      "step": 23480
    },
    {
      "epoch": 11.072607260726073,
      "grad_norm": 0.0022760089486837387,
      "learning_rate": 0.10946538315862062,
      "loss": 0.2267,
      "num_input_tokens_seen": 20337568,
      "step": 23485
    },
    {
      "epoch": 11.074964639321076,
      "grad_norm": 0.0013420458417385817,
      "learning_rate": 0.10940867294461679,
      "loss": 0.2153,
      "num_input_tokens_seen": 20341600,
      "step": 23490
    },
    {
      "epoch": 11.077322017916078,
      "grad_norm": 0.0016229223692789674,
      "learning_rate": 0.10935196899030565,
      "loss": 0.1974,
      "num_input_tokens_seen": 20345696,
      "step": 23495
    },
    {
      "epoch": 11.07967939651108,
      "grad_norm": 0.0014904115814715624,
      "learning_rate": 0.10929527130443177,
      "loss": 0.2541,
      "num_input_tokens_seen": 20350384,
      "step": 23500
    },
    {
      "epoch": 11.082036775106083,
      "grad_norm": 0.0012175634037703276,
      "learning_rate": 0.1092385798957385,
      "loss": 0.2389,
      "num_input_tokens_seen": 20354224,
      "step": 23505
    },
    {
      "epoch": 11.084394153701085,
      "grad_norm": 0.0015207205433398485,
      "learning_rate": 0.10918189477296848,
      "loss": 0.2279,
      "num_input_tokens_seen": 20358384,
      "step": 23510
    },
    {
      "epoch": 11.086751532296086,
      "grad_norm": 0.0023589455522596836,
      "learning_rate": 0.1091252159448633,
      "loss": 0.2215,
      "num_input_tokens_seen": 20362944,
      "step": 23515
    },
    {
      "epoch": 11.089108910891088,
      "grad_norm": 0.001307054772041738,
      "learning_rate": 0.10906854342016345,
      "loss": 0.2264,
      "num_input_tokens_seen": 20367584,
      "step": 23520
    },
    {
      "epoch": 11.09146628948609,
      "grad_norm": 0.0009953967528417706,
      "learning_rate": 0.10901187720760858,
      "loss": 0.2224,
      "num_input_tokens_seen": 20371488,
      "step": 23525
    },
    {
      "epoch": 11.093823668081093,
      "grad_norm": 0.002436491660773754,
      "learning_rate": 0.10895521731593734,
      "loss": 0.2217,
      "num_input_tokens_seen": 20375920,
      "step": 23530
    },
    {
      "epoch": 11.096181046676096,
      "grad_norm": 0.0015205296222120523,
      "learning_rate": 0.10889856375388733,
      "loss": 0.2392,
      "num_input_tokens_seen": 20380320,
      "step": 23535
    },
    {
      "epoch": 11.098538425271098,
      "grad_norm": 0.0014971033670008183,
      "learning_rate": 0.1088419165301954,
      "loss": 0.1931,
      "num_input_tokens_seen": 20384400,
      "step": 23540
    },
    {
      "epoch": 11.1008958038661,
      "grad_norm": 0.0008927164017222822,
      "learning_rate": 0.1087852756535971,
      "loss": 0.2307,
      "num_input_tokens_seen": 20389120,
      "step": 23545
    },
    {
      "epoch": 11.103253182461103,
      "grad_norm": 0.0015340876998379827,
      "learning_rate": 0.10872864113282725,
      "loss": 0.1929,
      "num_input_tokens_seen": 20393408,
      "step": 23550
    },
    {
      "epoch": 11.105610561056105,
      "grad_norm": 0.001431796234101057,
      "learning_rate": 0.10867201297661958,
      "loss": 0.2583,
      "num_input_tokens_seen": 20398784,
      "step": 23555
    },
    {
      "epoch": 11.107967939651108,
      "grad_norm": 0.001124314614571631,
      "learning_rate": 0.10861539119370689,
      "loss": 0.238,
      "num_input_tokens_seen": 20404048,
      "step": 23560
    },
    {
      "epoch": 11.11032531824611,
      "grad_norm": 0.0017419190844520926,
      "learning_rate": 0.10855877579282096,
      "loss": 0.187,
      "num_input_tokens_seen": 20408448,
      "step": 23565
    },
    {
      "epoch": 11.112682696841112,
      "grad_norm": 0.003219668287783861,
      "learning_rate": 0.10850216678269252,
      "loss": 0.2385,
      "num_input_tokens_seen": 20412864,
      "step": 23570
    },
    {
      "epoch": 11.115040075436115,
      "grad_norm": 0.0009226940455846488,
      "learning_rate": 0.10844556417205146,
      "loss": 0.1942,
      "num_input_tokens_seen": 20416800,
      "step": 23575
    },
    {
      "epoch": 11.117397454031117,
      "grad_norm": 0.0017599056009203196,
      "learning_rate": 0.10838896796962669,
      "loss": 0.2192,
      "num_input_tokens_seen": 20421296,
      "step": 23580
    },
    {
      "epoch": 11.11975483262612,
      "grad_norm": 0.000826413684990257,
      "learning_rate": 0.1083323781841459,
      "loss": 0.2163,
      "num_input_tokens_seen": 20424912,
      "step": 23585
    },
    {
      "epoch": 11.122112211221122,
      "grad_norm": 0.0011797323822975159,
      "learning_rate": 0.10827579482433607,
      "loss": 0.238,
      "num_input_tokens_seen": 20428992,
      "step": 23590
    },
    {
      "epoch": 11.124469589816124,
      "grad_norm": 0.0013507818803191185,
      "learning_rate": 0.10821921789892304,
      "loss": 0.2149,
      "num_input_tokens_seen": 20433168,
      "step": 23595
    },
    {
      "epoch": 11.126826968411127,
      "grad_norm": 0.00153036555275321,
      "learning_rate": 0.10816264741663158,
      "loss": 0.2159,
      "num_input_tokens_seen": 20437952,
      "step": 23600
    },
    {
      "epoch": 11.126826968411127,
      "eval_loss": 0.21580545604228973,
      "eval_runtime": 21.9296,
      "eval_samples_per_second": 43.001,
      "eval_steps_per_second": 21.523,
      "num_input_tokens_seen": 20437952,
      "step": 23600
    },
    {
      "epoch": 11.12918434700613,
      "grad_norm": 0.0009373281500302255,
      "learning_rate": 0.10810608338618573,
      "loss": 0.212,
      "num_input_tokens_seen": 20442672,
      "step": 23605
    },
    {
      "epoch": 11.131541725601132,
      "grad_norm": 0.0017372275469824672,
      "learning_rate": 0.10804952581630821,
      "loss": 0.2144,
      "num_input_tokens_seen": 20447072,
      "step": 23610
    },
    {
      "epoch": 11.133899104196134,
      "grad_norm": 0.00094787054695189,
      "learning_rate": 0.10799297471572102,
      "loss": 0.1753,
      "num_input_tokens_seen": 20451088,
      "step": 23615
    },
    {
      "epoch": 11.136256482791136,
      "grad_norm": 0.0007915576570667326,
      "learning_rate": 0.10793643009314507,
      "loss": 0.2063,
      "num_input_tokens_seen": 20455472,
      "step": 23620
    },
    {
      "epoch": 11.138613861386139,
      "grad_norm": 0.0008447933942079544,
      "learning_rate": 0.10787989195730015,
      "loss": 0.1925,
      "num_input_tokens_seen": 20459728,
      "step": 23625
    },
    {
      "epoch": 11.140971239981141,
      "grad_norm": 0.001458832179196179,
      "learning_rate": 0.10782336031690525,
      "loss": 0.2308,
      "num_input_tokens_seen": 20464112,
      "step": 23630
    },
    {
      "epoch": 11.143328618576144,
      "grad_norm": 0.0020741235930472612,
      "learning_rate": 0.10776683518067821,
      "loss": 0.2527,
      "num_input_tokens_seen": 20468112,
      "step": 23635
    },
    {
      "epoch": 11.145685997171146,
      "grad_norm": 0.0016224384307861328,
      "learning_rate": 0.10771031655733587,
      "loss": 0.2306,
      "num_input_tokens_seen": 20473696,
      "step": 23640
    },
    {
      "epoch": 11.148043375766148,
      "grad_norm": 0.0009556170552968979,
      "learning_rate": 0.10765380445559422,
      "loss": 0.2155,
      "num_input_tokens_seen": 20477616,
      "step": 23645
    },
    {
      "epoch": 11.15040075436115,
      "grad_norm": 0.0008031262550503016,
      "learning_rate": 0.10759729888416801,
      "loss": 0.2446,
      "num_input_tokens_seen": 20481472,
      "step": 23650
    },
    {
      "epoch": 11.152758132956153,
      "grad_norm": 0.0009678932838141918,
      "learning_rate": 0.10754079985177119,
      "loss": 0.2285,
      "num_input_tokens_seen": 20485984,
      "step": 23655
    },
    {
      "epoch": 11.155115511551156,
      "grad_norm": 0.0018371345940977335,
      "learning_rate": 0.10748430736711667,
      "loss": 0.2442,
      "num_input_tokens_seen": 20490512,
      "step": 23660
    },
    {
      "epoch": 11.157472890146158,
      "grad_norm": 0.001014583744108677,
      "learning_rate": 0.10742782143891623,
      "loss": 0.2431,
      "num_input_tokens_seen": 20494288,
      "step": 23665
    },
    {
      "epoch": 11.15983026874116,
      "grad_norm": 0.0009361589327454567,
      "learning_rate": 0.10737134207588069,
      "loss": 0.2055,
      "num_input_tokens_seen": 20498368,
      "step": 23670
    },
    {
      "epoch": 11.162187647336163,
      "grad_norm": 0.001036115805618465,
      "learning_rate": 0.10731486928671992,
      "loss": 0.2362,
      "num_input_tokens_seen": 20502176,
      "step": 23675
    },
    {
      "epoch": 11.164545025931165,
      "grad_norm": 0.001800219644792378,
      "learning_rate": 0.10725840308014269,
      "loss": 0.1781,
      "num_input_tokens_seen": 20506560,
      "step": 23680
    },
    {
      "epoch": 11.166902404526168,
      "grad_norm": 0.0014015953056514263,
      "learning_rate": 0.10720194346485688,
      "loss": 0.2065,
      "num_input_tokens_seen": 20510544,
      "step": 23685
    },
    {
      "epoch": 11.16925978312117,
      "grad_norm": 0.0006467823986895382,
      "learning_rate": 0.10714549044956918,
      "loss": 0.1786,
      "num_input_tokens_seen": 20515072,
      "step": 23690
    },
    {
      "epoch": 11.171617161716172,
      "grad_norm": 0.0012705764966085553,
      "learning_rate": 0.10708904404298542,
      "loss": 0.2245,
      "num_input_tokens_seen": 20519472,
      "step": 23695
    },
    {
      "epoch": 11.173974540311175,
      "grad_norm": 0.0012578489258885384,
      "learning_rate": 0.1070326042538103,
      "loss": 0.2208,
      "num_input_tokens_seen": 20523744,
      "step": 23700
    },
    {
      "epoch": 11.176331918906177,
      "grad_norm": 0.0012912634992972016,
      "learning_rate": 0.10697617109074758,
      "loss": 0.1975,
      "num_input_tokens_seen": 20527968,
      "step": 23705
    },
    {
      "epoch": 11.17868929750118,
      "grad_norm": 0.0010073906742036343,
      "learning_rate": 0.10691974456249999,
      "loss": 0.1878,
      "num_input_tokens_seen": 20531616,
      "step": 23710
    },
    {
      "epoch": 11.18104667609618,
      "grad_norm": 0.0018123212503269315,
      "learning_rate": 0.10686332467776909,
      "loss": 0.2307,
      "num_input_tokens_seen": 20535920,
      "step": 23715
    },
    {
      "epoch": 11.183404054691183,
      "grad_norm": 0.0014034094056114554,
      "learning_rate": 0.10680691144525563,
      "loss": 0.1892,
      "num_input_tokens_seen": 20541616,
      "step": 23720
    },
    {
      "epoch": 11.185761433286185,
      "grad_norm": 0.0009230842697434127,
      "learning_rate": 0.10675050487365928,
      "loss": 0.2116,
      "num_input_tokens_seen": 20545200,
      "step": 23725
    },
    {
      "epoch": 11.188118811881187,
      "grad_norm": 0.000952008122112602,
      "learning_rate": 0.10669410497167851,
      "loss": 0.1504,
      "num_input_tokens_seen": 20549392,
      "step": 23730
    },
    {
      "epoch": 11.19047619047619,
      "grad_norm": 0.000988367130048573,
      "learning_rate": 0.10663771174801102,
      "loss": 0.1922,
      "num_input_tokens_seen": 20553008,
      "step": 23735
    },
    {
      "epoch": 11.192833569071192,
      "grad_norm": 0.0018495024414733052,
      "learning_rate": 0.10658132521135329,
      "loss": 0.2258,
      "num_input_tokens_seen": 20557264,
      "step": 23740
    },
    {
      "epoch": 11.195190947666195,
      "grad_norm": 0.0008102197316475213,
      "learning_rate": 0.10652494537040084,
      "loss": 0.1537,
      "num_input_tokens_seen": 20562144,
      "step": 23745
    },
    {
      "epoch": 11.197548326261197,
      "grad_norm": 0.000914545264095068,
      "learning_rate": 0.1064685722338482,
      "loss": 0.1839,
      "num_input_tokens_seen": 20565952,
      "step": 23750
    },
    {
      "epoch": 11.1999057048562,
      "grad_norm": 0.0013692631619051099,
      "learning_rate": 0.10641220581038871,
      "loss": 0.2519,
      "num_input_tokens_seen": 20570544,
      "step": 23755
    },
    {
      "epoch": 11.202263083451202,
      "grad_norm": 0.0018690354190766811,
      "learning_rate": 0.10635584610871483,
      "loss": 0.2358,
      "num_input_tokens_seen": 20576096,
      "step": 23760
    },
    {
      "epoch": 11.204620462046204,
      "grad_norm": 0.0010012367274612188,
      "learning_rate": 0.10629949313751803,
      "loss": 0.268,
      "num_input_tokens_seen": 20579984,
      "step": 23765
    },
    {
      "epoch": 11.206977840641207,
      "grad_norm": 0.0016885208897292614,
      "learning_rate": 0.10624314690548849,
      "loss": 0.2114,
      "num_input_tokens_seen": 20584528,
      "step": 23770
    },
    {
      "epoch": 11.209335219236209,
      "grad_norm": 0.0011862764367833734,
      "learning_rate": 0.1061868074213156,
      "loss": 0.2242,
      "num_input_tokens_seen": 20588384,
      "step": 23775
    },
    {
      "epoch": 11.211692597831211,
      "grad_norm": 0.0019721442367881536,
      "learning_rate": 0.10613047469368765,
      "loss": 0.2195,
      "num_input_tokens_seen": 20592992,
      "step": 23780
    },
    {
      "epoch": 11.214049976426214,
      "grad_norm": 0.001509863301180303,
      "learning_rate": 0.10607414873129171,
      "loss": 0.2112,
      "num_input_tokens_seen": 20597712,
      "step": 23785
    },
    {
      "epoch": 11.216407355021216,
      "grad_norm": 0.0010134277399629354,
      "learning_rate": 0.10601782954281413,
      "loss": 0.2388,
      "num_input_tokens_seen": 20603216,
      "step": 23790
    },
    {
      "epoch": 11.218764733616219,
      "grad_norm": 0.0012597590684890747,
      "learning_rate": 0.1059615171369399,
      "loss": 0.216,
      "num_input_tokens_seen": 20607344,
      "step": 23795
    },
    {
      "epoch": 11.221122112211221,
      "grad_norm": 0.0010668057948350906,
      "learning_rate": 0.10590521152235312,
      "loss": 0.1985,
      "num_input_tokens_seen": 20611040,
      "step": 23800
    },
    {
      "epoch": 11.221122112211221,
      "eval_loss": 0.2159261852502823,
      "eval_runtime": 21.9157,
      "eval_samples_per_second": 43.029,
      "eval_steps_per_second": 21.537,
      "num_input_tokens_seen": 20611040,
      "step": 23800
    },
    {
      "epoch": 11.223479490806223,
      "grad_norm": 0.0008851926540955901,
      "learning_rate": 0.1058489127077369,
      "loss": 0.1808,
      "num_input_tokens_seen": 20614848,
      "step": 23805
    },
    {
      "epoch": 11.225836869401226,
      "grad_norm": 0.0010301230940967798,
      "learning_rate": 0.1057926207017732,
      "loss": 0.227,
      "num_input_tokens_seen": 20619584,
      "step": 23810
    },
    {
      "epoch": 11.228194247996228,
      "grad_norm": 0.001175128621980548,
      "learning_rate": 0.10573633551314285,
      "loss": 0.26,
      "num_input_tokens_seen": 20623408,
      "step": 23815
    },
    {
      "epoch": 11.23055162659123,
      "grad_norm": 0.0010005505755543709,
      "learning_rate": 0.1056800571505259,
      "loss": 0.217,
      "num_input_tokens_seen": 20628128,
      "step": 23820
    },
    {
      "epoch": 11.232909005186233,
      "grad_norm": 0.001662829308770597,
      "learning_rate": 0.10562378562260105,
      "loss": 0.2361,
      "num_input_tokens_seen": 20632848,
      "step": 23825
    },
    {
      "epoch": 11.235266383781235,
      "grad_norm": 0.001914051827043295,
      "learning_rate": 0.10556752093804615,
      "loss": 0.2019,
      "num_input_tokens_seen": 20637168,
      "step": 23830
    },
    {
      "epoch": 11.237623762376238,
      "grad_norm": 0.0011869823792949319,
      "learning_rate": 0.10551126310553786,
      "loss": 0.2238,
      "num_input_tokens_seen": 20641744,
      "step": 23835
    },
    {
      "epoch": 11.23998114097124,
      "grad_norm": 0.0008630105294287205,
      "learning_rate": 0.10545501213375187,
      "loss": 0.2218,
      "num_input_tokens_seen": 20646256,
      "step": 23840
    },
    {
      "epoch": 11.242338519566243,
      "grad_norm": 0.0015701164957135916,
      "learning_rate": 0.10539876803136287,
      "loss": 0.2107,
      "num_input_tokens_seen": 20650688,
      "step": 23845
    },
    {
      "epoch": 11.244695898161245,
      "grad_norm": 0.0009891436202451587,
      "learning_rate": 0.10534253080704428,
      "loss": 0.2365,
      "num_input_tokens_seen": 20655152,
      "step": 23850
    },
    {
      "epoch": 11.247053276756247,
      "grad_norm": 0.0011949061881750822,
      "learning_rate": 0.10528630046946862,
      "loss": 0.2496,
      "num_input_tokens_seen": 20659712,
      "step": 23855
    },
    {
      "epoch": 11.24941065535125,
      "grad_norm": 0.001867667306214571,
      "learning_rate": 0.1052300770273074,
      "loss": 0.213,
      "num_input_tokens_seen": 20664048,
      "step": 23860
    },
    {
      "epoch": 11.251768033946252,
      "grad_norm": 0.0015228319680318236,
      "learning_rate": 0.10517386048923086,
      "loss": 0.2335,
      "num_input_tokens_seen": 20667888,
      "step": 23865
    },
    {
      "epoch": 11.254125412541255,
      "grad_norm": 0.0015251276781782508,
      "learning_rate": 0.10511765086390841,
      "loss": 0.2218,
      "num_input_tokens_seen": 20672256,
      "step": 23870
    },
    {
      "epoch": 11.256482791136257,
      "grad_norm": 0.0011013098992407322,
      "learning_rate": 0.10506144816000816,
      "loss": 0.2409,
      "num_input_tokens_seen": 20676400,
      "step": 23875
    },
    {
      "epoch": 11.25884016973126,
      "grad_norm": 0.0013770844088867307,
      "learning_rate": 0.10500525238619736,
      "loss": 0.2091,
      "num_input_tokens_seen": 20680352,
      "step": 23880
    },
    {
      "epoch": 11.261197548326262,
      "grad_norm": 0.0021768221631646156,
      "learning_rate": 0.10494906355114209,
      "loss": 0.2362,
      "num_input_tokens_seen": 20686368,
      "step": 23885
    },
    {
      "epoch": 11.263554926921264,
      "grad_norm": 0.0010891961865127087,
      "learning_rate": 0.10489288166350737,
      "loss": 0.2655,
      "num_input_tokens_seen": 20692064,
      "step": 23890
    },
    {
      "epoch": 11.265912305516267,
      "grad_norm": 0.0009515869314782321,
      "learning_rate": 0.10483670673195711,
      "loss": 0.2127,
      "num_input_tokens_seen": 20697072,
      "step": 23895
    },
    {
      "epoch": 11.268269684111269,
      "grad_norm": 0.0008498075767420232,
      "learning_rate": 0.10478053876515431,
      "loss": 0.2101,
      "num_input_tokens_seen": 20701232,
      "step": 23900
    },
    {
      "epoch": 11.270627062706271,
      "grad_norm": 0.0014994601951912045,
      "learning_rate": 0.10472437777176061,
      "loss": 0.2234,
      "num_input_tokens_seen": 20705104,
      "step": 23905
    },
    {
      "epoch": 11.272984441301272,
      "grad_norm": 0.001110375509597361,
      "learning_rate": 0.1046682237604369,
      "loss": 0.2407,
      "num_input_tokens_seen": 20709616,
      "step": 23910
    },
    {
      "epoch": 11.275341819896274,
      "grad_norm": 0.001589594641700387,
      "learning_rate": 0.1046120767398427,
      "loss": 0.1926,
      "num_input_tokens_seen": 20714368,
      "step": 23915
    },
    {
      "epoch": 11.277699198491277,
      "grad_norm": 0.0013552963500842452,
      "learning_rate": 0.10455593671863667,
      "loss": 0.2052,
      "num_input_tokens_seen": 20719072,
      "step": 23920
    },
    {
      "epoch": 11.28005657708628,
      "grad_norm": 0.0008104029111564159,
      "learning_rate": 0.1044998037054763,
      "loss": 0.2154,
      "num_input_tokens_seen": 20722896,
      "step": 23925
    },
    {
      "epoch": 11.282413955681282,
      "grad_norm": 0.003819710109382868,
      "learning_rate": 0.10444367770901794,
      "loss": 0.2412,
      "num_input_tokens_seen": 20727664,
      "step": 23930
    },
    {
      "epoch": 11.284771334276284,
      "grad_norm": 0.0014984929002821445,
      "learning_rate": 0.10438755873791698,
      "loss": 0.2096,
      "num_input_tokens_seen": 20731520,
      "step": 23935
    },
    {
      "epoch": 11.287128712871286,
      "grad_norm": 0.0009691211162135005,
      "learning_rate": 0.10433144680082775,
      "loss": 0.205,
      "num_input_tokens_seen": 20736208,
      "step": 23940
    },
    {
      "epoch": 11.289486091466289,
      "grad_norm": 0.001547016086988151,
      "learning_rate": 0.10427534190640322,
      "loss": 0.224,
      "num_input_tokens_seen": 20740448,
      "step": 23945
    },
    {
      "epoch": 11.291843470061291,
      "grad_norm": 0.0010443029459565878,
      "learning_rate": 0.10421924406329568,
      "loss": 0.2283,
      "num_input_tokens_seen": 20744640,
      "step": 23950
    },
    {
      "epoch": 11.294200848656294,
      "grad_norm": 0.0018526441417634487,
      "learning_rate": 0.10416315328015598,
      "loss": 0.2098,
      "num_input_tokens_seen": 20749296,
      "step": 23955
    },
    {
      "epoch": 11.296558227251296,
      "grad_norm": 0.0009104938362725079,
      "learning_rate": 0.10410706956563402,
      "loss": 0.2392,
      "num_input_tokens_seen": 20753712,
      "step": 23960
    },
    {
      "epoch": 11.298915605846299,
      "grad_norm": 0.0008192817331291735,
      "learning_rate": 0.10405099292837874,
      "loss": 0.2087,
      "num_input_tokens_seen": 20757968,
      "step": 23965
    },
    {
      "epoch": 11.301272984441301,
      "grad_norm": 0.0010730481008067727,
      "learning_rate": 0.10399492337703771,
      "loss": 0.2126,
      "num_input_tokens_seen": 20762432,
      "step": 23970
    },
    {
      "epoch": 11.303630363036303,
      "grad_norm": 0.0011436528293415904,
      "learning_rate": 0.10393886092025764,
      "loss": 0.2398,
      "num_input_tokens_seen": 20766320,
      "step": 23975
    },
    {
      "epoch": 11.305987741631306,
      "grad_norm": 0.002610074356198311,
      "learning_rate": 0.10388280556668412,
      "loss": 0.241,
      "num_input_tokens_seen": 20769872,
      "step": 23980
    },
    {
      "epoch": 11.308345120226308,
      "grad_norm": 0.0010855704313144088,
      "learning_rate": 0.10382675732496145,
      "loss": 0.2269,
      "num_input_tokens_seen": 20774912,
      "step": 23985
    },
    {
      "epoch": 11.31070249882131,
      "grad_norm": 0.0012254139874130487,
      "learning_rate": 0.10377071620373311,
      "loss": 0.1937,
      "num_input_tokens_seen": 20779072,
      "step": 23990
    },
    {
      "epoch": 11.313059877416313,
      "grad_norm": 0.001671142759732902,
      "learning_rate": 0.10371468221164128,
      "loss": 0.2208,
      "num_input_tokens_seen": 20783584,
      "step": 23995
    },
    {
      "epoch": 11.315417256011315,
      "grad_norm": 0.0024140847381204367,
      "learning_rate": 0.10365865535732706,
      "loss": 0.2262,
      "num_input_tokens_seen": 20787488,
      "step": 24000
    },
    {
      "epoch": 11.315417256011315,
      "eval_loss": 0.2140500694513321,
      "eval_runtime": 21.9108,
      "eval_samples_per_second": 43.038,
      "eval_steps_per_second": 21.542,
      "num_input_tokens_seen": 20787488,
      "step": 24000
    },
    {
      "epoch": 11.317774634606318,
      "grad_norm": 0.0011820073705166578,
      "learning_rate": 0.10360263564943062,
      "loss": 0.18,
      "num_input_tokens_seen": 20791376,
      "step": 24005
    },
    {
      "epoch": 11.32013201320132,
      "grad_norm": 0.0009163091890513897,
      "learning_rate": 0.10354662309659075,
      "loss": 0.1769,
      "num_input_tokens_seen": 20795712,
      "step": 24010
    },
    {
      "epoch": 11.322489391796323,
      "grad_norm": 0.0012044081231579185,
      "learning_rate": 0.10349061770744537,
      "loss": 0.2304,
      "num_input_tokens_seen": 20799520,
      "step": 24015
    },
    {
      "epoch": 11.324846770391325,
      "grad_norm": 0.0011261162580922246,
      "learning_rate": 0.10343461949063128,
      "loss": 0.2523,
      "num_input_tokens_seen": 20804240,
      "step": 24020
    },
    {
      "epoch": 11.327204148986327,
      "grad_norm": 0.0012648450210690498,
      "learning_rate": 0.103378628454784,
      "loss": 0.2145,
      "num_input_tokens_seen": 20807936,
      "step": 24025
    },
    {
      "epoch": 11.32956152758133,
      "grad_norm": 0.0010526706464588642,
      "learning_rate": 0.10332264460853811,
      "loss": 0.2259,
      "num_input_tokens_seen": 20812464,
      "step": 24030
    },
    {
      "epoch": 11.331918906176332,
      "grad_norm": 0.001100714085623622,
      "learning_rate": 0.10326666796052701,
      "loss": 0.2176,
      "num_input_tokens_seen": 20816640,
      "step": 24035
    },
    {
      "epoch": 11.334276284771335,
      "grad_norm": 0.001818842370994389,
      "learning_rate": 0.10321069851938296,
      "loss": 0.2146,
      "num_input_tokens_seen": 20820736,
      "step": 24040
    },
    {
      "epoch": 11.336633663366337,
      "grad_norm": 0.0011157109402120113,
      "learning_rate": 0.10315473629373724,
      "loss": 0.2065,
      "num_input_tokens_seen": 20824448,
      "step": 24045
    },
    {
      "epoch": 11.33899104196134,
      "grad_norm": 0.0014897137880325317,
      "learning_rate": 0.10309878129221982,
      "loss": 0.2164,
      "num_input_tokens_seen": 20830240,
      "step": 24050
    },
    {
      "epoch": 11.341348420556342,
      "grad_norm": 0.0010584710398688912,
      "learning_rate": 0.10304283352345973,
      "loss": 0.2214,
      "num_input_tokens_seen": 20833584,
      "step": 24055
    },
    {
      "epoch": 11.343705799151344,
      "grad_norm": 0.0017780077178031206,
      "learning_rate": 0.10298689299608486,
      "loss": 0.2359,
      "num_input_tokens_seen": 20837424,
      "step": 24060
    },
    {
      "epoch": 11.346063177746347,
      "grad_norm": 0.0011143834562972188,
      "learning_rate": 0.10293095971872188,
      "loss": 0.1859,
      "num_input_tokens_seen": 20842736,
      "step": 24065
    },
    {
      "epoch": 11.348420556341349,
      "grad_norm": 0.0010674743680283427,
      "learning_rate": 0.10287503369999645,
      "loss": 0.197,
      "num_input_tokens_seen": 20846624,
      "step": 24070
    },
    {
      "epoch": 11.350777934936351,
      "grad_norm": 0.0014067590236663818,
      "learning_rate": 0.10281911494853295,
      "loss": 0.2168,
      "num_input_tokens_seen": 20850816,
      "step": 24075
    },
    {
      "epoch": 11.353135313531354,
      "grad_norm": 0.0013903224607929587,
      "learning_rate": 0.10276320347295485,
      "loss": 0.1993,
      "num_input_tokens_seen": 20855344,
      "step": 24080
    },
    {
      "epoch": 11.355492692126356,
      "grad_norm": 0.0022055944427847862,
      "learning_rate": 0.10270729928188446,
      "loss": 0.2974,
      "num_input_tokens_seen": 20859968,
      "step": 24085
    },
    {
      "epoch": 11.357850070721359,
      "grad_norm": 0.0014435197226703167,
      "learning_rate": 0.10265140238394276,
      "loss": 0.2092,
      "num_input_tokens_seen": 20864912,
      "step": 24090
    },
    {
      "epoch": 11.360207449316361,
      "grad_norm": 0.0015210837591439486,
      "learning_rate": 0.10259551278774988,
      "loss": 0.2339,
      "num_input_tokens_seen": 20870352,
      "step": 24095
    },
    {
      "epoch": 11.362564827911363,
      "grad_norm": 0.0012145775835961103,
      "learning_rate": 0.10253963050192462,
      "loss": 0.2002,
      "num_input_tokens_seen": 20874976,
      "step": 24100
    },
    {
      "epoch": 11.364922206506366,
      "grad_norm": 0.0010406265500932932,
      "learning_rate": 0.10248375553508478,
      "loss": 0.1912,
      "num_input_tokens_seen": 20878848,
      "step": 24105
    },
    {
      "epoch": 11.367279585101368,
      "grad_norm": 0.0015806496376171708,
      "learning_rate": 0.102427887895847,
      "loss": 0.2172,
      "num_input_tokens_seen": 20883072,
      "step": 24110
    },
    {
      "epoch": 11.369636963696369,
      "grad_norm": 0.001586403464898467,
      "learning_rate": 0.10237202759282668,
      "loss": 0.2356,
      "num_input_tokens_seen": 20888032,
      "step": 24115
    },
    {
      "epoch": 11.371994342291371,
      "grad_norm": 0.0012704903492704034,
      "learning_rate": 0.10231617463463821,
      "loss": 0.2332,
      "num_input_tokens_seen": 20892432,
      "step": 24120
    },
    {
      "epoch": 11.374351720886374,
      "grad_norm": 0.001378027256578207,
      "learning_rate": 0.10226032902989492,
      "loss": 0.227,
      "num_input_tokens_seen": 20896432,
      "step": 24125
    },
    {
      "epoch": 11.376709099481376,
      "grad_norm": 0.0010164872510358691,
      "learning_rate": 0.10220449078720877,
      "loss": 0.1988,
      "num_input_tokens_seen": 20900656,
      "step": 24130
    },
    {
      "epoch": 11.379066478076378,
      "grad_norm": 0.0011842716485261917,
      "learning_rate": 0.1021486599151908,
      "loss": 0.2359,
      "num_input_tokens_seen": 20904128,
      "step": 24135
    },
    {
      "epoch": 11.38142385667138,
      "grad_norm": 0.0012928934302181005,
      "learning_rate": 0.10209283642245084,
      "loss": 0.2119,
      "num_input_tokens_seen": 20907728,
      "step": 24140
    },
    {
      "epoch": 11.383781235266383,
      "grad_norm": 0.001320133451372385,
      "learning_rate": 0.10203702031759748,
      "loss": 0.2194,
      "num_input_tokens_seen": 20912224,
      "step": 24145
    },
    {
      "epoch": 11.386138613861386,
      "grad_norm": 0.0010877851163968444,
      "learning_rate": 0.1019812116092384,
      "loss": 0.1726,
      "num_input_tokens_seen": 20916576,
      "step": 24150
    },
    {
      "epoch": 11.388495992456388,
      "grad_norm": 0.0012953471159562469,
      "learning_rate": 0.10192541030597986,
      "loss": 0.2286,
      "num_input_tokens_seen": 20920464,
      "step": 24155
    },
    {
      "epoch": 11.39085337105139,
      "grad_norm": 0.0017360274214297533,
      "learning_rate": 0.1018696164164272,
      "loss": 0.1996,
      "num_input_tokens_seen": 20925232,
      "step": 24160
    },
    {
      "epoch": 11.393210749646393,
      "grad_norm": 0.0010649935575202107,
      "learning_rate": 0.10181382994918459,
      "loss": 0.254,
      "num_input_tokens_seen": 20929056,
      "step": 24165
    },
    {
      "epoch": 11.395568128241395,
      "grad_norm": 0.0008467952720820904,
      "learning_rate": 0.10175805091285492,
      "loss": 0.2167,
      "num_input_tokens_seen": 20933264,
      "step": 24170
    },
    {
      "epoch": 11.397925506836398,
      "grad_norm": 0.0010305946925655007,
      "learning_rate": 0.10170227931603999,
      "loss": 0.2519,
      "num_input_tokens_seen": 20936944,
      "step": 24175
    },
    {
      "epoch": 11.4002828854314,
      "grad_norm": 0.0015197590691968799,
      "learning_rate": 0.10164651516734062,
      "loss": 0.2225,
      "num_input_tokens_seen": 20940768,
      "step": 24180
    },
    {
      "epoch": 11.402640264026402,
      "grad_norm": 0.0010058096377179027,
      "learning_rate": 0.1015907584753562,
      "loss": 0.2172,
      "num_input_tokens_seen": 20944736,
      "step": 24185
    },
    {
      "epoch": 11.404997642621405,
      "grad_norm": 0.0014620432630181313,
      "learning_rate": 0.10153500924868523,
      "loss": 0.2181,
      "num_input_tokens_seen": 20948960,
      "step": 24190
    },
    {
      "epoch": 11.407355021216407,
      "grad_norm": 0.0013659505639225245,
      "learning_rate": 0.10147926749592483,
      "loss": 0.2034,
      "num_input_tokens_seen": 20953440,
      "step": 24195
    },
    {
      "epoch": 11.40971239981141,
      "grad_norm": 0.0016850425163283944,
      "learning_rate": 0.10142353322567112,
      "loss": 0.2036,
      "num_input_tokens_seen": 20958240,
      "step": 24200
    },
    {
      "epoch": 11.40971239981141,
      "eval_loss": 0.21602044999599457,
      "eval_runtime": 21.9124,
      "eval_samples_per_second": 43.035,
      "eval_steps_per_second": 21.54,
      "num_input_tokens_seen": 20958240,
      "step": 24200
    },
    {
      "epoch": 11.412069778406412,
      "grad_norm": 0.0011241210158914328,
      "learning_rate": 0.1013678064465191,
      "loss": 0.1905,
      "num_input_tokens_seen": 20962608,
      "step": 24205
    },
    {
      "epoch": 11.414427157001414,
      "grad_norm": 0.0012215763563290238,
      "learning_rate": 0.10131208716706244,
      "loss": 0.1994,
      "num_input_tokens_seen": 20966848,
      "step": 24210
    },
    {
      "epoch": 11.416784535596417,
      "grad_norm": 0.0012189783155918121,
      "learning_rate": 0.10125637539589379,
      "loss": 0.2403,
      "num_input_tokens_seen": 20971184,
      "step": 24215
    },
    {
      "epoch": 11.41914191419142,
      "grad_norm": 0.0010245434241369367,
      "learning_rate": 0.10120067114160464,
      "loss": 0.2315,
      "num_input_tokens_seen": 20975568,
      "step": 24220
    },
    {
      "epoch": 11.421499292786422,
      "grad_norm": 0.0012795149814337492,
      "learning_rate": 0.10114497441278517,
      "loss": 0.1795,
      "num_input_tokens_seen": 20979264,
      "step": 24225
    },
    {
      "epoch": 11.423856671381424,
      "grad_norm": 0.0012076591374352574,
      "learning_rate": 0.10108928521802468,
      "loss": 0.2352,
      "num_input_tokens_seen": 20983312,
      "step": 24230
    },
    {
      "epoch": 11.426214049976426,
      "grad_norm": 0.001956550171598792,
      "learning_rate": 0.101033603565911,
      "loss": 0.2101,
      "num_input_tokens_seen": 20988480,
      "step": 24235
    },
    {
      "epoch": 11.428571428571429,
      "grad_norm": 0.0011335097951814532,
      "learning_rate": 0.10097792946503102,
      "loss": 0.226,
      "num_input_tokens_seen": 20994288,
      "step": 24240
    },
    {
      "epoch": 11.430928807166431,
      "grad_norm": 0.0012657798361033201,
      "learning_rate": 0.10092226292397039,
      "loss": 0.2108,
      "num_input_tokens_seen": 20999264,
      "step": 24245
    },
    {
      "epoch": 11.433286185761434,
      "grad_norm": 0.0017539538675919175,
      "learning_rate": 0.10086660395131354,
      "loss": 0.2201,
      "num_input_tokens_seen": 21003504,
      "step": 24250
    },
    {
      "epoch": 11.435643564356436,
      "grad_norm": 0.0028838827274739742,
      "learning_rate": 0.10081095255564385,
      "loss": 0.2327,
      "num_input_tokens_seen": 21007328,
      "step": 24255
    },
    {
      "epoch": 11.438000942951438,
      "grad_norm": 0.00103617156855762,
      "learning_rate": 0.10075530874554335,
      "loss": 0.2229,
      "num_input_tokens_seen": 21011200,
      "step": 24260
    },
    {
      "epoch": 11.44035832154644,
      "grad_norm": 0.0018556129653006792,
      "learning_rate": 0.10069967252959311,
      "loss": 0.2281,
      "num_input_tokens_seen": 21015712,
      "step": 24265
    },
    {
      "epoch": 11.442715700141443,
      "grad_norm": 0.0013250431511551142,
      "learning_rate": 0.10064404391637297,
      "loss": 0.2243,
      "num_input_tokens_seen": 21019792,
      "step": 24270
    },
    {
      "epoch": 11.445073078736446,
      "grad_norm": 0.0011378051713109016,
      "learning_rate": 0.10058842291446145,
      "loss": 0.2301,
      "num_input_tokens_seen": 21024128,
      "step": 24275
    },
    {
      "epoch": 11.447430457331448,
      "grad_norm": 0.0011550643248483539,
      "learning_rate": 0.10053280953243608,
      "loss": 0.2088,
      "num_input_tokens_seen": 21028816,
      "step": 24280
    },
    {
      "epoch": 11.44978783592645,
      "grad_norm": 0.002494734711945057,
      "learning_rate": 0.10047720377887315,
      "loss": 0.2408,
      "num_input_tokens_seen": 21033824,
      "step": 24285
    },
    {
      "epoch": 11.452145214521453,
      "grad_norm": 0.0014824791578575969,
      "learning_rate": 0.10042160566234767,
      "loss": 0.2254,
      "num_input_tokens_seen": 21037248,
      "step": 24290
    },
    {
      "epoch": 11.454502593116455,
      "grad_norm": 0.0015586329391226172,
      "learning_rate": 0.10036601519143372,
      "loss": 0.2101,
      "num_input_tokens_seen": 21041168,
      "step": 24295
    },
    {
      "epoch": 11.456859971711458,
      "grad_norm": 0.0010274745291098952,
      "learning_rate": 0.1003104323747039,
      "loss": 0.2102,
      "num_input_tokens_seen": 21044640,
      "step": 24300
    },
    {
      "epoch": 11.45921735030646,
      "grad_norm": 0.0010759857250377536,
      "learning_rate": 0.10025485722072984,
      "loss": 0.1999,
      "num_input_tokens_seen": 21048944,
      "step": 24305
    },
    {
      "epoch": 11.46157472890146,
      "grad_norm": 0.001420038752257824,
      "learning_rate": 0.10019928973808201,
      "loss": 0.2424,
      "num_input_tokens_seen": 21054048,
      "step": 24310
    },
    {
      "epoch": 11.463932107496463,
      "grad_norm": 0.0010925345122814178,
      "learning_rate": 0.10014372993532945,
      "loss": 0.1953,
      "num_input_tokens_seen": 21058400,
      "step": 24315
    },
    {
      "epoch": 11.466289486091465,
      "grad_norm": 0.002286000642925501,
      "learning_rate": 0.1000881778210403,
      "loss": 0.2678,
      "num_input_tokens_seen": 21063056,
      "step": 24320
    },
    {
      "epoch": 11.468646864686468,
      "grad_norm": 0.0015329233137890697,
      "learning_rate": 0.10003263340378142,
      "loss": 0.2154,
      "num_input_tokens_seen": 21067328,
      "step": 24325
    },
    {
      "epoch": 11.47100424328147,
      "grad_norm": 0.0012834753142669797,
      "learning_rate": 0.09997709669211834,
      "loss": 0.2321,
      "num_input_tokens_seen": 21072160,
      "step": 24330
    },
    {
      "epoch": 11.473361621876473,
      "grad_norm": 0.0021054602693766356,
      "learning_rate": 0.0999215676946156,
      "loss": 0.2166,
      "num_input_tokens_seen": 21076128,
      "step": 24335
    },
    {
      "epoch": 11.475719000471475,
      "grad_norm": 0.0012935794657096267,
      "learning_rate": 0.0998660464198364,
      "loss": 0.2219,
      "num_input_tokens_seen": 21080208,
      "step": 24340
    },
    {
      "epoch": 11.478076379066477,
      "grad_norm": 0.0017072336049750447,
      "learning_rate": 0.09981053287634288,
      "loss": 0.2017,
      "num_input_tokens_seen": 21085040,
      "step": 24345
    },
    {
      "epoch": 11.48043375766148,
      "grad_norm": 0.0012616710737347603,
      "learning_rate": 0.09975502707269596,
      "loss": 0.2226,
      "num_input_tokens_seen": 21089200,
      "step": 24350
    },
    {
      "epoch": 11.482791136256482,
      "grad_norm": 0.0012929923832416534,
      "learning_rate": 0.09969952901745524,
      "loss": 0.2039,
      "num_input_tokens_seen": 21092944,
      "step": 24355
    },
    {
      "epoch": 11.485148514851485,
      "grad_norm": 0.002151044551283121,
      "learning_rate": 0.09964403871917925,
      "loss": 0.2218,
      "num_input_tokens_seen": 21097136,
      "step": 24360
    },
    {
      "epoch": 11.487505893446487,
      "grad_norm": 0.0024337039794772863,
      "learning_rate": 0.09958855618642536,
      "loss": 0.2308,
      "num_input_tokens_seen": 21102496,
      "step": 24365
    },
    {
      "epoch": 11.48986327204149,
      "grad_norm": 0.0012786609586328268,
      "learning_rate": 0.09953308142774955,
      "loss": 0.2097,
      "num_input_tokens_seen": 21106624,
      "step": 24370
    },
    {
      "epoch": 11.492220650636492,
      "grad_norm": 0.0010939042549580336,
      "learning_rate": 0.09947761445170686,
      "loss": 0.1855,
      "num_input_tokens_seen": 21111632,
      "step": 24375
    },
    {
      "epoch": 11.494578029231494,
      "grad_norm": 0.005516489502042532,
      "learning_rate": 0.09942215526685086,
      "loss": 0.257,
      "num_input_tokens_seen": 21115680,
      "step": 24380
    },
    {
      "epoch": 11.496935407826497,
      "grad_norm": 0.001244401908479631,
      "learning_rate": 0.09936670388173414,
      "loss": 0.2255,
      "num_input_tokens_seen": 21120192,
      "step": 24385
    },
    {
      "epoch": 11.499292786421499,
      "grad_norm": 0.0016690681222826242,
      "learning_rate": 0.09931126030490799,
      "loss": 0.2138,
      "num_input_tokens_seen": 21124144,
      "step": 24390
    },
    {
      "epoch": 11.501650165016502,
      "grad_norm": 0.0011564086889848113,
      "learning_rate": 0.0992558245449225,
      "loss": 0.1822,
      "num_input_tokens_seen": 21129024,
      "step": 24395
    },
    {
      "epoch": 11.504007543611504,
      "grad_norm": 0.0018102757167071104,
      "learning_rate": 0.09920039661032651,
      "loss": 0.2398,
      "num_input_tokens_seen": 21133392,
      "step": 24400
    },
    {
      "epoch": 11.504007543611504,
      "eval_loss": 0.21459263563156128,
      "eval_runtime": 21.9296,
      "eval_samples_per_second": 43.001,
      "eval_steps_per_second": 21.523,
      "num_input_tokens_seen": 21133392,
      "step": 24400
    },
    {
      "epoch": 11.506364922206506,
      "grad_norm": 0.0013346398482099175,
      "learning_rate": 0.09914497650966782,
      "loss": 0.2277,
      "num_input_tokens_seen": 21137664,
      "step": 24405
    },
    {
      "epoch": 11.508722300801509,
      "grad_norm": 0.001345939002931118,
      "learning_rate": 0.09908956425149276,
      "loss": 0.196,
      "num_input_tokens_seen": 21141136,
      "step": 24410
    },
    {
      "epoch": 11.511079679396511,
      "grad_norm": 0.0013239822583273053,
      "learning_rate": 0.09903415984434677,
      "loss": 0.2132,
      "num_input_tokens_seen": 21146080,
      "step": 24415
    },
    {
      "epoch": 11.513437057991514,
      "grad_norm": 0.0013430117396637797,
      "learning_rate": 0.09897876329677373,
      "loss": 0.2228,
      "num_input_tokens_seen": 21149920,
      "step": 24420
    },
    {
      "epoch": 11.515794436586516,
      "grad_norm": 0.0011524935252964497,
      "learning_rate": 0.09892337461731658,
      "loss": 0.2044,
      "num_input_tokens_seen": 21153952,
      "step": 24425
    },
    {
      "epoch": 11.518151815181518,
      "grad_norm": 0.002316533587872982,
      "learning_rate": 0.09886799381451693,
      "loss": 0.2391,
      "num_input_tokens_seen": 21158768,
      "step": 24430
    },
    {
      "epoch": 11.52050919377652,
      "grad_norm": 0.001123458961956203,
      "learning_rate": 0.09881262089691521,
      "loss": 0.2415,
      "num_input_tokens_seen": 21162720,
      "step": 24435
    },
    {
      "epoch": 11.522866572371523,
      "grad_norm": 0.0015193960862234235,
      "learning_rate": 0.09875725587305059,
      "loss": 0.2133,
      "num_input_tokens_seen": 21166944,
      "step": 24440
    },
    {
      "epoch": 11.525223950966526,
      "grad_norm": 0.0013115541078150272,
      "learning_rate": 0.09870189875146111,
      "loss": 0.1907,
      "num_input_tokens_seen": 21171904,
      "step": 24445
    },
    {
      "epoch": 11.527581329561528,
      "grad_norm": 0.0010188404703512788,
      "learning_rate": 0.09864654954068346,
      "loss": 0.2154,
      "num_input_tokens_seen": 21175728,
      "step": 24450
    },
    {
      "epoch": 11.52993870815653,
      "grad_norm": 0.0016422872431576252,
      "learning_rate": 0.09859120824925326,
      "loss": 0.1823,
      "num_input_tokens_seen": 21180192,
      "step": 24455
    },
    {
      "epoch": 11.532296086751533,
      "grad_norm": 0.0015043941093608737,
      "learning_rate": 0.09853587488570474,
      "loss": 0.1908,
      "num_input_tokens_seen": 21184224,
      "step": 24460
    },
    {
      "epoch": 11.534653465346535,
      "grad_norm": 0.002071185503154993,
      "learning_rate": 0.09848054945857107,
      "loss": 0.2591,
      "num_input_tokens_seen": 21188656,
      "step": 24465
    },
    {
      "epoch": 11.537010843941538,
      "grad_norm": 0.004276470746845007,
      "learning_rate": 0.09842523197638416,
      "loss": 0.1978,
      "num_input_tokens_seen": 21192912,
      "step": 24470
    },
    {
      "epoch": 11.53936822253654,
      "grad_norm": 0.0013434748398140073,
      "learning_rate": 0.09836992244767452,
      "loss": 0.204,
      "num_input_tokens_seen": 21197680,
      "step": 24475
    },
    {
      "epoch": 11.541725601131542,
      "grad_norm": 0.0012395968660712242,
      "learning_rate": 0.09831462088097168,
      "loss": 0.2321,
      "num_input_tokens_seen": 21201408,
      "step": 24480
    },
    {
      "epoch": 11.544082979726545,
      "grad_norm": 0.0012872024672105908,
      "learning_rate": 0.09825932728480385,
      "loss": 0.2039,
      "num_input_tokens_seen": 21205392,
      "step": 24485
    },
    {
      "epoch": 11.546440358321547,
      "grad_norm": 0.0021323596592992544,
      "learning_rate": 0.09820404166769794,
      "loss": 0.2641,
      "num_input_tokens_seen": 21210048,
      "step": 24490
    },
    {
      "epoch": 11.54879773691655,
      "grad_norm": 0.0010779349831864238,
      "learning_rate": 0.09814876403817978,
      "loss": 0.1939,
      "num_input_tokens_seen": 21213840,
      "step": 24495
    },
    {
      "epoch": 11.551155115511552,
      "grad_norm": 0.0014021014794707298,
      "learning_rate": 0.09809349440477376,
      "loss": 0.2299,
      "num_input_tokens_seen": 21218160,
      "step": 24500
    },
    {
      "epoch": 11.553512494106554,
      "grad_norm": 0.0014081717235967517,
      "learning_rate": 0.09803823277600317,
      "loss": 0.211,
      "num_input_tokens_seen": 21222688,
      "step": 24505
    },
    {
      "epoch": 11.555869872701557,
      "grad_norm": 0.0012839860282838345,
      "learning_rate": 0.09798297916039014,
      "loss": 0.2287,
      "num_input_tokens_seen": 21226464,
      "step": 24510
    },
    {
      "epoch": 11.558227251296557,
      "grad_norm": 0.001742522232234478,
      "learning_rate": 0.09792773356645534,
      "loss": 0.2388,
      "num_input_tokens_seen": 21230544,
      "step": 24515
    },
    {
      "epoch": 11.56058462989156,
      "grad_norm": 0.002315906574949622,
      "learning_rate": 0.09787249600271843,
      "loss": 0.223,
      "num_input_tokens_seen": 21235088,
      "step": 24520
    },
    {
      "epoch": 11.562942008486562,
      "grad_norm": 0.001618875190615654,
      "learning_rate": 0.09781726647769776,
      "loss": 0.2319,
      "num_input_tokens_seen": 21239504,
      "step": 24525
    },
    {
      "epoch": 11.565299387081565,
      "grad_norm": 0.0014063516864553094,
      "learning_rate": 0.0977620449999103,
      "loss": 0.209,
      "num_input_tokens_seen": 21243392,
      "step": 24530
    },
    {
      "epoch": 11.567656765676567,
      "grad_norm": 0.002089003100991249,
      "learning_rate": 0.09770683157787204,
      "loss": 0.2008,
      "num_input_tokens_seen": 21246816,
      "step": 24535
    },
    {
      "epoch": 11.57001414427157,
      "grad_norm": 0.0020877420902252197,
      "learning_rate": 0.09765162622009745,
      "loss": 0.2045,
      "num_input_tokens_seen": 21251920,
      "step": 24540
    },
    {
      "epoch": 11.572371522866572,
      "grad_norm": 0.001878962037153542,
      "learning_rate": 0.09759642893509995,
      "loss": 0.2491,
      "num_input_tokens_seen": 21256640,
      "step": 24545
    },
    {
      "epoch": 11.574728901461574,
      "grad_norm": 0.0014620802830904722,
      "learning_rate": 0.09754123973139169,
      "loss": 0.2273,
      "num_input_tokens_seen": 21260784,
      "step": 24550
    },
    {
      "epoch": 11.577086280056577,
      "grad_norm": 0.001695121987722814,
      "learning_rate": 0.09748605861748345,
      "loss": 0.1955,
      "num_input_tokens_seen": 21265712,
      "step": 24555
    },
    {
      "epoch": 11.579443658651579,
      "grad_norm": 0.0012805418809875846,
      "learning_rate": 0.0974308856018849,
      "loss": 0.2137,
      "num_input_tokens_seen": 21269360,
      "step": 24560
    },
    {
      "epoch": 11.581801037246581,
      "grad_norm": 0.0013549226569011807,
      "learning_rate": 0.09737572069310449,
      "loss": 0.1893,
      "num_input_tokens_seen": 21273440,
      "step": 24565
    },
    {
      "epoch": 11.584158415841584,
      "grad_norm": 0.0036154750268906355,
      "learning_rate": 0.09732056389964922,
      "loss": 0.261,
      "num_input_tokens_seen": 21278064,
      "step": 24570
    },
    {
      "epoch": 11.586515794436586,
      "grad_norm": 0.0014415037585422397,
      "learning_rate": 0.097265415230025,
      "loss": 0.2366,
      "num_input_tokens_seen": 21282336,
      "step": 24575
    },
    {
      "epoch": 11.588873173031589,
      "grad_norm": 0.0010968868155032396,
      "learning_rate": 0.09721027469273648,
      "loss": 0.2373,
      "num_input_tokens_seen": 21286544,
      "step": 24580
    },
    {
      "epoch": 11.591230551626591,
      "grad_norm": 0.001220164936967194,
      "learning_rate": 0.09715514229628695,
      "loss": 0.1998,
      "num_input_tokens_seen": 21290352,
      "step": 24585
    },
    {
      "epoch": 11.593587930221593,
      "grad_norm": 0.0009680346702225506,
      "learning_rate": 0.09710001804917864,
      "loss": 0.2155,
      "num_input_tokens_seen": 21295072,
      "step": 24590
    },
    {
      "epoch": 11.595945308816596,
      "grad_norm": 0.0019337751436978579,
      "learning_rate": 0.09704490195991226,
      "loss": 0.2261,
      "num_input_tokens_seen": 21299440,
      "step": 24595
    },
    {
      "epoch": 11.598302687411598,
      "grad_norm": 0.0018808655440807343,
      "learning_rate": 0.09698979403698753,
      "loss": 0.236,
      "num_input_tokens_seen": 21303360,
      "step": 24600
    },
    {
      "epoch": 11.598302687411598,
      "eval_loss": 0.21430175006389618,
      "eval_runtime": 21.9486,
      "eval_samples_per_second": 42.964,
      "eval_steps_per_second": 21.505,
      "num_input_tokens_seen": 21303360,
      "step": 24600
    },
    {
      "epoch": 11.6006600660066,
      "grad_norm": 0.001367606921121478,
      "learning_rate": 0.0969346942889027,
      "loss": 0.2072,
      "num_input_tokens_seen": 21307776,
      "step": 24605
    },
    {
      "epoch": 11.603017444601603,
      "grad_norm": 0.001170686911791563,
      "learning_rate": 0.09687960272415487,
      "loss": 0.1893,
      "num_input_tokens_seen": 21311696,
      "step": 24610
    },
    {
      "epoch": 11.605374823196605,
      "grad_norm": 0.001034544431604445,
      "learning_rate": 0.0968245193512399,
      "loss": 0.2116,
      "num_input_tokens_seen": 21315808,
      "step": 24615
    },
    {
      "epoch": 11.607732201791608,
      "grad_norm": 0.0016862135380506516,
      "learning_rate": 0.09676944417865221,
      "loss": 0.2406,
      "num_input_tokens_seen": 21319968,
      "step": 24620
    },
    {
      "epoch": 11.61008958038661,
      "grad_norm": 0.0016530383145436645,
      "learning_rate": 0.09671437721488517,
      "loss": 0.2105,
      "num_input_tokens_seen": 21324416,
      "step": 24625
    },
    {
      "epoch": 11.612446958981613,
      "grad_norm": 0.0010541364317759871,
      "learning_rate": 0.09665931846843086,
      "loss": 0.2025,
      "num_input_tokens_seen": 21328784,
      "step": 24630
    },
    {
      "epoch": 11.614804337576615,
      "grad_norm": 0.00205571879632771,
      "learning_rate": 0.0966042679477799,
      "loss": 0.208,
      "num_input_tokens_seen": 21333744,
      "step": 24635
    },
    {
      "epoch": 11.617161716171617,
      "grad_norm": 0.0012913331156596541,
      "learning_rate": 0.09654922566142186,
      "loss": 0.2052,
      "num_input_tokens_seen": 21337696,
      "step": 24640
    },
    {
      "epoch": 11.61951909476662,
      "grad_norm": 0.0015795450890436769,
      "learning_rate": 0.09649419161784498,
      "loss": 0.2287,
      "num_input_tokens_seen": 21342272,
      "step": 24645
    },
    {
      "epoch": 11.621876473361622,
      "grad_norm": 0.0016384940827265382,
      "learning_rate": 0.09643916582553606,
      "loss": 0.2167,
      "num_input_tokens_seen": 21345936,
      "step": 24650
    },
    {
      "epoch": 11.624233851956625,
      "grad_norm": 0.0016610833117738366,
      "learning_rate": 0.09638414829298093,
      "loss": 0.1613,
      "num_input_tokens_seen": 21350672,
      "step": 24655
    },
    {
      "epoch": 11.626591230551627,
      "grad_norm": 0.0015123037155717611,
      "learning_rate": 0.09632913902866386,
      "loss": 0.2399,
      "num_input_tokens_seen": 21355232,
      "step": 24660
    },
    {
      "epoch": 11.62894860914663,
      "grad_norm": 0.0014813189627602696,
      "learning_rate": 0.096274138041068,
      "loss": 0.2107,
      "num_input_tokens_seen": 21359744,
      "step": 24665
    },
    {
      "epoch": 11.631305987741632,
      "grad_norm": 0.0021001307759433985,
      "learning_rate": 0.09621914533867527,
      "loss": 0.2346,
      "num_input_tokens_seen": 21363104,
      "step": 24670
    },
    {
      "epoch": 11.633663366336634,
      "grad_norm": 0.0019138817442581058,
      "learning_rate": 0.09616416092996616,
      "loss": 0.188,
      "num_input_tokens_seen": 21367232,
      "step": 24675
    },
    {
      "epoch": 11.636020744931637,
      "grad_norm": 0.0010424847714602947,
      "learning_rate": 0.09610918482342,
      "loss": 0.2028,
      "num_input_tokens_seen": 21371808,
      "step": 24680
    },
    {
      "epoch": 11.638378123526639,
      "grad_norm": 0.0010596808278933167,
      "learning_rate": 0.09605421702751478,
      "loss": 0.2033,
      "num_input_tokens_seen": 21375840,
      "step": 24685
    },
    {
      "epoch": 11.640735502121641,
      "grad_norm": 0.0012585166841745377,
      "learning_rate": 0.09599925755072718,
      "loss": 0.2248,
      "num_input_tokens_seen": 21380112,
      "step": 24690
    },
    {
      "epoch": 11.643092880716644,
      "grad_norm": 0.0010412329575046897,
      "learning_rate": 0.09594430640153273,
      "loss": 0.2337,
      "num_input_tokens_seen": 21384096,
      "step": 24695
    },
    {
      "epoch": 11.645450259311646,
      "grad_norm": 0.0023194735404103994,
      "learning_rate": 0.09588936358840547,
      "loss": 0.2443,
      "num_input_tokens_seen": 21387728,
      "step": 24700
    },
    {
      "epoch": 11.647807637906649,
      "grad_norm": 0.0021233169827610254,
      "learning_rate": 0.09583442911981836,
      "loss": 0.2184,
      "num_input_tokens_seen": 21391808,
      "step": 24705
    },
    {
      "epoch": 11.65016501650165,
      "grad_norm": 0.0023815142922103405,
      "learning_rate": 0.09577950300424302,
      "loss": 0.2107,
      "num_input_tokens_seen": 21396832,
      "step": 24710
    },
    {
      "epoch": 11.652522395096653,
      "grad_norm": 0.0021937149576842785,
      "learning_rate": 0.09572458525014967,
      "loss": 0.2175,
      "num_input_tokens_seen": 21401184,
      "step": 24715
    },
    {
      "epoch": 11.654879773691654,
      "grad_norm": 0.0019147262210026383,
      "learning_rate": 0.0956696758660073,
      "loss": 0.193,
      "num_input_tokens_seen": 21406176,
      "step": 24720
    },
    {
      "epoch": 11.657237152286656,
      "grad_norm": 0.0009050863445736468,
      "learning_rate": 0.09561477486028373,
      "loss": 0.2136,
      "num_input_tokens_seen": 21411312,
      "step": 24725
    },
    {
      "epoch": 11.659594530881659,
      "grad_norm": 0.001486301771365106,
      "learning_rate": 0.09555988224144528,
      "loss": 0.2033,
      "num_input_tokens_seen": 21415616,
      "step": 24730
    },
    {
      "epoch": 11.661951909476661,
      "grad_norm": 0.0014088016469031572,
      "learning_rate": 0.09550499801795717,
      "loss": 0.238,
      "num_input_tokens_seen": 21419904,
      "step": 24735
    },
    {
      "epoch": 11.664309288071664,
      "grad_norm": 0.0010885733645409346,
      "learning_rate": 0.09545012219828314,
      "loss": 0.2057,
      "num_input_tokens_seen": 21424704,
      "step": 24740
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 0.0017249888041988015,
      "learning_rate": 0.09539525479088577,
      "loss": 0.2285,
      "num_input_tokens_seen": 21429472,
      "step": 24745
    },
    {
      "epoch": 11.669024045261668,
      "grad_norm": 0.0011772416764870286,
      "learning_rate": 0.0953403958042264,
      "loss": 0.2186,
      "num_input_tokens_seen": 21433952,
      "step": 24750
    },
    {
      "epoch": 11.67138142385667,
      "grad_norm": 0.0008628592477180064,
      "learning_rate": 0.09528554524676484,
      "loss": 0.1772,
      "num_input_tokens_seen": 21438128,
      "step": 24755
    },
    {
      "epoch": 11.673738802451673,
      "grad_norm": 0.00136595091316849,
      "learning_rate": 0.09523070312695978,
      "loss": 0.2253,
      "num_input_tokens_seen": 21441504,
      "step": 24760
    },
    {
      "epoch": 11.676096181046676,
      "grad_norm": 0.0008873876067809761,
      "learning_rate": 0.09517586945326863,
      "loss": 0.2628,
      "num_input_tokens_seen": 21445232,
      "step": 24765
    },
    {
      "epoch": 11.678453559641678,
      "grad_norm": 0.0008397348574362695,
      "learning_rate": 0.0951210442341473,
      "loss": 0.1787,
      "num_input_tokens_seen": 21449280,
      "step": 24770
    },
    {
      "epoch": 11.68081093823668,
      "grad_norm": 0.0010120074730366468,
      "learning_rate": 0.09506622747805066,
      "loss": 0.2093,
      "num_input_tokens_seen": 21453200,
      "step": 24775
    },
    {
      "epoch": 11.683168316831683,
      "grad_norm": 0.0016064882511273026,
      "learning_rate": 0.09501141919343203,
      "loss": 0.2644,
      "num_input_tokens_seen": 21457888,
      "step": 24780
    },
    {
      "epoch": 11.685525695426685,
      "grad_norm": 0.0015735324705019593,
      "learning_rate": 0.09495661938874361,
      "loss": 0.1943,
      "num_input_tokens_seen": 21462176,
      "step": 24785
    },
    {
      "epoch": 11.687883074021688,
      "grad_norm": 0.000923692889045924,
      "learning_rate": 0.0949018280724362,
      "loss": 0.2391,
      "num_input_tokens_seen": 21466336,
      "step": 24790
    },
    {
      "epoch": 11.69024045261669,
      "grad_norm": 0.0018216018797829747,
      "learning_rate": 0.09484704525295934,
      "loss": 0.2168,
      "num_input_tokens_seen": 21471296,
      "step": 24795
    },
    {
      "epoch": 11.692597831211692,
      "grad_norm": 0.0009637311450205743,
      "learning_rate": 0.09479227093876112,
      "loss": 0.2261,
      "num_input_tokens_seen": 21475184,
      "step": 24800
    },
    {
      "epoch": 11.692597831211692,
      "eval_loss": 0.2172553837299347,
      "eval_runtime": 21.9208,
      "eval_samples_per_second": 43.019,
      "eval_steps_per_second": 21.532,
      "num_input_tokens_seen": 21475184,
      "step": 24800
    },
    {
      "epoch": 11.694955209806695,
      "grad_norm": 0.0010780264856293797,
      "learning_rate": 0.0947375051382886,
      "loss": 0.243,
      "num_input_tokens_seen": 21480320,
      "step": 24805
    },
    {
      "epoch": 11.697312588401697,
      "grad_norm": 0.001218119403347373,
      "learning_rate": 0.09468274785998718,
      "loss": 0.2172,
      "num_input_tokens_seen": 21484800,
      "step": 24810
    },
    {
      "epoch": 11.6996699669967,
      "grad_norm": 0.002056143945083022,
      "learning_rate": 0.09462799911230127,
      "loss": 0.2335,
      "num_input_tokens_seen": 21489104,
      "step": 24815
    },
    {
      "epoch": 11.702027345591702,
      "grad_norm": 0.0011257024016231298,
      "learning_rate": 0.0945732589036737,
      "loss": 0.2094,
      "num_input_tokens_seen": 21492928,
      "step": 24820
    },
    {
      "epoch": 11.704384724186705,
      "grad_norm": 0.0016977416817098856,
      "learning_rate": 0.09451852724254614,
      "loss": 0.2103,
      "num_input_tokens_seen": 21497456,
      "step": 24825
    },
    {
      "epoch": 11.706742102781707,
      "grad_norm": 0.0009925810154527426,
      "learning_rate": 0.09446380413735894,
      "loss": 0.2033,
      "num_input_tokens_seen": 21501440,
      "step": 24830
    },
    {
      "epoch": 11.70909948137671,
      "grad_norm": 0.0012298488290980458,
      "learning_rate": 0.09440908959655099,
      "loss": 0.221,
      "num_input_tokens_seen": 21506640,
      "step": 24835
    },
    {
      "epoch": 11.711456859971712,
      "grad_norm": 0.0011750245466828346,
      "learning_rate": 0.09435438362856004,
      "loss": 0.1949,
      "num_input_tokens_seen": 21511536,
      "step": 24840
    },
    {
      "epoch": 11.713814238566714,
      "grad_norm": 0.001713107223622501,
      "learning_rate": 0.0942996862418225,
      "loss": 0.222,
      "num_input_tokens_seen": 21516192,
      "step": 24845
    },
    {
      "epoch": 11.716171617161717,
      "grad_norm": 0.002661296399310231,
      "learning_rate": 0.09424499744477322,
      "loss": 0.2352,
      "num_input_tokens_seen": 21520032,
      "step": 24850
    },
    {
      "epoch": 11.718528995756719,
      "grad_norm": 0.0015966345090419054,
      "learning_rate": 0.09419031724584608,
      "loss": 0.1575,
      "num_input_tokens_seen": 21524368,
      "step": 24855
    },
    {
      "epoch": 11.720886374351721,
      "grad_norm": 0.0010222073178738356,
      "learning_rate": 0.09413564565347331,
      "loss": 0.1602,
      "num_input_tokens_seen": 21528832,
      "step": 24860
    },
    {
      "epoch": 11.723243752946724,
      "grad_norm": 0.0017616093391552567,
      "learning_rate": 0.094080982676086,
      "loss": 0.1948,
      "num_input_tokens_seen": 21533328,
      "step": 24865
    },
    {
      "epoch": 11.725601131541726,
      "grad_norm": 0.0010742853628471494,
      "learning_rate": 0.09402632832211395,
      "loss": 0.1844,
      "num_input_tokens_seen": 21537072,
      "step": 24870
    },
    {
      "epoch": 11.727958510136729,
      "grad_norm": 0.0038774514105170965,
      "learning_rate": 0.09397168259998541,
      "loss": 0.2809,
      "num_input_tokens_seen": 21541584,
      "step": 24875
    },
    {
      "epoch": 11.730315888731731,
      "grad_norm": 0.0019322356674820185,
      "learning_rate": 0.09391704551812759,
      "loss": 0.2217,
      "num_input_tokens_seen": 21545728,
      "step": 24880
    },
    {
      "epoch": 11.732673267326733,
      "grad_norm": 0.001406824099831283,
      "learning_rate": 0.09386241708496605,
      "loss": 0.2386,
      "num_input_tokens_seen": 21549536,
      "step": 24885
    },
    {
      "epoch": 11.735030645921736,
      "grad_norm": 0.001991902943700552,
      "learning_rate": 0.09380779730892527,
      "loss": 0.1957,
      "num_input_tokens_seen": 21554736,
      "step": 24890
    },
    {
      "epoch": 11.737388024516738,
      "grad_norm": 0.001078232191503048,
      "learning_rate": 0.09375318619842836,
      "loss": 0.2059,
      "num_input_tokens_seen": 21558768,
      "step": 24895
    },
    {
      "epoch": 11.73974540311174,
      "grad_norm": 0.0013251790078356862,
      "learning_rate": 0.09369858376189696,
      "loss": 0.2173,
      "num_input_tokens_seen": 21563232,
      "step": 24900
    },
    {
      "epoch": 11.742102781706743,
      "grad_norm": 0.0011969280894845724,
      "learning_rate": 0.09364399000775143,
      "loss": 0.1949,
      "num_input_tokens_seen": 21567040,
      "step": 24905
    },
    {
      "epoch": 11.744460160301745,
      "grad_norm": 0.0009954662527889013,
      "learning_rate": 0.09358940494441093,
      "loss": 0.1923,
      "num_input_tokens_seen": 21571392,
      "step": 24910
    },
    {
      "epoch": 11.746817538896746,
      "grad_norm": 0.0011927669402211905,
      "learning_rate": 0.09353482858029301,
      "loss": 0.2354,
      "num_input_tokens_seen": 21576016,
      "step": 24915
    },
    {
      "epoch": 11.749174917491748,
      "grad_norm": 0.0015159152681007981,
      "learning_rate": 0.09348026092381419,
      "loss": 0.2365,
      "num_input_tokens_seen": 21579920,
      "step": 24920
    },
    {
      "epoch": 11.75153229608675,
      "grad_norm": 0.0022116571199148893,
      "learning_rate": 0.09342570198338931,
      "loss": 0.2025,
      "num_input_tokens_seen": 21584592,
      "step": 24925
    },
    {
      "epoch": 11.753889674681753,
      "grad_norm": 0.0014214979019016027,
      "learning_rate": 0.0933711517674322,
      "loss": 0.1741,
      "num_input_tokens_seen": 21588944,
      "step": 24930
    },
    {
      "epoch": 11.756247053276756,
      "grad_norm": 0.0016390391392633319,
      "learning_rate": 0.09331661028435513,
      "loss": 0.1837,
      "num_input_tokens_seen": 21593808,
      "step": 24935
    },
    {
      "epoch": 11.758604431871758,
      "grad_norm": 0.00120930268894881,
      "learning_rate": 0.09326207754256909,
      "loss": 0.2079,
      "num_input_tokens_seen": 21598560,
      "step": 24940
    },
    {
      "epoch": 11.76096181046676,
      "grad_norm": 0.0008626120397821069,
      "learning_rate": 0.09320755355048366,
      "loss": 0.2238,
      "num_input_tokens_seen": 21602944,
      "step": 24945
    },
    {
      "epoch": 11.763319189061763,
      "grad_norm": 0.0011741317575797439,
      "learning_rate": 0.09315303831650722,
      "loss": 0.2793,
      "num_input_tokens_seen": 21607392,
      "step": 24950
    },
    {
      "epoch": 11.765676567656765,
      "grad_norm": 0.0017197990091517568,
      "learning_rate": 0.09309853184904661,
      "loss": 0.1757,
      "num_input_tokens_seen": 21611680,
      "step": 24955
    },
    {
      "epoch": 11.768033946251768,
      "grad_norm": 0.0014889169251546264,
      "learning_rate": 0.09304403415650753,
      "loss": 0.2059,
      "num_input_tokens_seen": 21615632,
      "step": 24960
    },
    {
      "epoch": 11.77039132484677,
      "grad_norm": 0.0012549147941172123,
      "learning_rate": 0.09298954524729405,
      "loss": 0.2095,
      "num_input_tokens_seen": 21619216,
      "step": 24965
    },
    {
      "epoch": 11.772748703441772,
      "grad_norm": 0.0012689559953287244,
      "learning_rate": 0.09293506512980916,
      "loss": 0.2186,
      "num_input_tokens_seen": 21623808,
      "step": 24970
    },
    {
      "epoch": 11.775106082036775,
      "grad_norm": 0.001233382266946137,
      "learning_rate": 0.0928805938124544,
      "loss": 0.2194,
      "num_input_tokens_seen": 21627888,
      "step": 24975
    },
    {
      "epoch": 11.777463460631777,
      "grad_norm": 0.0013282728614285588,
      "learning_rate": 0.09282613130362982,
      "loss": 0.1779,
      "num_input_tokens_seen": 21632256,
      "step": 24980
    },
    {
      "epoch": 11.77982083922678,
      "grad_norm": 0.001700561260804534,
      "learning_rate": 0.09277167761173427,
      "loss": 0.2652,
      "num_input_tokens_seen": 21636576,
      "step": 24985
    },
    {
      "epoch": 11.782178217821782,
      "grad_norm": 0.0013899715850129724,
      "learning_rate": 0.0927172327451653,
      "loss": 0.2131,
      "num_input_tokens_seen": 21640560,
      "step": 24990
    },
    {
      "epoch": 11.784535596416784,
      "grad_norm": 0.0010246846359223127,
      "learning_rate": 0.09266279671231882,
      "loss": 0.1575,
      "num_input_tokens_seen": 21644864,
      "step": 24995
    },
    {
      "epoch": 11.786892975011787,
      "grad_norm": 0.0012227657716721296,
      "learning_rate": 0.09260836952158967,
      "loss": 0.1897,
      "num_input_tokens_seen": 21649744,
      "step": 25000
    },
    {
      "epoch": 11.786892975011787,
      "eval_loss": 0.2155287116765976,
      "eval_runtime": 21.8904,
      "eval_samples_per_second": 43.078,
      "eval_steps_per_second": 21.562,
      "num_input_tokens_seen": 21649744,
      "step": 25000
    },
    {
      "epoch": 11.78925035360679,
      "grad_norm": 0.002352917566895485,
      "learning_rate": 0.09255395118137114,
      "loss": 0.2355,
      "num_input_tokens_seen": 21653408,
      "step": 25005
    },
    {
      "epoch": 11.791607732201792,
      "grad_norm": 0.0028916732408106327,
      "learning_rate": 0.09249954170005527,
      "loss": 0.2129,
      "num_input_tokens_seen": 21657904,
      "step": 25010
    },
    {
      "epoch": 11.793965110796794,
      "grad_norm": 0.0015345962019637227,
      "learning_rate": 0.0924451410860327,
      "loss": 0.1601,
      "num_input_tokens_seen": 21662016,
      "step": 25015
    },
    {
      "epoch": 11.796322489391796,
      "grad_norm": 0.0010948312701657414,
      "learning_rate": 0.09239074934769258,
      "loss": 0.2399,
      "num_input_tokens_seen": 21665904,
      "step": 25020
    },
    {
      "epoch": 11.798679867986799,
      "grad_norm": 0.0016351558733731508,
      "learning_rate": 0.09233636649342288,
      "loss": 0.2073,
      "num_input_tokens_seen": 21670480,
      "step": 25025
    },
    {
      "epoch": 11.801037246581801,
      "grad_norm": 0.002235376974567771,
      "learning_rate": 0.09228199253161017,
      "loss": 0.188,
      "num_input_tokens_seen": 21674928,
      "step": 25030
    },
    {
      "epoch": 11.803394625176804,
      "grad_norm": 0.0016001559561118484,
      "learning_rate": 0.09222762747063949,
      "loss": 0.219,
      "num_input_tokens_seen": 21678784,
      "step": 25035
    },
    {
      "epoch": 11.805752003771806,
      "grad_norm": 0.0015034000389277935,
      "learning_rate": 0.09217327131889473,
      "loss": 0.2214,
      "num_input_tokens_seen": 21682832,
      "step": 25040
    },
    {
      "epoch": 11.808109382366808,
      "grad_norm": 0.0022226108703762293,
      "learning_rate": 0.09211892408475818,
      "loss": 0.1985,
      "num_input_tokens_seen": 21687184,
      "step": 25045
    },
    {
      "epoch": 11.81046676096181,
      "grad_norm": 0.0018172124400734901,
      "learning_rate": 0.09206458577661089,
      "loss": 0.2302,
      "num_input_tokens_seen": 21691424,
      "step": 25050
    },
    {
      "epoch": 11.812824139556813,
      "grad_norm": 0.001658797962591052,
      "learning_rate": 0.09201025640283263,
      "loss": 0.2051,
      "num_input_tokens_seen": 21695584,
      "step": 25055
    },
    {
      "epoch": 11.815181518151816,
      "grad_norm": 0.0014435937628149986,
      "learning_rate": 0.09195593597180148,
      "loss": 0.1922,
      "num_input_tokens_seen": 21699520,
      "step": 25060
    },
    {
      "epoch": 11.817538896746818,
      "grad_norm": 0.0020107757300138474,
      "learning_rate": 0.09190162449189444,
      "loss": 0.2924,
      "num_input_tokens_seen": 21703520,
      "step": 25065
    },
    {
      "epoch": 11.81989627534182,
      "grad_norm": 0.0017625651089474559,
      "learning_rate": 0.09184732197148705,
      "loss": 0.1979,
      "num_input_tokens_seen": 21708512,
      "step": 25070
    },
    {
      "epoch": 11.822253653936823,
      "grad_norm": 0.0014290951658040285,
      "learning_rate": 0.09179302841895343,
      "loss": 0.2386,
      "num_input_tokens_seen": 21712752,
      "step": 25075
    },
    {
      "epoch": 11.824611032531825,
      "grad_norm": 0.001372298807837069,
      "learning_rate": 0.09173874384266625,
      "loss": 0.1671,
      "num_input_tokens_seen": 21717680,
      "step": 25080
    },
    {
      "epoch": 11.826968411126828,
      "grad_norm": 0.0011316505260765553,
      "learning_rate": 0.09168446825099695,
      "loss": 0.2017,
      "num_input_tokens_seen": 21721776,
      "step": 25085
    },
    {
      "epoch": 11.82932578972183,
      "grad_norm": 0.0014601885341107845,
      "learning_rate": 0.09163020165231545,
      "loss": 0.2058,
      "num_input_tokens_seen": 21726112,
      "step": 25090
    },
    {
      "epoch": 11.831683168316832,
      "grad_norm": 0.0016520641511306167,
      "learning_rate": 0.09157594405499044,
      "loss": 0.1986,
      "num_input_tokens_seen": 21730416,
      "step": 25095
    },
    {
      "epoch": 11.834040546911835,
      "grad_norm": 0.0016342884628102183,
      "learning_rate": 0.09152169546738899,
      "loss": 0.2146,
      "num_input_tokens_seen": 21734608,
      "step": 25100
    },
    {
      "epoch": 11.836397925506837,
      "grad_norm": 0.0018079245928674936,
      "learning_rate": 0.09146745589787698,
      "loss": 0.2024,
      "num_input_tokens_seen": 21738752,
      "step": 25105
    },
    {
      "epoch": 11.838755304101838,
      "grad_norm": 0.0017751138657331467,
      "learning_rate": 0.09141322535481891,
      "loss": 0.2382,
      "num_input_tokens_seen": 21743168,
      "step": 25110
    },
    {
      "epoch": 11.841112682696842,
      "grad_norm": 0.001870118547230959,
      "learning_rate": 0.0913590038465777,
      "loss": 0.2084,
      "num_input_tokens_seen": 21747824,
      "step": 25115
    },
    {
      "epoch": 11.843470061291843,
      "grad_norm": 0.00155125861056149,
      "learning_rate": 0.09130479138151505,
      "loss": 0.1855,
      "num_input_tokens_seen": 21751952,
      "step": 25120
    },
    {
      "epoch": 11.845827439886845,
      "grad_norm": 0.001150398631580174,
      "learning_rate": 0.09125058796799114,
      "loss": 0.1948,
      "num_input_tokens_seen": 21756544,
      "step": 25125
    },
    {
      "epoch": 11.848184818481847,
      "grad_norm": 0.0018827456515282393,
      "learning_rate": 0.09119639361436485,
      "loss": 0.2076,
      "num_input_tokens_seen": 21760368,
      "step": 25130
    },
    {
      "epoch": 11.85054219707685,
      "grad_norm": 0.0027398993261158466,
      "learning_rate": 0.09114220832899368,
      "loss": 0.234,
      "num_input_tokens_seen": 21765008,
      "step": 25135
    },
    {
      "epoch": 11.852899575671852,
      "grad_norm": 0.003395430976524949,
      "learning_rate": 0.0910880321202336,
      "loss": 0.2388,
      "num_input_tokens_seen": 21769824,
      "step": 25140
    },
    {
      "epoch": 11.855256954266855,
      "grad_norm": 0.0017761178314685822,
      "learning_rate": 0.09103386499643933,
      "loss": 0.25,
      "num_input_tokens_seen": 21773728,
      "step": 25145
    },
    {
      "epoch": 11.857614332861857,
      "grad_norm": 0.0029723462648689747,
      "learning_rate": 0.09097970696596407,
      "loss": 0.2152,
      "num_input_tokens_seen": 21778016,
      "step": 25150
    },
    {
      "epoch": 11.85997171145686,
      "grad_norm": 0.0017005990957841277,
      "learning_rate": 0.09092555803715971,
      "loss": 0.2434,
      "num_input_tokens_seen": 21782000,
      "step": 25155
    },
    {
      "epoch": 11.862329090051862,
      "grad_norm": 0.001295386697165668,
      "learning_rate": 0.0908714182183767,
      "loss": 0.2056,
      "num_input_tokens_seen": 21785840,
      "step": 25160
    },
    {
      "epoch": 11.864686468646864,
      "grad_norm": 0.0013133526081219316,
      "learning_rate": 0.090817287517964,
      "loss": 0.2227,
      "num_input_tokens_seen": 21789680,
      "step": 25165
    },
    {
      "epoch": 11.867043847241867,
      "grad_norm": 0.001896976842544973,
      "learning_rate": 0.09076316594426931,
      "loss": 0.2228,
      "num_input_tokens_seen": 21794352,
      "step": 25170
    },
    {
      "epoch": 11.869401225836869,
      "grad_norm": 0.0010368068469688296,
      "learning_rate": 0.09070905350563888,
      "loss": 0.2121,
      "num_input_tokens_seen": 21798640,
      "step": 25175
    },
    {
      "epoch": 11.871758604431871,
      "grad_norm": 0.0020882452372461557,
      "learning_rate": 0.09065495021041745,
      "loss": 0.1843,
      "num_input_tokens_seen": 21803024,
      "step": 25180
    },
    {
      "epoch": 11.874115983026874,
      "grad_norm": 0.0016612487379461527,
      "learning_rate": 0.09060085606694851,
      "loss": 0.2107,
      "num_input_tokens_seen": 21806640,
      "step": 25185
    },
    {
      "epoch": 11.876473361621876,
      "grad_norm": 0.0016265788581222296,
      "learning_rate": 0.09054677108357405,
      "loss": 0.2155,
      "num_input_tokens_seen": 21811232,
      "step": 25190
    },
    {
      "epoch": 11.878830740216879,
      "grad_norm": 0.001999874599277973,
      "learning_rate": 0.09049269526863457,
      "loss": 0.2213,
      "num_input_tokens_seen": 21815152,
      "step": 25195
    },
    {
      "epoch": 11.881188118811881,
      "grad_norm": 0.0024321454111486673,
      "learning_rate": 0.09043862863046935,
      "loss": 0.208,
      "num_input_tokens_seen": 21819728,
      "step": 25200
    },
    {
      "epoch": 11.881188118811881,
      "eval_loss": 0.2112353891134262,
      "eval_runtime": 21.9807,
      "eval_samples_per_second": 42.901,
      "eval_steps_per_second": 21.473,
      "num_input_tokens_seen": 21819728,
      "step": 25200
    },
    {
      "epoch": 11.883545497406883,
      "grad_norm": 0.0018754929769784212,
      "learning_rate": 0.09038457117741602,
      "loss": 0.2478,
      "num_input_tokens_seen": 21823952,
      "step": 25205
    },
    {
      "epoch": 11.885902876001886,
      "grad_norm": 0.002216377994045615,
      "learning_rate": 0.09033052291781099,
      "loss": 0.2261,
      "num_input_tokens_seen": 21827984,
      "step": 25210
    },
    {
      "epoch": 11.888260254596888,
      "grad_norm": 0.0019501898204907775,
      "learning_rate": 0.09027648385998926,
      "loss": 0.1974,
      "num_input_tokens_seen": 21832000,
      "step": 25215
    },
    {
      "epoch": 11.89061763319189,
      "grad_norm": 0.0015178475296124816,
      "learning_rate": 0.09022245401228417,
      "loss": 0.1926,
      "num_input_tokens_seen": 21836176,
      "step": 25220
    },
    {
      "epoch": 11.892975011786893,
      "grad_norm": 0.0013542333617806435,
      "learning_rate": 0.09016843338302792,
      "loss": 0.236,
      "num_input_tokens_seen": 21840848,
      "step": 25225
    },
    {
      "epoch": 11.895332390381895,
      "grad_norm": 0.0013553169555962086,
      "learning_rate": 0.09011442198055115,
      "loss": 0.2098,
      "num_input_tokens_seen": 21845344,
      "step": 25230
    },
    {
      "epoch": 11.897689768976898,
      "grad_norm": 0.0011615008115768433,
      "learning_rate": 0.09006041981318305,
      "loss": 0.2171,
      "num_input_tokens_seen": 21850208,
      "step": 25235
    },
    {
      "epoch": 11.9000471475719,
      "grad_norm": 0.000910402974113822,
      "learning_rate": 0.09000642688925149,
      "loss": 0.1922,
      "num_input_tokens_seen": 21854240,
      "step": 25240
    },
    {
      "epoch": 11.902404526166903,
      "grad_norm": 0.004234171472489834,
      "learning_rate": 0.0899524432170828,
      "loss": 0.22,
      "num_input_tokens_seen": 21858960,
      "step": 25245
    },
    {
      "epoch": 11.904761904761905,
      "grad_norm": 0.0010156435891985893,
      "learning_rate": 0.08989846880500196,
      "loss": 0.2259,
      "num_input_tokens_seen": 21863472,
      "step": 25250
    },
    {
      "epoch": 11.907119283356908,
      "grad_norm": 0.0018457279074937105,
      "learning_rate": 0.08984450366133256,
      "loss": 0.1856,
      "num_input_tokens_seen": 21867632,
      "step": 25255
    },
    {
      "epoch": 11.90947666195191,
      "grad_norm": 0.0013895700685679913,
      "learning_rate": 0.08979054779439664,
      "loss": 0.2409,
      "num_input_tokens_seen": 21871760,
      "step": 25260
    },
    {
      "epoch": 11.911834040546912,
      "grad_norm": 0.0027131098322570324,
      "learning_rate": 0.08973660121251485,
      "loss": 0.2499,
      "num_input_tokens_seen": 21876096,
      "step": 25265
    },
    {
      "epoch": 11.914191419141915,
      "grad_norm": 0.0011141665745526552,
      "learning_rate": 0.08968266392400655,
      "loss": 0.2126,
      "num_input_tokens_seen": 21880048,
      "step": 25270
    },
    {
      "epoch": 11.916548797736917,
      "grad_norm": 0.0009513244731351733,
      "learning_rate": 0.0896287359371894,
      "loss": 0.2199,
      "num_input_tokens_seen": 21883696,
      "step": 25275
    },
    {
      "epoch": 11.91890617633192,
      "grad_norm": 0.0018973363330587745,
      "learning_rate": 0.08957481726037989,
      "loss": 0.1932,
      "num_input_tokens_seen": 21888368,
      "step": 25280
    },
    {
      "epoch": 11.921263554926922,
      "grad_norm": 0.0013521626824513078,
      "learning_rate": 0.08952090790189286,
      "loss": 0.1854,
      "num_input_tokens_seen": 21892976,
      "step": 25285
    },
    {
      "epoch": 11.923620933521924,
      "grad_norm": 0.0014583324082195759,
      "learning_rate": 0.08946700787004187,
      "loss": 0.2366,
      "num_input_tokens_seen": 21896976,
      "step": 25290
    },
    {
      "epoch": 11.925978312116927,
      "grad_norm": 0.0013630498433485627,
      "learning_rate": 0.08941311717313899,
      "loss": 0.1757,
      "num_input_tokens_seen": 21900672,
      "step": 25295
    },
    {
      "epoch": 11.92833569071193,
      "grad_norm": 0.0014233427355065942,
      "learning_rate": 0.08935923581949483,
      "loss": 0.1879,
      "num_input_tokens_seen": 21904576,
      "step": 25300
    },
    {
      "epoch": 11.930693069306932,
      "grad_norm": 0.0018050877843052149,
      "learning_rate": 0.0893053638174185,
      "loss": 0.2378,
      "num_input_tokens_seen": 21908784,
      "step": 25305
    },
    {
      "epoch": 11.933050447901934,
      "grad_norm": 0.0016010937979444861,
      "learning_rate": 0.0892515011752179,
      "loss": 0.2355,
      "num_input_tokens_seen": 21913808,
      "step": 25310
    },
    {
      "epoch": 11.935407826496935,
      "grad_norm": 0.002102271653711796,
      "learning_rate": 0.08919764790119918,
      "loss": 0.2385,
      "num_input_tokens_seen": 21917824,
      "step": 25315
    },
    {
      "epoch": 11.937765205091937,
      "grad_norm": 0.001730751246213913,
      "learning_rate": 0.08914380400366727,
      "loss": 0.2352,
      "num_input_tokens_seen": 21922464,
      "step": 25320
    },
    {
      "epoch": 11.94012258368694,
      "grad_norm": 0.0031019810121506453,
      "learning_rate": 0.08908996949092551,
      "loss": 0.2189,
      "num_input_tokens_seen": 21928352,
      "step": 25325
    },
    {
      "epoch": 11.942479962281942,
      "grad_norm": 0.001600078190676868,
      "learning_rate": 0.08903614437127592,
      "loss": 0.2136,
      "num_input_tokens_seen": 21932704,
      "step": 25330
    },
    {
      "epoch": 11.944837340876944,
      "grad_norm": 0.0011619935976341367,
      "learning_rate": 0.088982328653019,
      "loss": 0.2252,
      "num_input_tokens_seen": 21937104,
      "step": 25335
    },
    {
      "epoch": 11.947194719471947,
      "grad_norm": 0.0013469844125211239,
      "learning_rate": 0.0889285223444538,
      "loss": 0.2136,
      "num_input_tokens_seen": 21940704,
      "step": 25340
    },
    {
      "epoch": 11.949552098066949,
      "grad_norm": 0.0015705502592027187,
      "learning_rate": 0.08887472545387787,
      "loss": 0.1868,
      "num_input_tokens_seen": 21944544,
      "step": 25345
    },
    {
      "epoch": 11.951909476661951,
      "grad_norm": 0.0013550722505897284,
      "learning_rate": 0.08882093798958751,
      "loss": 0.2047,
      "num_input_tokens_seen": 21948720,
      "step": 25350
    },
    {
      "epoch": 11.954266855256954,
      "grad_norm": 0.0015780176036059856,
      "learning_rate": 0.08876715995987726,
      "loss": 0.2067,
      "num_input_tokens_seen": 21953712,
      "step": 25355
    },
    {
      "epoch": 11.956624233851956,
      "grad_norm": 0.0014395128237083554,
      "learning_rate": 0.08871339137304052,
      "loss": 0.2557,
      "num_input_tokens_seen": 21957552,
      "step": 25360
    },
    {
      "epoch": 11.958981612446959,
      "grad_norm": 0.001709404750727117,
      "learning_rate": 0.0886596322373689,
      "loss": 0.2524,
      "num_input_tokens_seen": 21962592,
      "step": 25365
    },
    {
      "epoch": 11.961338991041961,
      "grad_norm": 0.002401844598352909,
      "learning_rate": 0.08860588256115293,
      "loss": 0.2287,
      "num_input_tokens_seen": 21966752,
      "step": 25370
    },
    {
      "epoch": 11.963696369636963,
      "grad_norm": 0.0016225959407165647,
      "learning_rate": 0.0885521423526814,
      "loss": 0.2177,
      "num_input_tokens_seen": 21971136,
      "step": 25375
    },
    {
      "epoch": 11.966053748231966,
      "grad_norm": 0.0015869117341935635,
      "learning_rate": 0.08849841162024165,
      "loss": 0.1922,
      "num_input_tokens_seen": 21976112,
      "step": 25380
    },
    {
      "epoch": 11.968411126826968,
      "grad_norm": 0.0013724610907956958,
      "learning_rate": 0.08844469037211973,
      "loss": 0.2455,
      "num_input_tokens_seen": 21980672,
      "step": 25385
    },
    {
      "epoch": 11.97076850542197,
      "grad_norm": 0.0011946968734264374,
      "learning_rate": 0.08839097861660014,
      "loss": 0.1961,
      "num_input_tokens_seen": 21985824,
      "step": 25390
    },
    {
      "epoch": 11.973125884016973,
      "grad_norm": 0.0021944670006632805,
      "learning_rate": 0.08833727636196585,
      "loss": 0.2153,
      "num_input_tokens_seen": 21989232,
      "step": 25395
    },
    {
      "epoch": 11.975483262611975,
      "grad_norm": 0.002419522497802973,
      "learning_rate": 0.08828358361649848,
      "loss": 0.1995,
      "num_input_tokens_seen": 21993120,
      "step": 25400
    },
    {
      "epoch": 11.975483262611975,
      "eval_loss": 0.2142464965581894,
      "eval_runtime": 21.906,
      "eval_samples_per_second": 43.048,
      "eval_steps_per_second": 21.547,
      "num_input_tokens_seen": 21993120,
      "step": 25400
    },
    {
      "epoch": 11.977840641206978,
      "grad_norm": 0.001564336707815528,
      "learning_rate": 0.08822990038847807,
      "loss": 0.2166,
      "num_input_tokens_seen": 21997584,
      "step": 25405
    },
    {
      "epoch": 11.98019801980198,
      "grad_norm": 0.00197174702771008,
      "learning_rate": 0.08817622668618325,
      "loss": 0.1952,
      "num_input_tokens_seen": 22001488,
      "step": 25410
    },
    {
      "epoch": 11.982555398396983,
      "grad_norm": 0.0013819426530972123,
      "learning_rate": 0.08812256251789125,
      "loss": 0.1971,
      "num_input_tokens_seen": 22006192,
      "step": 25415
    },
    {
      "epoch": 11.984912776991985,
      "grad_norm": 0.0014173618983477354,
      "learning_rate": 0.08806890789187766,
      "loss": 0.2066,
      "num_input_tokens_seen": 22010176,
      "step": 25420
    },
    {
      "epoch": 11.987270155586987,
      "grad_norm": 0.001422629109583795,
      "learning_rate": 0.08801526281641672,
      "loss": 0.2519,
      "num_input_tokens_seen": 22013968,
      "step": 25425
    },
    {
      "epoch": 11.98962753418199,
      "grad_norm": 0.0019721915014088154,
      "learning_rate": 0.0879616272997813,
      "loss": 0.2267,
      "num_input_tokens_seen": 22017920,
      "step": 25430
    },
    {
      "epoch": 11.991984912776992,
      "grad_norm": 0.0023337099701166153,
      "learning_rate": 0.08790800135024247,
      "loss": 0.2685,
      "num_input_tokens_seen": 22021968,
      "step": 25435
    },
    {
      "epoch": 11.994342291371995,
      "grad_norm": 0.0019517289474606514,
      "learning_rate": 0.08785438497607023,
      "loss": 0.2032,
      "num_input_tokens_seen": 22026832,
      "step": 25440
    },
    {
      "epoch": 11.996699669966997,
      "grad_norm": 0.001907902187667787,
      "learning_rate": 0.08780077818553277,
      "loss": 0.1898,
      "num_input_tokens_seen": 22031008,
      "step": 25445
    },
    {
      "epoch": 11.999057048562,
      "grad_norm": 0.0009369938634335995,
      "learning_rate": 0.0877471809868969,
      "loss": 0.2171,
      "num_input_tokens_seen": 22034768,
      "step": 25450
    },
    {
      "epoch": 12.001414427157002,
      "grad_norm": 0.0015117436414584517,
      "learning_rate": 0.08769359338842811,
      "loss": 0.2495,
      "num_input_tokens_seen": 22039216,
      "step": 25455
    },
    {
      "epoch": 12.003771805752004,
      "grad_norm": 0.001222338411025703,
      "learning_rate": 0.08764001539839016,
      "loss": 0.1734,
      "num_input_tokens_seen": 22043216,
      "step": 25460
    },
    {
      "epoch": 12.006129184347007,
      "grad_norm": 0.0009650439023971558,
      "learning_rate": 0.08758644702504548,
      "loss": 0.181,
      "num_input_tokens_seen": 22047968,
      "step": 25465
    },
    {
      "epoch": 12.008486562942009,
      "grad_norm": 0.0012462990125641227,
      "learning_rate": 0.0875328882766551,
      "loss": 0.2136,
      "num_input_tokens_seen": 22052000,
      "step": 25470
    },
    {
      "epoch": 12.010843941537011,
      "grad_norm": 0.004697110503911972,
      "learning_rate": 0.08747933916147828,
      "loss": 0.2,
      "num_input_tokens_seen": 22055824,
      "step": 25475
    },
    {
      "epoch": 12.013201320132014,
      "grad_norm": 0.0018836923409253359,
      "learning_rate": 0.0874257996877731,
      "loss": 0.1673,
      "num_input_tokens_seen": 22060128,
      "step": 25480
    },
    {
      "epoch": 12.015558698727016,
      "grad_norm": 0.0012637325562536716,
      "learning_rate": 0.08737226986379593,
      "loss": 0.1933,
      "num_input_tokens_seen": 22065184,
      "step": 25485
    },
    {
      "epoch": 12.017916077322019,
      "grad_norm": 0.0012028979836031795,
      "learning_rate": 0.08731874969780173,
      "loss": 0.1829,
      "num_input_tokens_seen": 22068912,
      "step": 25490
    },
    {
      "epoch": 12.020273455917021,
      "grad_norm": 0.001387283205986023,
      "learning_rate": 0.08726523919804412,
      "loss": 0.2691,
      "num_input_tokens_seen": 22072992,
      "step": 25495
    },
    {
      "epoch": 12.022630834512023,
      "grad_norm": 0.0016071537975221872,
      "learning_rate": 0.08721173837277492,
      "loss": 0.198,
      "num_input_tokens_seen": 22077312,
      "step": 25500
    },
    {
      "epoch": 12.024988213107026,
      "grad_norm": 0.002463748212903738,
      "learning_rate": 0.08715824723024479,
      "loss": 0.2326,
      "num_input_tokens_seen": 22081488,
      "step": 25505
    },
    {
      "epoch": 12.027345591702028,
      "grad_norm": 0.001992699922993779,
      "learning_rate": 0.08710476577870258,
      "loss": 0.2123,
      "num_input_tokens_seen": 22085136,
      "step": 25510
    },
    {
      "epoch": 12.029702970297029,
      "grad_norm": 0.0016763792373239994,
      "learning_rate": 0.08705129402639587,
      "loss": 0.203,
      "num_input_tokens_seen": 22089520,
      "step": 25515
    },
    {
      "epoch": 12.032060348892031,
      "grad_norm": 0.0013211703626438975,
      "learning_rate": 0.08699783198157078,
      "loss": 0.2044,
      "num_input_tokens_seen": 22094560,
      "step": 25520
    },
    {
      "epoch": 12.034417727487034,
      "grad_norm": 0.001403424539603293,
      "learning_rate": 0.08694437965247163,
      "loss": 0.2387,
      "num_input_tokens_seen": 22098560,
      "step": 25525
    },
    {
      "epoch": 12.036775106082036,
      "grad_norm": 0.001501058111898601,
      "learning_rate": 0.08689093704734165,
      "loss": 0.2034,
      "num_input_tokens_seen": 22103328,
      "step": 25530
    },
    {
      "epoch": 12.039132484677038,
      "grad_norm": 0.0016912651481106877,
      "learning_rate": 0.08683750417442222,
      "loss": 0.235,
      "num_input_tokens_seen": 22106880,
      "step": 25535
    },
    {
      "epoch": 12.04148986327204,
      "grad_norm": 0.0010738568380475044,
      "learning_rate": 0.08678408104195334,
      "loss": 0.2216,
      "num_input_tokens_seen": 22111536,
      "step": 25540
    },
    {
      "epoch": 12.043847241867043,
      "grad_norm": 0.0014230493688955903,
      "learning_rate": 0.08673066765817365,
      "loss": 0.2117,
      "num_input_tokens_seen": 22115968,
      "step": 25545
    },
    {
      "epoch": 12.046204620462046,
      "grad_norm": 0.0017954371869564056,
      "learning_rate": 0.08667726403132005,
      "loss": 0.218,
      "num_input_tokens_seen": 22119488,
      "step": 25550
    },
    {
      "epoch": 12.048561999057048,
      "grad_norm": 0.0016354575054720044,
      "learning_rate": 0.0866238701696281,
      "loss": 0.1977,
      "num_input_tokens_seen": 22123088,
      "step": 25555
    },
    {
      "epoch": 12.05091937765205,
      "grad_norm": 0.0012881620787084103,
      "learning_rate": 0.08657048608133185,
      "loss": 0.2345,
      "num_input_tokens_seen": 22128080,
      "step": 25560
    },
    {
      "epoch": 12.053276756247053,
      "grad_norm": 0.0012136102886870503,
      "learning_rate": 0.08651711177466369,
      "loss": 0.1891,
      "num_input_tokens_seen": 22132560,
      "step": 25565
    },
    {
      "epoch": 12.055634134842055,
      "grad_norm": 0.0014696959406137466,
      "learning_rate": 0.08646374725785466,
      "loss": 0.1955,
      "num_input_tokens_seen": 22137648,
      "step": 25570
    },
    {
      "epoch": 12.057991513437058,
      "grad_norm": 0.0016685780137777328,
      "learning_rate": 0.08641039253913434,
      "loss": 0.2461,
      "num_input_tokens_seen": 22141504,
      "step": 25575
    },
    {
      "epoch": 12.06034889203206,
      "grad_norm": 0.008065101690590382,
      "learning_rate": 0.08635704762673052,
      "loss": 0.1963,
      "num_input_tokens_seen": 22145664,
      "step": 25580
    },
    {
      "epoch": 12.062706270627062,
      "grad_norm": 0.0026334684807807207,
      "learning_rate": 0.08630371252886981,
      "loss": 0.2235,
      "num_input_tokens_seen": 22150448,
      "step": 25585
    },
    {
      "epoch": 12.065063649222065,
      "grad_norm": 0.0020950455218553543,
      "learning_rate": 0.08625038725377704,
      "loss": 0.1953,
      "num_input_tokens_seen": 22156480,
      "step": 25590
    },
    {
      "epoch": 12.067421027817067,
      "grad_norm": 0.002943019149824977,
      "learning_rate": 0.08619707180967566,
      "loss": 0.2224,
      "num_input_tokens_seen": 22160016,
      "step": 25595
    },
    {
      "epoch": 12.06977840641207,
      "grad_norm": 0.0020576019305735826,
      "learning_rate": 0.08614376620478768,
      "loss": 0.2066,
      "num_input_tokens_seen": 22164624,
      "step": 25600
    },
    {
      "epoch": 12.06977840641207,
      "eval_loss": 0.21620330214500427,
      "eval_runtime": 21.8809,
      "eval_samples_per_second": 43.097,
      "eval_steps_per_second": 21.571,
      "num_input_tokens_seen": 22164624,
      "step": 25600
    },
    {
      "epoch": 12.072135785007072,
      "grad_norm": 0.0020870338194072247,
      "learning_rate": 0.08609047044733344,
      "loss": 0.2072,
      "num_input_tokens_seen": 22168592,
      "step": 25605
    },
    {
      "epoch": 12.074493163602074,
      "grad_norm": 0.0016808483051136136,
      "learning_rate": 0.08603718454553168,
      "loss": 0.1757,
      "num_input_tokens_seen": 22172672,
      "step": 25610
    },
    {
      "epoch": 12.076850542197077,
      "grad_norm": 0.002127154730260372,
      "learning_rate": 0.08598390850759997,
      "loss": 0.2198,
      "num_input_tokens_seen": 22177136,
      "step": 25615
    },
    {
      "epoch": 12.07920792079208,
      "grad_norm": 0.0019992238376289606,
      "learning_rate": 0.08593064234175397,
      "loss": 0.2231,
      "num_input_tokens_seen": 22181696,
      "step": 25620
    },
    {
      "epoch": 12.081565299387082,
      "grad_norm": 0.0017763111973181367,
      "learning_rate": 0.08587738605620815,
      "loss": 0.1796,
      "num_input_tokens_seen": 22186032,
      "step": 25625
    },
    {
      "epoch": 12.083922677982084,
      "grad_norm": 0.0019198667723685503,
      "learning_rate": 0.08582413965917512,
      "loss": 0.2118,
      "num_input_tokens_seen": 22190736,
      "step": 25630
    },
    {
      "epoch": 12.086280056577086,
      "grad_norm": 0.0022524837404489517,
      "learning_rate": 0.08577090315886628,
      "loss": 0.2253,
      "num_input_tokens_seen": 22195280,
      "step": 25635
    },
    {
      "epoch": 12.088637435172089,
      "grad_norm": 0.0020613379310816526,
      "learning_rate": 0.08571767656349136,
      "loss": 0.1935,
      "num_input_tokens_seen": 22198944,
      "step": 25640
    },
    {
      "epoch": 12.090994813767091,
      "grad_norm": 0.0033502918668091297,
      "learning_rate": 0.08566445988125847,
      "loss": 0.1846,
      "num_input_tokens_seen": 22203264,
      "step": 25645
    },
    {
      "epoch": 12.093352192362094,
      "grad_norm": 0.002366072963923216,
      "learning_rate": 0.08561125312037436,
      "loss": 0.1669,
      "num_input_tokens_seen": 22208016,
      "step": 25650
    },
    {
      "epoch": 12.095709570957096,
      "grad_norm": 0.0017855376936495304,
      "learning_rate": 0.08555805628904424,
      "loss": 0.1905,
      "num_input_tokens_seen": 22212304,
      "step": 25655
    },
    {
      "epoch": 12.098066949552098,
      "grad_norm": 0.003180636325851083,
      "learning_rate": 0.08550486939547161,
      "loss": 0.2064,
      "num_input_tokens_seen": 22216672,
      "step": 25660
    },
    {
      "epoch": 12.100424328147101,
      "grad_norm": 0.0029660738073289394,
      "learning_rate": 0.08545169244785869,
      "loss": 0.2071,
      "num_input_tokens_seen": 22221920,
      "step": 25665
    },
    {
      "epoch": 12.102781706742103,
      "grad_norm": 0.0028700476977974176,
      "learning_rate": 0.08539852545440589,
      "loss": 0.2167,
      "num_input_tokens_seen": 22226640,
      "step": 25670
    },
    {
      "epoch": 12.105139085337106,
      "grad_norm": 0.0026035821065306664,
      "learning_rate": 0.08534536842331235,
      "loss": 0.2176,
      "num_input_tokens_seen": 22230672,
      "step": 25675
    },
    {
      "epoch": 12.107496463932108,
      "grad_norm": 0.0015027793124318123,
      "learning_rate": 0.08529222136277545,
      "loss": 0.2118,
      "num_input_tokens_seen": 22235568,
      "step": 25680
    },
    {
      "epoch": 12.10985384252711,
      "grad_norm": 0.001726228860206902,
      "learning_rate": 0.08523908428099125,
      "loss": 0.2051,
      "num_input_tokens_seen": 22239888,
      "step": 25685
    },
    {
      "epoch": 12.112211221122113,
      "grad_norm": 0.0020738658495247364,
      "learning_rate": 0.08518595718615402,
      "loss": 0.1858,
      "num_input_tokens_seen": 22243696,
      "step": 25690
    },
    {
      "epoch": 12.114568599717115,
      "grad_norm": 0.0029829961713403463,
      "learning_rate": 0.08513284008645675,
      "loss": 0.1969,
      "num_input_tokens_seen": 22248128,
      "step": 25695
    },
    {
      "epoch": 12.116925978312118,
      "grad_norm": 0.0016095552127808332,
      "learning_rate": 0.08507973299009065,
      "loss": 0.1828,
      "num_input_tokens_seen": 22251664,
      "step": 25700
    },
    {
      "epoch": 12.11928335690712,
      "grad_norm": 0.002045788336545229,
      "learning_rate": 0.08502663590524563,
      "loss": 0.1888,
      "num_input_tokens_seen": 22256064,
      "step": 25705
    },
    {
      "epoch": 12.121640735502123,
      "grad_norm": 0.002793314168229699,
      "learning_rate": 0.08497354884010981,
      "loss": 0.1946,
      "num_input_tokens_seen": 22261136,
      "step": 25710
    },
    {
      "epoch": 12.123998114097123,
      "grad_norm": 0.0013766068732365966,
      "learning_rate": 0.0849204718028699,
      "loss": 0.1704,
      "num_input_tokens_seen": 22265840,
      "step": 25715
    },
    {
      "epoch": 12.126355492692126,
      "grad_norm": 0.003446029732003808,
      "learning_rate": 0.08486740480171118,
      "loss": 0.215,
      "num_input_tokens_seen": 22270192,
      "step": 25720
    },
    {
      "epoch": 12.128712871287128,
      "grad_norm": 0.0038862829096615314,
      "learning_rate": 0.08481434784481706,
      "loss": 0.2974,
      "num_input_tokens_seen": 22274272,
      "step": 25725
    },
    {
      "epoch": 12.13107024988213,
      "grad_norm": 0.001949790515936911,
      "learning_rate": 0.08476130094036968,
      "loss": 0.1723,
      "num_input_tokens_seen": 22278976,
      "step": 25730
    },
    {
      "epoch": 12.133427628477133,
      "grad_norm": 0.003534179413691163,
      "learning_rate": 0.08470826409654961,
      "loss": 0.2309,
      "num_input_tokens_seen": 22283728,
      "step": 25735
    },
    {
      "epoch": 12.135785007072135,
      "grad_norm": 0.0029591440688818693,
      "learning_rate": 0.08465523732153564,
      "loss": 0.2026,
      "num_input_tokens_seen": 22287760,
      "step": 25740
    },
    {
      "epoch": 12.138142385667138,
      "grad_norm": 0.00276637333445251,
      "learning_rate": 0.08460222062350532,
      "loss": 0.1971,
      "num_input_tokens_seen": 22291472,
      "step": 25745
    },
    {
      "epoch": 12.14049976426214,
      "grad_norm": 0.002962039550766349,
      "learning_rate": 0.08454921401063442,
      "loss": 0.2495,
      "num_input_tokens_seen": 22295792,
      "step": 25750
    },
    {
      "epoch": 12.142857142857142,
      "grad_norm": 0.0019281741697341204,
      "learning_rate": 0.08449621749109716,
      "loss": 0.2251,
      "num_input_tokens_seen": 22300256,
      "step": 25755
    },
    {
      "epoch": 12.145214521452145,
      "grad_norm": 0.001813719398342073,
      "learning_rate": 0.08444323107306641,
      "loss": 0.1921,
      "num_input_tokens_seen": 22304400,
      "step": 25760
    },
    {
      "epoch": 12.147571900047147,
      "grad_norm": 0.003000825410708785,
      "learning_rate": 0.0843902547647132,
      "loss": 0.2199,
      "num_input_tokens_seen": 22308864,
      "step": 25765
    },
    {
      "epoch": 12.14992927864215,
      "grad_norm": 0.004205869045108557,
      "learning_rate": 0.0843372885742072,
      "loss": 0.2364,
      "num_input_tokens_seen": 22313200,
      "step": 25770
    },
    {
      "epoch": 12.152286657237152,
      "grad_norm": 0.002508777193725109,
      "learning_rate": 0.08428433250971652,
      "loss": 0.1821,
      "num_input_tokens_seen": 22317536,
      "step": 25775
    },
    {
      "epoch": 12.154644035832154,
      "grad_norm": 0.002177955349907279,
      "learning_rate": 0.08423138657940757,
      "loss": 0.2363,
      "num_input_tokens_seen": 22322208,
      "step": 25780
    },
    {
      "epoch": 12.157001414427157,
      "grad_norm": 0.002514116931706667,
      "learning_rate": 0.08417845079144536,
      "loss": 0.1795,
      "num_input_tokens_seen": 22327072,
      "step": 25785
    },
    {
      "epoch": 12.15935879302216,
      "grad_norm": 0.002000371925532818,
      "learning_rate": 0.08412552515399314,
      "loss": 0.1957,
      "num_input_tokens_seen": 22331024,
      "step": 25790
    },
    {
      "epoch": 12.161716171617162,
      "grad_norm": 0.0019110015127807856,
      "learning_rate": 0.08407260967521278,
      "loss": 0.2822,
      "num_input_tokens_seen": 22335088,
      "step": 25795
    },
    {
      "epoch": 12.164073550212164,
      "grad_norm": 0.0019304907182231545,
      "learning_rate": 0.08401970436326454,
      "loss": 0.1747,
      "num_input_tokens_seen": 22340064,
      "step": 25800
    },
    {
      "epoch": 12.164073550212164,
      "eval_loss": 0.21300135552883148,
      "eval_runtime": 21.8819,
      "eval_samples_per_second": 43.095,
      "eval_steps_per_second": 21.57,
      "num_input_tokens_seen": 22340064,
      "step": 25800
    },
    {
      "epoch": 12.166430928807166,
      "grad_norm": 0.0016985925612971187,
      "learning_rate": 0.08396680922630702,
      "loss": 0.2168,
      "num_input_tokens_seen": 22344432,
      "step": 25805
    },
    {
      "epoch": 12.168788307402169,
      "grad_norm": 0.0024870571214705706,
      "learning_rate": 0.08391392427249732,
      "loss": 0.1956,
      "num_input_tokens_seen": 22349040,
      "step": 25810
    },
    {
      "epoch": 12.171145685997171,
      "grad_norm": 0.001961824717000127,
      "learning_rate": 0.08386104950999107,
      "loss": 0.161,
      "num_input_tokens_seen": 22352928,
      "step": 25815
    },
    {
      "epoch": 12.173503064592174,
      "grad_norm": 0.0022859713062644005,
      "learning_rate": 0.0838081849469421,
      "loss": 0.2377,
      "num_input_tokens_seen": 22357312,
      "step": 25820
    },
    {
      "epoch": 12.175860443187176,
      "grad_norm": 0.0018203194485977292,
      "learning_rate": 0.08375533059150281,
      "loss": 0.2073,
      "num_input_tokens_seen": 22361824,
      "step": 25825
    },
    {
      "epoch": 12.178217821782178,
      "grad_norm": 0.002292245626449585,
      "learning_rate": 0.08370248645182406,
      "loss": 0.1979,
      "num_input_tokens_seen": 22365696,
      "step": 25830
    },
    {
      "epoch": 12.18057520037718,
      "grad_norm": 0.002885006135329604,
      "learning_rate": 0.083649652536055,
      "loss": 0.1799,
      "num_input_tokens_seen": 22370816,
      "step": 25835
    },
    {
      "epoch": 12.182932578972183,
      "grad_norm": 0.002114993054419756,
      "learning_rate": 0.08359682885234339,
      "loss": 0.2518,
      "num_input_tokens_seen": 22375936,
      "step": 25840
    },
    {
      "epoch": 12.185289957567186,
      "grad_norm": 0.001957348082214594,
      "learning_rate": 0.08354401540883516,
      "loss": 0.1682,
      "num_input_tokens_seen": 22379936,
      "step": 25845
    },
    {
      "epoch": 12.187647336162188,
      "grad_norm": 0.0018400036497041583,
      "learning_rate": 0.0834912122136749,
      "loss": 0.1758,
      "num_input_tokens_seen": 22385072,
      "step": 25850
    },
    {
      "epoch": 12.19000471475719,
      "grad_norm": 0.002717594848945737,
      "learning_rate": 0.0834384192750056,
      "loss": 0.2124,
      "num_input_tokens_seen": 22389072,
      "step": 25855
    },
    {
      "epoch": 12.192362093352193,
      "grad_norm": 0.002345660235732794,
      "learning_rate": 0.08338563660096844,
      "loss": 0.1789,
      "num_input_tokens_seen": 22393424,
      "step": 25860
    },
    {
      "epoch": 12.194719471947195,
      "grad_norm": 0.0026977569796144962,
      "learning_rate": 0.08333286419970329,
      "loss": 0.2616,
      "num_input_tokens_seen": 22398112,
      "step": 25865
    },
    {
      "epoch": 12.197076850542198,
      "grad_norm": 0.0020180665887892246,
      "learning_rate": 0.08328010207934824,
      "loss": 0.1984,
      "num_input_tokens_seen": 22403456,
      "step": 25870
    },
    {
      "epoch": 12.1994342291372,
      "grad_norm": 0.002175070345401764,
      "learning_rate": 0.08322735024803989,
      "loss": 0.2221,
      "num_input_tokens_seen": 22406992,
      "step": 25875
    },
    {
      "epoch": 12.201791607732202,
      "grad_norm": 0.0022843487095087767,
      "learning_rate": 0.08317460871391331,
      "loss": 0.2064,
      "num_input_tokens_seen": 22411184,
      "step": 25880
    },
    {
      "epoch": 12.204148986327205,
      "grad_norm": 0.002314218319952488,
      "learning_rate": 0.08312187748510179,
      "loss": 0.2192,
      "num_input_tokens_seen": 22416688,
      "step": 25885
    },
    {
      "epoch": 12.206506364922207,
      "grad_norm": 0.0029838450718671083,
      "learning_rate": 0.08306915656973726,
      "loss": 0.2165,
      "num_input_tokens_seen": 22421536,
      "step": 25890
    },
    {
      "epoch": 12.20886374351721,
      "grad_norm": 0.0017975930823013186,
      "learning_rate": 0.08301644597594988,
      "loss": 0.1706,
      "num_input_tokens_seen": 22426624,
      "step": 25895
    },
    {
      "epoch": 12.211221122112212,
      "grad_norm": 0.002504409058019519,
      "learning_rate": 0.08296374571186826,
      "loss": 0.2064,
      "num_input_tokens_seen": 22430448,
      "step": 25900
    },
    {
      "epoch": 12.213578500707214,
      "grad_norm": 0.00192357343621552,
      "learning_rate": 0.08291105578561955,
      "loss": 0.1826,
      "num_input_tokens_seen": 22434592,
      "step": 25905
    },
    {
      "epoch": 12.215935879302217,
      "grad_norm": 0.001866841223090887,
      "learning_rate": 0.08285837620532904,
      "loss": 0.1882,
      "num_input_tokens_seen": 22438816,
      "step": 25910
    },
    {
      "epoch": 12.218293257897217,
      "grad_norm": 0.0022493931464850903,
      "learning_rate": 0.0828057069791207,
      "loss": 0.1731,
      "num_input_tokens_seen": 22442848,
      "step": 25915
    },
    {
      "epoch": 12.22065063649222,
      "grad_norm": 0.0029825069941580296,
      "learning_rate": 0.0827530481151168,
      "loss": 0.2079,
      "num_input_tokens_seen": 22446768,
      "step": 25920
    },
    {
      "epoch": 12.223008015087222,
      "grad_norm": 0.0026687353383749723,
      "learning_rate": 0.08270039962143792,
      "loss": 0.1992,
      "num_input_tokens_seen": 22451600,
      "step": 25925
    },
    {
      "epoch": 12.225365393682225,
      "grad_norm": 0.004171442240476608,
      "learning_rate": 0.08264776150620314,
      "loss": 0.2573,
      "num_input_tokens_seen": 22455504,
      "step": 25930
    },
    {
      "epoch": 12.227722772277227,
      "grad_norm": 0.00356071419082582,
      "learning_rate": 0.08259513377753,
      "loss": 0.2198,
      "num_input_tokens_seen": 22458944,
      "step": 25935
    },
    {
      "epoch": 12.23008015087223,
      "grad_norm": 0.002362266182899475,
      "learning_rate": 0.08254251644353423,
      "loss": 0.2312,
      "num_input_tokens_seen": 22462560,
      "step": 25940
    },
    {
      "epoch": 12.232437529467232,
      "grad_norm": 0.002474409295246005,
      "learning_rate": 0.08248990951233022,
      "loss": 0.2489,
      "num_input_tokens_seen": 22466976,
      "step": 25945
    },
    {
      "epoch": 12.234794908062234,
      "grad_norm": 0.0023583583533763885,
      "learning_rate": 0.08243731299203048,
      "loss": 0.219,
      "num_input_tokens_seen": 22471664,
      "step": 25950
    },
    {
      "epoch": 12.237152286657237,
      "grad_norm": 0.0027400508988648653,
      "learning_rate": 0.08238472689074612,
      "loss": 0.1868,
      "num_input_tokens_seen": 22476448,
      "step": 25955
    },
    {
      "epoch": 12.239509665252239,
      "grad_norm": 0.002432436216622591,
      "learning_rate": 0.08233215121658666,
      "loss": 0.2162,
      "num_input_tokens_seen": 22480960,
      "step": 25960
    },
    {
      "epoch": 12.241867043847241,
      "grad_norm": 0.0034653102047741413,
      "learning_rate": 0.08227958597765982,
      "loss": 0.2647,
      "num_input_tokens_seen": 22484784,
      "step": 25965
    },
    {
      "epoch": 12.244224422442244,
      "grad_norm": 0.0019362765597179532,
      "learning_rate": 0.08222703118207181,
      "loss": 0.1985,
      "num_input_tokens_seen": 22489088,
      "step": 25970
    },
    {
      "epoch": 12.246581801037246,
      "grad_norm": 0.0019056793535128236,
      "learning_rate": 0.08217448683792734,
      "loss": 0.193,
      "num_input_tokens_seen": 22493520,
      "step": 25975
    },
    {
      "epoch": 12.248939179632249,
      "grad_norm": 0.0020754325669258833,
      "learning_rate": 0.08212195295332926,
      "loss": 0.2236,
      "num_input_tokens_seen": 22497248,
      "step": 25980
    },
    {
      "epoch": 12.251296558227251,
      "grad_norm": 0.0020912587642669678,
      "learning_rate": 0.08206942953637915,
      "loss": 0.2384,
      "num_input_tokens_seen": 22502880,
      "step": 25985
    },
    {
      "epoch": 12.253653936822253,
      "grad_norm": 0.0033875734079629183,
      "learning_rate": 0.08201691659517658,
      "loss": 0.2415,
      "num_input_tokens_seen": 22507184,
      "step": 25990
    },
    {
      "epoch": 12.256011315417256,
      "grad_norm": 0.0015324442647397518,
      "learning_rate": 0.08196441413781981,
      "loss": 0.2317,
      "num_input_tokens_seen": 22511520,
      "step": 25995
    },
    {
      "epoch": 12.258368694012258,
      "grad_norm": 0.002376343123614788,
      "learning_rate": 0.08191192217240544,
      "loss": 0.1969,
      "num_input_tokens_seen": 22515088,
      "step": 26000
    },
    {
      "epoch": 12.258368694012258,
      "eval_loss": 0.21766319870948792,
      "eval_runtime": 21.9432,
      "eval_samples_per_second": 42.975,
      "eval_steps_per_second": 21.51,
      "num_input_tokens_seen": 22515088,
      "step": 26000
    },
    {
      "epoch": 12.26072607260726,
      "grad_norm": 0.001220780424773693,
      "learning_rate": 0.08185944070702823,
      "loss": 0.2045,
      "num_input_tokens_seen": 22519088,
      "step": 26005
    },
    {
      "epoch": 12.263083451202263,
      "grad_norm": 0.0016136636259034276,
      "learning_rate": 0.08180696974978159,
      "loss": 0.1784,
      "num_input_tokens_seen": 22524144,
      "step": 26010
    },
    {
      "epoch": 12.265440829797265,
      "grad_norm": 0.0027210155967622995,
      "learning_rate": 0.08175450930875724,
      "loss": 0.2077,
      "num_input_tokens_seen": 22528816,
      "step": 26015
    },
    {
      "epoch": 12.267798208392268,
      "grad_norm": 0.0032448158599436283,
      "learning_rate": 0.08170205939204513,
      "loss": 0.2679,
      "num_input_tokens_seen": 22532704,
      "step": 26020
    },
    {
      "epoch": 12.27015558698727,
      "grad_norm": 0.0018816920928657055,
      "learning_rate": 0.08164962000773379,
      "loss": 0.1815,
      "num_input_tokens_seen": 22537456,
      "step": 26025
    },
    {
      "epoch": 12.272512965582273,
      "grad_norm": 0.0013753611128777266,
      "learning_rate": 0.08159719116390995,
      "loss": 0.1908,
      "num_input_tokens_seen": 22542176,
      "step": 26030
    },
    {
      "epoch": 12.274870344177275,
      "grad_norm": 0.002307357033714652,
      "learning_rate": 0.08154477286865887,
      "loss": 0.2189,
      "num_input_tokens_seen": 22546560,
      "step": 26035
    },
    {
      "epoch": 12.277227722772277,
      "grad_norm": 0.0016140807420015335,
      "learning_rate": 0.08149236513006404,
      "loss": 0.1898,
      "num_input_tokens_seen": 22550848,
      "step": 26040
    },
    {
      "epoch": 12.27958510136728,
      "grad_norm": 0.002089468529447913,
      "learning_rate": 0.08143996795620746,
      "loss": 0.2309,
      "num_input_tokens_seen": 22555696,
      "step": 26045
    },
    {
      "epoch": 12.281942479962282,
      "grad_norm": 0.0026942442636936903,
      "learning_rate": 0.08138758135516938,
      "loss": 0.1887,
      "num_input_tokens_seen": 22559232,
      "step": 26050
    },
    {
      "epoch": 12.284299858557285,
      "grad_norm": 0.001832404755987227,
      "learning_rate": 0.08133520533502851,
      "loss": 0.1871,
      "num_input_tokens_seen": 22563728,
      "step": 26055
    },
    {
      "epoch": 12.286657237152287,
      "grad_norm": 0.002027918351814151,
      "learning_rate": 0.08128283990386184,
      "loss": 0.1827,
      "num_input_tokens_seen": 22569664,
      "step": 26060
    },
    {
      "epoch": 12.28901461574729,
      "grad_norm": 0.0012998961610719562,
      "learning_rate": 0.08123048506974488,
      "loss": 0.1849,
      "num_input_tokens_seen": 22574096,
      "step": 26065
    },
    {
      "epoch": 12.291371994342292,
      "grad_norm": 0.0022959953639656305,
      "learning_rate": 0.08117814084075124,
      "loss": 0.2053,
      "num_input_tokens_seen": 22578384,
      "step": 26070
    },
    {
      "epoch": 12.293729372937294,
      "grad_norm": 0.0034285038709640503,
      "learning_rate": 0.08112580722495318,
      "loss": 0.2432,
      "num_input_tokens_seen": 22582400,
      "step": 26075
    },
    {
      "epoch": 12.296086751532297,
      "grad_norm": 0.0022926349192857742,
      "learning_rate": 0.08107348423042122,
      "loss": 0.1634,
      "num_input_tokens_seen": 22587168,
      "step": 26080
    },
    {
      "epoch": 12.298444130127299,
      "grad_norm": 0.004621081054210663,
      "learning_rate": 0.08102117186522413,
      "loss": 0.2267,
      "num_input_tokens_seen": 22591952,
      "step": 26085
    },
    {
      "epoch": 12.300801508722301,
      "grad_norm": 0.0017184215830639005,
      "learning_rate": 0.08096887013742916,
      "loss": 0.2481,
      "num_input_tokens_seen": 22596368,
      "step": 26090
    },
    {
      "epoch": 12.303158887317304,
      "grad_norm": 0.0016717249527573586,
      "learning_rate": 0.08091657905510198,
      "loss": 0.2535,
      "num_input_tokens_seen": 22600192,
      "step": 26095
    },
    {
      "epoch": 12.305516265912306,
      "grad_norm": 0.0027228593826293945,
      "learning_rate": 0.08086429862630642,
      "loss": 0.2217,
      "num_input_tokens_seen": 22604768,
      "step": 26100
    },
    {
      "epoch": 12.307873644507309,
      "grad_norm": 0.002259837230667472,
      "learning_rate": 0.08081202885910488,
      "loss": 0.26,
      "num_input_tokens_seen": 22608992,
      "step": 26105
    },
    {
      "epoch": 12.310231023102311,
      "grad_norm": 0.002469619968906045,
      "learning_rate": 0.08075976976155795,
      "loss": 0.2288,
      "num_input_tokens_seen": 22614000,
      "step": 26110
    },
    {
      "epoch": 12.312588401697312,
      "grad_norm": 0.0030219489708542824,
      "learning_rate": 0.08070752134172461,
      "loss": 0.2268,
      "num_input_tokens_seen": 22618224,
      "step": 26115
    },
    {
      "epoch": 12.314945780292314,
      "grad_norm": 0.002238794695585966,
      "learning_rate": 0.08065528360766229,
      "loss": 0.2349,
      "num_input_tokens_seen": 22622768,
      "step": 26120
    },
    {
      "epoch": 12.317303158887317,
      "grad_norm": 0.002145440084859729,
      "learning_rate": 0.08060305656742664,
      "loss": 0.1975,
      "num_input_tokens_seen": 22627456,
      "step": 26125
    },
    {
      "epoch": 12.319660537482319,
      "grad_norm": 0.002198203466832638,
      "learning_rate": 0.08055084022907182,
      "loss": 0.2423,
      "num_input_tokens_seen": 22631616,
      "step": 26130
    },
    {
      "epoch": 12.322017916077321,
      "grad_norm": 0.0020793797448277473,
      "learning_rate": 0.08049863460065014,
      "loss": 0.2025,
      "num_input_tokens_seen": 22636048,
      "step": 26135
    },
    {
      "epoch": 12.324375294672324,
      "grad_norm": 0.0023819527123123407,
      "learning_rate": 0.0804464396902124,
      "loss": 0.2201,
      "num_input_tokens_seen": 22640640,
      "step": 26140
    },
    {
      "epoch": 12.326732673267326,
      "grad_norm": 0.002019654493778944,
      "learning_rate": 0.08039425550580777,
      "loss": 0.2008,
      "num_input_tokens_seen": 22645728,
      "step": 26145
    },
    {
      "epoch": 12.329090051862329,
      "grad_norm": 0.0018002190627157688,
      "learning_rate": 0.08034208205548363,
      "loss": 0.2229,
      "num_input_tokens_seen": 22649968,
      "step": 26150
    },
    {
      "epoch": 12.331447430457331,
      "grad_norm": 0.0020829036366194487,
      "learning_rate": 0.08028991934728581,
      "loss": 0.1993,
      "num_input_tokens_seen": 22654016,
      "step": 26155
    },
    {
      "epoch": 12.333804809052333,
      "grad_norm": 0.0031222905963659286,
      "learning_rate": 0.0802377673892585,
      "loss": 0.2329,
      "num_input_tokens_seen": 22657888,
      "step": 26160
    },
    {
      "epoch": 12.336162187647336,
      "grad_norm": 0.0019448851235210896,
      "learning_rate": 0.0801856261894441,
      "loss": 0.2415,
      "num_input_tokens_seen": 22661968,
      "step": 26165
    },
    {
      "epoch": 12.338519566242338,
      "grad_norm": 0.0019721277058124542,
      "learning_rate": 0.08013349575588354,
      "loss": 0.217,
      "num_input_tokens_seen": 22665648,
      "step": 26170
    },
    {
      "epoch": 12.34087694483734,
      "grad_norm": 0.002281284425407648,
      "learning_rate": 0.08008137609661586,
      "loss": 0.2269,
      "num_input_tokens_seen": 22669760,
      "step": 26175
    },
    {
      "epoch": 12.343234323432343,
      "grad_norm": 0.0017794506857171655,
      "learning_rate": 0.08002926721967872,
      "loss": 0.2276,
      "num_input_tokens_seen": 22674240,
      "step": 26180
    },
    {
      "epoch": 12.345591702027345,
      "grad_norm": 0.0020533092319965363,
      "learning_rate": 0.07997716913310782,
      "loss": 0.2143,
      "num_input_tokens_seen": 22678464,
      "step": 26185
    },
    {
      "epoch": 12.347949080622348,
      "grad_norm": 0.0024264021776616573,
      "learning_rate": 0.07992508184493745,
      "loss": 0.2251,
      "num_input_tokens_seen": 22683152,
      "step": 26190
    },
    {
      "epoch": 12.35030645921735,
      "grad_norm": 0.0021962476894259453,
      "learning_rate": 0.07987300536320001,
      "loss": 0.1794,
      "num_input_tokens_seen": 22687664,
      "step": 26195
    },
    {
      "epoch": 12.352663837812353,
      "grad_norm": 0.002134514506906271,
      "learning_rate": 0.07982093969592649,
      "loss": 0.2165,
      "num_input_tokens_seen": 22692240,
      "step": 26200
    },
    {
      "epoch": 12.352663837812353,
      "eval_loss": 0.2127642035484314,
      "eval_runtime": 21.9311,
      "eval_samples_per_second": 42.998,
      "eval_steps_per_second": 21.522,
      "num_input_tokens_seen": 22692240,
      "step": 26200
    },
    {
      "epoch": 12.355021216407355,
      "grad_norm": 0.0023708236403763294,
      "learning_rate": 0.07976888485114592,
      "loss": 0.1936,
      "num_input_tokens_seen": 22697136,
      "step": 26205
    },
    {
      "epoch": 12.357378595002357,
      "grad_norm": 0.0016226297011598945,
      "learning_rate": 0.07971684083688595,
      "loss": 0.2075,
      "num_input_tokens_seen": 22701184,
      "step": 26210
    },
    {
      "epoch": 12.35973597359736,
      "grad_norm": 0.001754903350956738,
      "learning_rate": 0.0796648076611723,
      "loss": 0.1827,
      "num_input_tokens_seen": 22704944,
      "step": 26215
    },
    {
      "epoch": 12.362093352192362,
      "grad_norm": 0.0017046554712578654,
      "learning_rate": 0.07961278533202922,
      "loss": 0.1707,
      "num_input_tokens_seen": 22709488,
      "step": 26220
    },
    {
      "epoch": 12.364450730787365,
      "grad_norm": 0.0022566879633814096,
      "learning_rate": 0.07956077385747919,
      "loss": 0.2063,
      "num_input_tokens_seen": 22713616,
      "step": 26225
    },
    {
      "epoch": 12.366808109382367,
      "grad_norm": 0.0013881293125450611,
      "learning_rate": 0.079508773245543,
      "loss": 0.1053,
      "num_input_tokens_seen": 22717856,
      "step": 26230
    },
    {
      "epoch": 12.36916548797737,
      "grad_norm": 0.002244348172098398,
      "learning_rate": 0.07945678350423982,
      "loss": 0.2518,
      "num_input_tokens_seen": 22721712,
      "step": 26235
    },
    {
      "epoch": 12.371522866572372,
      "grad_norm": 0.003388015553355217,
      "learning_rate": 0.07940480464158717,
      "loss": 0.2653,
      "num_input_tokens_seen": 22726736,
      "step": 26240
    },
    {
      "epoch": 12.373880245167374,
      "grad_norm": 0.002988613909110427,
      "learning_rate": 0.07935283666560076,
      "loss": 0.2296,
      "num_input_tokens_seen": 22730848,
      "step": 26245
    },
    {
      "epoch": 12.376237623762377,
      "grad_norm": 0.0033228073734790087,
      "learning_rate": 0.07930087958429478,
      "loss": 0.2456,
      "num_input_tokens_seen": 22735056,
      "step": 26250
    },
    {
      "epoch": 12.378595002357379,
      "grad_norm": 0.0013690374325960875,
      "learning_rate": 0.07924893340568159,
      "loss": 0.2361,
      "num_input_tokens_seen": 22739872,
      "step": 26255
    },
    {
      "epoch": 12.380952380952381,
      "grad_norm": 0.0015261591179296374,
      "learning_rate": 0.07919699813777205,
      "loss": 0.217,
      "num_input_tokens_seen": 22744224,
      "step": 26260
    },
    {
      "epoch": 12.383309759547384,
      "grad_norm": 0.0014579371782019734,
      "learning_rate": 0.07914507378857515,
      "loss": 0.1969,
      "num_input_tokens_seen": 22748720,
      "step": 26265
    },
    {
      "epoch": 12.385667138142386,
      "grad_norm": 0.0016698380932211876,
      "learning_rate": 0.07909316036609822,
      "loss": 0.2281,
      "num_input_tokens_seen": 22753152,
      "step": 26270
    },
    {
      "epoch": 12.388024516737389,
      "grad_norm": 0.0016808349173516035,
      "learning_rate": 0.07904125787834704,
      "loss": 0.2179,
      "num_input_tokens_seen": 22756848,
      "step": 26275
    },
    {
      "epoch": 12.390381895332391,
      "grad_norm": 0.0018952348036691546,
      "learning_rate": 0.07898936633332569,
      "loss": 0.2201,
      "num_input_tokens_seen": 22760256,
      "step": 26280
    },
    {
      "epoch": 12.392739273927393,
      "grad_norm": 0.0018000735435634851,
      "learning_rate": 0.07893748573903635,
      "loss": 0.2027,
      "num_input_tokens_seen": 22765280,
      "step": 26285
    },
    {
      "epoch": 12.395096652522396,
      "grad_norm": 0.0015228565316647291,
      "learning_rate": 0.0788856161034798,
      "loss": 0.2082,
      "num_input_tokens_seen": 22769536,
      "step": 26290
    },
    {
      "epoch": 12.397454031117398,
      "grad_norm": 0.0019077903125435114,
      "learning_rate": 0.07883375743465487,
      "loss": 0.2172,
      "num_input_tokens_seen": 22773792,
      "step": 26295
    },
    {
      "epoch": 12.3998114097124,
      "grad_norm": 0.00255806022323668,
      "learning_rate": 0.07878190974055888,
      "loss": 0.2349,
      "num_input_tokens_seen": 22778112,
      "step": 26300
    },
    {
      "epoch": 12.402168788307403,
      "grad_norm": 0.0014959365362301469,
      "learning_rate": 0.07873007302918746,
      "loss": 0.1928,
      "num_input_tokens_seen": 22782720,
      "step": 26305
    },
    {
      "epoch": 12.404526166902404,
      "grad_norm": 0.0020221550948917866,
      "learning_rate": 0.07867824730853433,
      "loss": 0.214,
      "num_input_tokens_seen": 22787296,
      "step": 26310
    },
    {
      "epoch": 12.406883545497408,
      "grad_norm": 0.001640526344999671,
      "learning_rate": 0.07862643258659176,
      "loss": 0.2034,
      "num_input_tokens_seen": 22791312,
      "step": 26315
    },
    {
      "epoch": 12.409240924092408,
      "grad_norm": 0.0015621319180354476,
      "learning_rate": 0.07857462887135026,
      "loss": 0.2259,
      "num_input_tokens_seen": 22795920,
      "step": 26320
    },
    {
      "epoch": 12.41159830268741,
      "grad_norm": 0.0016736218240112066,
      "learning_rate": 0.0785228361707986,
      "loss": 0.2118,
      "num_input_tokens_seen": 22800512,
      "step": 26325
    },
    {
      "epoch": 12.413955681282413,
      "grad_norm": 0.0012658763444051147,
      "learning_rate": 0.07847105449292378,
      "loss": 0.2206,
      "num_input_tokens_seen": 22804544,
      "step": 26330
    },
    {
      "epoch": 12.416313059877416,
      "grad_norm": 0.002013157354667783,
      "learning_rate": 0.0784192838457113,
      "loss": 0.1792,
      "num_input_tokens_seen": 22809456,
      "step": 26335
    },
    {
      "epoch": 12.418670438472418,
      "grad_norm": 0.001976506784558296,
      "learning_rate": 0.07836752423714473,
      "loss": 0.1913,
      "num_input_tokens_seen": 22813648,
      "step": 26340
    },
    {
      "epoch": 12.42102781706742,
      "grad_norm": 0.0015136905713006854,
      "learning_rate": 0.07831577567520616,
      "loss": 0.2368,
      "num_input_tokens_seen": 22817808,
      "step": 26345
    },
    {
      "epoch": 12.423385195662423,
      "grad_norm": 0.0022582600358873606,
      "learning_rate": 0.07826403816787579,
      "loss": 0.2639,
      "num_input_tokens_seen": 22821648,
      "step": 26350
    },
    {
      "epoch": 12.425742574257425,
      "grad_norm": 0.002037752652540803,
      "learning_rate": 0.0782123117231322,
      "loss": 0.248,
      "num_input_tokens_seen": 22826560,
      "step": 26355
    },
    {
      "epoch": 12.428099952852428,
      "grad_norm": 0.0016276385867968202,
      "learning_rate": 0.07816059634895237,
      "loss": 0.2233,
      "num_input_tokens_seen": 22830096,
      "step": 26360
    },
    {
      "epoch": 12.43045733144743,
      "grad_norm": 0.002240564441308379,
      "learning_rate": 0.0781088920533113,
      "loss": 0.209,
      "num_input_tokens_seen": 22834560,
      "step": 26365
    },
    {
      "epoch": 12.432814710042432,
      "grad_norm": 0.002288860036060214,
      "learning_rate": 0.07805719884418257,
      "loss": 0.2197,
      "num_input_tokens_seen": 22839264,
      "step": 26370
    },
    {
      "epoch": 12.435172088637435,
      "grad_norm": 0.001849495805799961,
      "learning_rate": 0.07800551672953779,
      "loss": 0.2251,
      "num_input_tokens_seen": 22843296,
      "step": 26375
    },
    {
      "epoch": 12.437529467232437,
      "grad_norm": 0.002244577743113041,
      "learning_rate": 0.07795384571734709,
      "loss": 0.2233,
      "num_input_tokens_seen": 22847712,
      "step": 26380
    },
    {
      "epoch": 12.43988684582744,
      "grad_norm": 0.0023002084344625473,
      "learning_rate": 0.07790218581557883,
      "loss": 0.2363,
      "num_input_tokens_seen": 22852080,
      "step": 26385
    },
    {
      "epoch": 12.442244224422442,
      "grad_norm": 0.0015494026010856032,
      "learning_rate": 0.07785053703219949,
      "loss": 0.188,
      "num_input_tokens_seen": 22856368,
      "step": 26390
    },
    {
      "epoch": 12.444601603017444,
      "grad_norm": 0.001809404231607914,
      "learning_rate": 0.07779889937517409,
      "loss": 0.2189,
      "num_input_tokens_seen": 22860624,
      "step": 26395
    },
    {
      "epoch": 12.446958981612447,
      "grad_norm": 0.0013917406322434545,
      "learning_rate": 0.0777472728524657,
      "loss": 0.1789,
      "num_input_tokens_seen": 22864512,
      "step": 26400
    },
    {
      "epoch": 12.446958981612447,
      "eval_loss": 0.21330220997333527,
      "eval_runtime": 21.8753,
      "eval_samples_per_second": 43.108,
      "eval_steps_per_second": 21.577,
      "num_input_tokens_seen": 22864512,
      "step": 26400
    },
    {
      "epoch": 12.44931636020745,
      "grad_norm": 0.001880201743915677,
      "learning_rate": 0.07769565747203584,
      "loss": 0.1968,
      "num_input_tokens_seen": 22869232,
      "step": 26405
    },
    {
      "epoch": 12.451673738802452,
      "grad_norm": 0.002829479519277811,
      "learning_rate": 0.07764405324184427,
      "loss": 0.2003,
      "num_input_tokens_seen": 22873360,
      "step": 26410
    },
    {
      "epoch": 12.454031117397454,
      "grad_norm": 0.0025351690128445625,
      "learning_rate": 0.07759246016984889,
      "loss": 0.2588,
      "num_input_tokens_seen": 22877184,
      "step": 26415
    },
    {
      "epoch": 12.456388495992456,
      "grad_norm": 0.0013408659724518657,
      "learning_rate": 0.07754087826400609,
      "loss": 0.1747,
      "num_input_tokens_seen": 22881600,
      "step": 26420
    },
    {
      "epoch": 12.458745874587459,
      "grad_norm": 0.0014373630983754992,
      "learning_rate": 0.0774893075322705,
      "loss": 0.1879,
      "num_input_tokens_seen": 22885904,
      "step": 26425
    },
    {
      "epoch": 12.461103253182461,
      "grad_norm": 0.0033324630931019783,
      "learning_rate": 0.07743774798259484,
      "loss": 0.2156,
      "num_input_tokens_seen": 22890672,
      "step": 26430
    },
    {
      "epoch": 12.463460631777464,
      "grad_norm": 0.0013285984750837088,
      "learning_rate": 0.07738619962293032,
      "loss": 0.1527,
      "num_input_tokens_seen": 22895216,
      "step": 26435
    },
    {
      "epoch": 12.465818010372466,
      "grad_norm": 0.0014367697294801474,
      "learning_rate": 0.0773346624612264,
      "loss": 0.2004,
      "num_input_tokens_seen": 22900000,
      "step": 26440
    },
    {
      "epoch": 12.468175388967468,
      "grad_norm": 0.001604254823178053,
      "learning_rate": 0.07728313650543066,
      "loss": 0.203,
      "num_input_tokens_seen": 22904336,
      "step": 26445
    },
    {
      "epoch": 12.47053276756247,
      "grad_norm": 0.0018329507438465953,
      "learning_rate": 0.07723162176348913,
      "loss": 0.1806,
      "num_input_tokens_seen": 22909472,
      "step": 26450
    },
    {
      "epoch": 12.472890146157473,
      "grad_norm": 0.0020310545805841684,
      "learning_rate": 0.07718011824334593,
      "loss": 0.2462,
      "num_input_tokens_seen": 22913632,
      "step": 26455
    },
    {
      "epoch": 12.475247524752476,
      "grad_norm": 0.002024115761741996,
      "learning_rate": 0.07712862595294363,
      "loss": 0.2249,
      "num_input_tokens_seen": 22918016,
      "step": 26460
    },
    {
      "epoch": 12.477604903347478,
      "grad_norm": 0.001963347429409623,
      "learning_rate": 0.07707714490022301,
      "loss": 0.2252,
      "num_input_tokens_seen": 22922192,
      "step": 26465
    },
    {
      "epoch": 12.47996228194248,
      "grad_norm": 0.0019649898167699575,
      "learning_rate": 0.07702567509312298,
      "loss": 0.1976,
      "num_input_tokens_seen": 22927072,
      "step": 26470
    },
    {
      "epoch": 12.482319660537483,
      "grad_norm": 0.0016838643932715058,
      "learning_rate": 0.07697421653958098,
      "loss": 0.2054,
      "num_input_tokens_seen": 22931376,
      "step": 26475
    },
    {
      "epoch": 12.484677039132485,
      "grad_norm": 0.001809090143069625,
      "learning_rate": 0.07692276924753247,
      "loss": 0.2241,
      "num_input_tokens_seen": 22935296,
      "step": 26480
    },
    {
      "epoch": 12.487034417727488,
      "grad_norm": 0.0018332254840061069,
      "learning_rate": 0.07687133322491124,
      "loss": 0.1798,
      "num_input_tokens_seen": 22939280,
      "step": 26485
    },
    {
      "epoch": 12.48939179632249,
      "grad_norm": 0.0018497697310522199,
      "learning_rate": 0.07681990847964948,
      "loss": 0.193,
      "num_input_tokens_seen": 22943680,
      "step": 26490
    },
    {
      "epoch": 12.491749174917492,
      "grad_norm": 0.001770719769410789,
      "learning_rate": 0.0767684950196774,
      "loss": 0.2231,
      "num_input_tokens_seen": 22948912,
      "step": 26495
    },
    {
      "epoch": 12.494106553512495,
      "grad_norm": 0.0024434123188257217,
      "learning_rate": 0.0767170928529237,
      "loss": 0.2347,
      "num_input_tokens_seen": 22953376,
      "step": 26500
    },
    {
      "epoch": 12.496463932107497,
      "grad_norm": 0.0018066201591864228,
      "learning_rate": 0.07666570198731526,
      "loss": 0.1956,
      "num_input_tokens_seen": 22957952,
      "step": 26505
    },
    {
      "epoch": 12.4988213107025,
      "grad_norm": 0.002617438090965152,
      "learning_rate": 0.07661432243077708,
      "loss": 0.1901,
      "num_input_tokens_seen": 22961632,
      "step": 26510
    },
    {
      "epoch": 12.5011786892975,
      "grad_norm": 0.001422450295649469,
      "learning_rate": 0.0765629541912326,
      "loss": 0.216,
      "num_input_tokens_seen": 22966080,
      "step": 26515
    },
    {
      "epoch": 12.503536067892503,
      "grad_norm": 0.0011538423132151365,
      "learning_rate": 0.07651159727660352,
      "loss": 0.2014,
      "num_input_tokens_seen": 22970000,
      "step": 26520
    },
    {
      "epoch": 12.505893446487505,
      "grad_norm": 0.00220974232070148,
      "learning_rate": 0.07646025169480959,
      "loss": 0.2557,
      "num_input_tokens_seen": 22974560,
      "step": 26525
    },
    {
      "epoch": 12.508250825082508,
      "grad_norm": 0.0011979369446635246,
      "learning_rate": 0.07640891745376908,
      "loss": 0.1656,
      "num_input_tokens_seen": 22978912,
      "step": 26530
    },
    {
      "epoch": 12.51060820367751,
      "grad_norm": 0.0016943954396992922,
      "learning_rate": 0.07635759456139822,
      "loss": 0.2151,
      "num_input_tokens_seen": 22982944,
      "step": 26535
    },
    {
      "epoch": 12.512965582272512,
      "grad_norm": 0.0015075486153364182,
      "learning_rate": 0.0763062830256118,
      "loss": 0.2039,
      "num_input_tokens_seen": 22987088,
      "step": 26540
    },
    {
      "epoch": 12.515322960867515,
      "grad_norm": 0.001513878465630114,
      "learning_rate": 0.07625498285432258,
      "loss": 0.1705,
      "num_input_tokens_seen": 22991168,
      "step": 26545
    },
    {
      "epoch": 12.517680339462517,
      "grad_norm": 0.002128720050677657,
      "learning_rate": 0.07620369405544176,
      "loss": 0.2421,
      "num_input_tokens_seen": 22995760,
      "step": 26550
    },
    {
      "epoch": 12.52003771805752,
      "grad_norm": 0.0013931647408753633,
      "learning_rate": 0.07615241663687868,
      "loss": 0.2422,
      "num_input_tokens_seen": 22999744,
      "step": 26555
    },
    {
      "epoch": 12.522395096652522,
      "grad_norm": 0.0020236228592693806,
      "learning_rate": 0.07610115060654106,
      "loss": 0.2069,
      "num_input_tokens_seen": 23004016,
      "step": 26560
    },
    {
      "epoch": 12.524752475247524,
      "grad_norm": 0.0019758623093366623,
      "learning_rate": 0.07604989597233458,
      "loss": 0.2194,
      "num_input_tokens_seen": 23008080,
      "step": 26565
    },
    {
      "epoch": 12.527109853842527,
      "grad_norm": 0.002313564997166395,
      "learning_rate": 0.07599865274216352,
      "loss": 0.2015,
      "num_input_tokens_seen": 23012832,
      "step": 26570
    },
    {
      "epoch": 12.52946723243753,
      "grad_norm": 0.0024272941518574953,
      "learning_rate": 0.07594742092393013,
      "loss": 0.2187,
      "num_input_tokens_seen": 23016672,
      "step": 26575
    },
    {
      "epoch": 12.531824611032532,
      "grad_norm": 0.002012553857639432,
      "learning_rate": 0.07589620052553503,
      "loss": 0.1738,
      "num_input_tokens_seen": 23020640,
      "step": 26580
    },
    {
      "epoch": 12.534181989627534,
      "grad_norm": 0.0015155673027038574,
      "learning_rate": 0.0758449915548771,
      "loss": 0.1883,
      "num_input_tokens_seen": 23025248,
      "step": 26585
    },
    {
      "epoch": 12.536539368222536,
      "grad_norm": 0.002881759312003851,
      "learning_rate": 0.07579379401985332,
      "loss": 0.1804,
      "num_input_tokens_seen": 23029424,
      "step": 26590
    },
    {
      "epoch": 12.538896746817539,
      "grad_norm": 0.0017582689179107547,
      "learning_rate": 0.07574260792835905,
      "loss": 0.2301,
      "num_input_tokens_seen": 23033120,
      "step": 26595
    },
    {
      "epoch": 12.541254125412541,
      "grad_norm": 0.003071241080760956,
      "learning_rate": 0.07569143328828784,
      "loss": 0.2362,
      "num_input_tokens_seen": 23037568,
      "step": 26600
    },
    {
      "epoch": 12.541254125412541,
      "eval_loss": 0.21578361093997955,
      "eval_runtime": 21.9259,
      "eval_samples_per_second": 43.009,
      "eval_steps_per_second": 21.527,
      "num_input_tokens_seen": 23037568,
      "step": 26600
    },
    {
      "epoch": 12.543611504007544,
      "grad_norm": 0.0013814913108944893,
      "learning_rate": 0.0756402701075314,
      "loss": 0.1902,
      "num_input_tokens_seen": 23041920,
      "step": 26605
    },
    {
      "epoch": 12.545968882602546,
      "grad_norm": 0.002518513472750783,
      "learning_rate": 0.07558911839397982,
      "loss": 0.2408,
      "num_input_tokens_seen": 23046816,
      "step": 26610
    },
    {
      "epoch": 12.548326261197548,
      "grad_norm": 0.001633631531149149,
      "learning_rate": 0.07553797815552123,
      "loss": 0.1819,
      "num_input_tokens_seen": 23051152,
      "step": 26615
    },
    {
      "epoch": 12.55068363979255,
      "grad_norm": 0.0013821866596117616,
      "learning_rate": 0.07548684940004222,
      "loss": 0.1794,
      "num_input_tokens_seen": 23055440,
      "step": 26620
    },
    {
      "epoch": 12.553041018387553,
      "grad_norm": 0.001595496665686369,
      "learning_rate": 0.07543573213542744,
      "loss": 0.2264,
      "num_input_tokens_seen": 23060224,
      "step": 26625
    },
    {
      "epoch": 12.555398396982556,
      "grad_norm": 0.0018699313513934612,
      "learning_rate": 0.0753846263695597,
      "loss": 0.234,
      "num_input_tokens_seen": 23064128,
      "step": 26630
    },
    {
      "epoch": 12.557755775577558,
      "grad_norm": 0.002152504865080118,
      "learning_rate": 0.07533353211032029,
      "loss": 0.2276,
      "num_input_tokens_seen": 23068336,
      "step": 26635
    },
    {
      "epoch": 12.56011315417256,
      "grad_norm": 0.0016267058672383428,
      "learning_rate": 0.07528244936558857,
      "loss": 0.2361,
      "num_input_tokens_seen": 23072048,
      "step": 26640
    },
    {
      "epoch": 12.562470532767563,
      "grad_norm": 0.0019499401096254587,
      "learning_rate": 0.07523137814324206,
      "loss": 0.1856,
      "num_input_tokens_seen": 23076064,
      "step": 26645
    },
    {
      "epoch": 12.564827911362565,
      "grad_norm": 0.0027052853256464005,
      "learning_rate": 0.07518031845115672,
      "loss": 0.2081,
      "num_input_tokens_seen": 23080496,
      "step": 26650
    },
    {
      "epoch": 12.567185289957568,
      "grad_norm": 0.0013792923418805003,
      "learning_rate": 0.07512927029720647,
      "loss": 0.1964,
      "num_input_tokens_seen": 23084560,
      "step": 26655
    },
    {
      "epoch": 12.56954266855257,
      "grad_norm": 0.002021533902734518,
      "learning_rate": 0.0750782336892636,
      "loss": 0.2307,
      "num_input_tokens_seen": 23089008,
      "step": 26660
    },
    {
      "epoch": 12.571900047147572,
      "grad_norm": 0.0021919459104537964,
      "learning_rate": 0.0750272086351987,
      "loss": 0.1996,
      "num_input_tokens_seen": 23093248,
      "step": 26665
    },
    {
      "epoch": 12.574257425742575,
      "grad_norm": 0.0017514448845759034,
      "learning_rate": 0.07497619514288031,
      "loss": 0.2021,
      "num_input_tokens_seen": 23097120,
      "step": 26670
    },
    {
      "epoch": 12.576614804337577,
      "grad_norm": 0.00236295023933053,
      "learning_rate": 0.07492519322017545,
      "loss": 0.2405,
      "num_input_tokens_seen": 23101600,
      "step": 26675
    },
    {
      "epoch": 12.57897218293258,
      "grad_norm": 0.00184104114305228,
      "learning_rate": 0.0748742028749493,
      "loss": 0.2031,
      "num_input_tokens_seen": 23105680,
      "step": 26680
    },
    {
      "epoch": 12.581329561527582,
      "grad_norm": 0.0016915567684918642,
      "learning_rate": 0.0748232241150651,
      "loss": 0.2356,
      "num_input_tokens_seen": 23110032,
      "step": 26685
    },
    {
      "epoch": 12.583686940122584,
      "grad_norm": 0.002853140700608492,
      "learning_rate": 0.07477225694838453,
      "loss": 0.2263,
      "num_input_tokens_seen": 23114640,
      "step": 26690
    },
    {
      "epoch": 12.586044318717587,
      "grad_norm": 0.002332896227017045,
      "learning_rate": 0.07472130138276731,
      "loss": 0.1981,
      "num_input_tokens_seen": 23118864,
      "step": 26695
    },
    {
      "epoch": 12.58840169731259,
      "grad_norm": 0.0020513543859124184,
      "learning_rate": 0.07467035742607138,
      "loss": 0.224,
      "num_input_tokens_seen": 23123488,
      "step": 26700
    },
    {
      "epoch": 12.590759075907592,
      "grad_norm": 0.001656990614719689,
      "learning_rate": 0.07461942508615303,
      "loss": 0.1947,
      "num_input_tokens_seen": 23129072,
      "step": 26705
    },
    {
      "epoch": 12.593116454502592,
      "grad_norm": 0.002102404832839966,
      "learning_rate": 0.07456850437086657,
      "loss": 0.2194,
      "num_input_tokens_seen": 23133040,
      "step": 26710
    },
    {
      "epoch": 12.595473833097596,
      "grad_norm": 0.0015924021136015654,
      "learning_rate": 0.07451759528806468,
      "loss": 0.1832,
      "num_input_tokens_seen": 23136720,
      "step": 26715
    },
    {
      "epoch": 12.597831211692597,
      "grad_norm": 0.001781273982487619,
      "learning_rate": 0.0744666978455982,
      "loss": 0.1898,
      "num_input_tokens_seen": 23141696,
      "step": 26720
    },
    {
      "epoch": 12.6001885902876,
      "grad_norm": 0.0023413351736962795,
      "learning_rate": 0.07441581205131609,
      "loss": 0.2027,
      "num_input_tokens_seen": 23146192,
      "step": 26725
    },
    {
      "epoch": 12.602545968882602,
      "grad_norm": 0.0028869984671473503,
      "learning_rate": 0.07436493791306566,
      "loss": 0.2211,
      "num_input_tokens_seen": 23150560,
      "step": 26730
    },
    {
      "epoch": 12.604903347477604,
      "grad_norm": 0.0022023487836122513,
      "learning_rate": 0.07431407543869223,
      "loss": 0.1944,
      "num_input_tokens_seen": 23154944,
      "step": 26735
    },
    {
      "epoch": 12.607260726072607,
      "grad_norm": 0.0018141076434403658,
      "learning_rate": 0.0742632246360395,
      "loss": 0.2063,
      "num_input_tokens_seen": 23159616,
      "step": 26740
    },
    {
      "epoch": 12.609618104667609,
      "grad_norm": 0.003723234636709094,
      "learning_rate": 0.07421238551294934,
      "loss": 0.1927,
      "num_input_tokens_seen": 23163760,
      "step": 26745
    },
    {
      "epoch": 12.611975483262611,
      "grad_norm": 0.0014296281151473522,
      "learning_rate": 0.07416155807726171,
      "loss": 0.1693,
      "num_input_tokens_seen": 23167824,
      "step": 26750
    },
    {
      "epoch": 12.614332861857614,
      "grad_norm": 0.0032542904373258352,
      "learning_rate": 0.07411074233681492,
      "loss": 0.2506,
      "num_input_tokens_seen": 23172384,
      "step": 26755
    },
    {
      "epoch": 12.616690240452616,
      "grad_norm": 0.002967148320749402,
      "learning_rate": 0.07405993829944528,
      "loss": 0.2282,
      "num_input_tokens_seen": 23177040,
      "step": 26760
    },
    {
      "epoch": 12.619047619047619,
      "grad_norm": 0.001895320601761341,
      "learning_rate": 0.07400914597298755,
      "loss": 0.1868,
      "num_input_tokens_seen": 23181424,
      "step": 26765
    },
    {
      "epoch": 12.621404997642621,
      "grad_norm": 0.0026605986058712006,
      "learning_rate": 0.07395836536527445,
      "loss": 0.2495,
      "num_input_tokens_seen": 23185456,
      "step": 26770
    },
    {
      "epoch": 12.623762376237623,
      "grad_norm": 0.0017383563099429011,
      "learning_rate": 0.07390759648413696,
      "loss": 0.226,
      "num_input_tokens_seen": 23188736,
      "step": 26775
    },
    {
      "epoch": 12.626119754832626,
      "grad_norm": 0.00409194128587842,
      "learning_rate": 0.07385683933740435,
      "loss": 0.214,
      "num_input_tokens_seen": 23193088,
      "step": 26780
    },
    {
      "epoch": 12.628477133427628,
      "grad_norm": 0.002773512387648225,
      "learning_rate": 0.07380609393290402,
      "loss": 0.1984,
      "num_input_tokens_seen": 23196800,
      "step": 26785
    },
    {
      "epoch": 12.63083451202263,
      "grad_norm": 0.0015155128203332424,
      "learning_rate": 0.07375536027846147,
      "loss": 0.1838,
      "num_input_tokens_seen": 23200784,
      "step": 26790
    },
    {
      "epoch": 12.633191890617633,
      "grad_norm": 0.002389263827353716,
      "learning_rate": 0.07370463838190057,
      "loss": 0.2299,
      "num_input_tokens_seen": 23204192,
      "step": 26795
    },
    {
      "epoch": 12.635549269212635,
      "grad_norm": 0.0018419284606352448,
      "learning_rate": 0.07365392825104317,
      "loss": 0.2457,
      "num_input_tokens_seen": 23207936,
      "step": 26800
    },
    {
      "epoch": 12.635549269212635,
      "eval_loss": 0.21454894542694092,
      "eval_runtime": 21.8902,
      "eval_samples_per_second": 43.079,
      "eval_steps_per_second": 21.562,
      "num_input_tokens_seen": 23207936,
      "step": 26800
    },
    {
      "epoch": 12.637906647807638,
      "grad_norm": 0.002074162010103464,
      "learning_rate": 0.07360322989370945,
      "loss": 0.1903,
      "num_input_tokens_seen": 23212432,
      "step": 26805
    },
    {
      "epoch": 12.64026402640264,
      "grad_norm": 0.0026755088474601507,
      "learning_rate": 0.07355254331771781,
      "loss": 0.1992,
      "num_input_tokens_seen": 23217584,
      "step": 26810
    },
    {
      "epoch": 12.642621404997643,
      "grad_norm": 0.0016151960007846355,
      "learning_rate": 0.07350186853088461,
      "loss": 0.224,
      "num_input_tokens_seen": 23221536,
      "step": 26815
    },
    {
      "epoch": 12.644978783592645,
      "grad_norm": 0.002171027008444071,
      "learning_rate": 0.07345120554102462,
      "loss": 0.2193,
      "num_input_tokens_seen": 23225360,
      "step": 26820
    },
    {
      "epoch": 12.647336162187647,
      "grad_norm": 0.001508059212937951,
      "learning_rate": 0.07340055435595079,
      "loss": 0.1834,
      "num_input_tokens_seen": 23230320,
      "step": 26825
    },
    {
      "epoch": 12.64969354078265,
      "grad_norm": 0.0021907805930823088,
      "learning_rate": 0.07334991498347401,
      "loss": 0.2095,
      "num_input_tokens_seen": 23235296,
      "step": 26830
    },
    {
      "epoch": 12.652050919377652,
      "grad_norm": 0.0019480556948110461,
      "learning_rate": 0.07329928743140365,
      "loss": 0.1687,
      "num_input_tokens_seen": 23240240,
      "step": 26835
    },
    {
      "epoch": 12.654408297972655,
      "grad_norm": 0.002430003834888339,
      "learning_rate": 0.07324867170754705,
      "loss": 0.2201,
      "num_input_tokens_seen": 23245088,
      "step": 26840
    },
    {
      "epoch": 12.656765676567657,
      "grad_norm": 0.0024915605317801237,
      "learning_rate": 0.07319806781970974,
      "loss": 0.2353,
      "num_input_tokens_seen": 23249744,
      "step": 26845
    },
    {
      "epoch": 12.65912305516266,
      "grad_norm": 0.001959258457645774,
      "learning_rate": 0.07314747577569555,
      "loss": 0.1696,
      "num_input_tokens_seen": 23253984,
      "step": 26850
    },
    {
      "epoch": 12.661480433757662,
      "grad_norm": 0.0032744049094617367,
      "learning_rate": 0.07309689558330636,
      "loss": 0.2231,
      "num_input_tokens_seen": 23258464,
      "step": 26855
    },
    {
      "epoch": 12.663837812352664,
      "grad_norm": 0.0025698316749185324,
      "learning_rate": 0.0730463272503423,
      "loss": 0.2689,
      "num_input_tokens_seen": 23263792,
      "step": 26860
    },
    {
      "epoch": 12.666195190947667,
      "grad_norm": 0.0014490815810859203,
      "learning_rate": 0.07299577078460168,
      "loss": 0.1932,
      "num_input_tokens_seen": 23267936,
      "step": 26865
    },
    {
      "epoch": 12.668552569542669,
      "grad_norm": 0.002938194666057825,
      "learning_rate": 0.07294522619388083,
      "loss": 0.1809,
      "num_input_tokens_seen": 23272624,
      "step": 26870
    },
    {
      "epoch": 12.670909948137671,
      "grad_norm": 0.0018101558089256287,
      "learning_rate": 0.07289469348597452,
      "loss": 0.2002,
      "num_input_tokens_seen": 23277280,
      "step": 26875
    },
    {
      "epoch": 12.673267326732674,
      "grad_norm": 0.0018491678638383746,
      "learning_rate": 0.07284417266867535,
      "loss": 0.2283,
      "num_input_tokens_seen": 23281984,
      "step": 26880
    },
    {
      "epoch": 12.675624705327676,
      "grad_norm": 0.0018341557588428259,
      "learning_rate": 0.07279366374977439,
      "loss": 0.1911,
      "num_input_tokens_seen": 23286160,
      "step": 26885
    },
    {
      "epoch": 12.677982083922679,
      "grad_norm": 0.002989309374243021,
      "learning_rate": 0.07274316673706074,
      "loss": 0.1943,
      "num_input_tokens_seen": 23290800,
      "step": 26890
    },
    {
      "epoch": 12.680339462517681,
      "grad_norm": 0.0018026836914941669,
      "learning_rate": 0.07269268163832161,
      "loss": 0.2088,
      "num_input_tokens_seen": 23295328,
      "step": 26895
    },
    {
      "epoch": 12.682696841112683,
      "grad_norm": 0.0016587793361395597,
      "learning_rate": 0.07264220846134248,
      "loss": 0.2021,
      "num_input_tokens_seen": 23299088,
      "step": 26900
    },
    {
      "epoch": 12.685054219707686,
      "grad_norm": 0.0017360051861032844,
      "learning_rate": 0.07259174721390699,
      "loss": 0.1907,
      "num_input_tokens_seen": 23302800,
      "step": 26905
    },
    {
      "epoch": 12.687411598302688,
      "grad_norm": 0.0017065188148990273,
      "learning_rate": 0.07254129790379686,
      "loss": 0.1765,
      "num_input_tokens_seen": 23306864,
      "step": 26910
    },
    {
      "epoch": 12.689768976897689,
      "grad_norm": 0.0022955574095249176,
      "learning_rate": 0.072490860538792,
      "loss": 0.2329,
      "num_input_tokens_seen": 23311168,
      "step": 26915
    },
    {
      "epoch": 12.692126355492691,
      "grad_norm": 0.002466501435264945,
      "learning_rate": 0.07244043512667042,
      "loss": 0.1832,
      "num_input_tokens_seen": 23315200,
      "step": 26920
    },
    {
      "epoch": 12.694483734087694,
      "grad_norm": 0.0017437050119042397,
      "learning_rate": 0.07239002167520843,
      "loss": 0.2237,
      "num_input_tokens_seen": 23320064,
      "step": 26925
    },
    {
      "epoch": 12.696841112682696,
      "grad_norm": 0.0024176998995244503,
      "learning_rate": 0.07233962019218045,
      "loss": 0.2412,
      "num_input_tokens_seen": 23324496,
      "step": 26930
    },
    {
      "epoch": 12.699198491277699,
      "grad_norm": 0.0020343330688774586,
      "learning_rate": 0.07228923068535892,
      "loss": 0.2154,
      "num_input_tokens_seen": 23328528,
      "step": 26935
    },
    {
      "epoch": 12.701555869872701,
      "grad_norm": 0.0023078371305018663,
      "learning_rate": 0.0722388531625146,
      "loss": 0.1613,
      "num_input_tokens_seen": 23333472,
      "step": 26940
    },
    {
      "epoch": 12.703913248467703,
      "grad_norm": 0.002338495571166277,
      "learning_rate": 0.07218848763141639,
      "loss": 0.217,
      "num_input_tokens_seen": 23337152,
      "step": 26945
    },
    {
      "epoch": 12.706270627062706,
      "grad_norm": 0.0021293286699801683,
      "learning_rate": 0.07213813409983118,
      "loss": 0.1737,
      "num_input_tokens_seen": 23340848,
      "step": 26950
    },
    {
      "epoch": 12.708628005657708,
      "grad_norm": 0.0024564340710639954,
      "learning_rate": 0.0720877925755242,
      "loss": 0.2302,
      "num_input_tokens_seen": 23344752,
      "step": 26955
    },
    {
      "epoch": 12.71098538425271,
      "grad_norm": 0.0028243325650691986,
      "learning_rate": 0.07203746306625866,
      "loss": 0.2097,
      "num_input_tokens_seen": 23348912,
      "step": 26960
    },
    {
      "epoch": 12.713342762847713,
      "grad_norm": 0.0031891961116343737,
      "learning_rate": 0.07198714557979606,
      "loss": 0.2204,
      "num_input_tokens_seen": 23353008,
      "step": 26965
    },
    {
      "epoch": 12.715700141442715,
      "grad_norm": 0.002105789491906762,
      "learning_rate": 0.07193684012389602,
      "loss": 0.2038,
      "num_input_tokens_seen": 23356736,
      "step": 26970
    },
    {
      "epoch": 12.718057520037718,
      "grad_norm": 0.002072356641292572,
      "learning_rate": 0.07188654670631621,
      "loss": 0.1635,
      "num_input_tokens_seen": 23360784,
      "step": 26975
    },
    {
      "epoch": 12.72041489863272,
      "grad_norm": 0.0025962605141103268,
      "learning_rate": 0.07183626533481258,
      "loss": 0.2559,
      "num_input_tokens_seen": 23364848,
      "step": 26980
    },
    {
      "epoch": 12.722772277227723,
      "grad_norm": 0.0018610811093822122,
      "learning_rate": 0.07178599601713909,
      "loss": 0.1958,
      "num_input_tokens_seen": 23368880,
      "step": 26985
    },
    {
      "epoch": 12.725129655822725,
      "grad_norm": 0.0016213153721764684,
      "learning_rate": 0.07173573876104786,
      "loss": 0.2169,
      "num_input_tokens_seen": 23372928,
      "step": 26990
    },
    {
      "epoch": 12.727487034417727,
      "grad_norm": 0.002885885536670685,
      "learning_rate": 0.0716854935742893,
      "loss": 0.1832,
      "num_input_tokens_seen": 23377088,
      "step": 26995
    },
    {
      "epoch": 12.72984441301273,
      "grad_norm": 0.002385099418461323,
      "learning_rate": 0.07163526046461174,
      "loss": 0.2113,
      "num_input_tokens_seen": 23381376,
      "step": 27000
    },
    {
      "epoch": 12.72984441301273,
      "eval_loss": 0.21353647112846375,
      "eval_runtime": 21.8983,
      "eval_samples_per_second": 43.063,
      "eval_steps_per_second": 21.554,
      "num_input_tokens_seen": 23381376,
      "step": 27000
    },
    {
      "epoch": 12.732201791607732,
      "grad_norm": 0.002014672150835395,
      "learning_rate": 0.07158503943976181,
      "loss": 0.2034,
      "num_input_tokens_seen": 23385232,
      "step": 27005
    },
    {
      "epoch": 12.734559170202735,
      "grad_norm": 0.0025827244389802217,
      "learning_rate": 0.07153483050748427,
      "loss": 0.2632,
      "num_input_tokens_seen": 23389280,
      "step": 27010
    },
    {
      "epoch": 12.736916548797737,
      "grad_norm": 0.002819756744429469,
      "learning_rate": 0.07148463367552188,
      "loss": 0.2314,
      "num_input_tokens_seen": 23393296,
      "step": 27015
    },
    {
      "epoch": 12.73927392739274,
      "grad_norm": 0.0018407930620014668,
      "learning_rate": 0.07143444895161565,
      "loss": 0.1972,
      "num_input_tokens_seen": 23397456,
      "step": 27020
    },
    {
      "epoch": 12.741631305987742,
      "grad_norm": 0.002018267521634698,
      "learning_rate": 0.07138427634350476,
      "loss": 0.2007,
      "num_input_tokens_seen": 23402000,
      "step": 27025
    },
    {
      "epoch": 12.743988684582744,
      "grad_norm": 0.001716488040983677,
      "learning_rate": 0.07133411585892636,
      "loss": 0.2098,
      "num_input_tokens_seen": 23406592,
      "step": 27030
    },
    {
      "epoch": 12.746346063177747,
      "grad_norm": 0.0017438913928344846,
      "learning_rate": 0.07128396750561593,
      "loss": 0.2056,
      "num_input_tokens_seen": 23410864,
      "step": 27035
    },
    {
      "epoch": 12.748703441772749,
      "grad_norm": 0.002013581804931164,
      "learning_rate": 0.07123383129130685,
      "loss": 0.2201,
      "num_input_tokens_seen": 23415120,
      "step": 27040
    },
    {
      "epoch": 12.751060820367751,
      "grad_norm": 0.0027766625862568617,
      "learning_rate": 0.07118370722373084,
      "loss": 0.1961,
      "num_input_tokens_seen": 23419696,
      "step": 27045
    },
    {
      "epoch": 12.753418198962754,
      "grad_norm": 0.001869815867394209,
      "learning_rate": 0.07113359531061769,
      "loss": 0.2415,
      "num_input_tokens_seen": 23423648,
      "step": 27050
    },
    {
      "epoch": 12.755775577557756,
      "grad_norm": 0.0015165242366492748,
      "learning_rate": 0.07108349555969525,
      "loss": 0.1995,
      "num_input_tokens_seen": 23428000,
      "step": 27055
    },
    {
      "epoch": 12.758132956152759,
      "grad_norm": 0.003131876001134515,
      "learning_rate": 0.07103340797868944,
      "loss": 0.2052,
      "num_input_tokens_seen": 23432368,
      "step": 27060
    },
    {
      "epoch": 12.760490334747761,
      "grad_norm": 0.002230751793831587,
      "learning_rate": 0.07098333257532453,
      "loss": 0.213,
      "num_input_tokens_seen": 23436848,
      "step": 27065
    },
    {
      "epoch": 12.762847713342763,
      "grad_norm": 0.003149432595819235,
      "learning_rate": 0.07093326935732269,
      "loss": 0.2126,
      "num_input_tokens_seen": 23441152,
      "step": 27070
    },
    {
      "epoch": 12.765205091937766,
      "grad_norm": 0.003091956954449415,
      "learning_rate": 0.0708832183324044,
      "loss": 0.1969,
      "num_input_tokens_seen": 23445696,
      "step": 27075
    },
    {
      "epoch": 12.767562470532768,
      "grad_norm": 0.0018901799339801073,
      "learning_rate": 0.07083317950828799,
      "loss": 0.2028,
      "num_input_tokens_seen": 23449888,
      "step": 27080
    },
    {
      "epoch": 12.76991984912777,
      "grad_norm": 0.0018480228027328849,
      "learning_rate": 0.0707831528926902,
      "loss": 0.1898,
      "num_input_tokens_seen": 23455232,
      "step": 27085
    },
    {
      "epoch": 12.772277227722773,
      "grad_norm": 0.0026183989830315113,
      "learning_rate": 0.07073313849332578,
      "loss": 0.1789,
      "num_input_tokens_seen": 23460448,
      "step": 27090
    },
    {
      "epoch": 12.774634606317775,
      "grad_norm": 0.003312087384983897,
      "learning_rate": 0.07068313631790749,
      "loss": 0.2408,
      "num_input_tokens_seen": 23465152,
      "step": 27095
    },
    {
      "epoch": 12.776991984912778,
      "grad_norm": 0.0024705049581825733,
      "learning_rate": 0.07063314637414632,
      "loss": 0.1859,
      "num_input_tokens_seen": 23470240,
      "step": 27100
    },
    {
      "epoch": 12.77934936350778,
      "grad_norm": 0.003400129731744528,
      "learning_rate": 0.07058316866975144,
      "loss": 0.2243,
      "num_input_tokens_seen": 23474368,
      "step": 27105
    },
    {
      "epoch": 12.78170674210278,
      "grad_norm": 0.0029374314472079277,
      "learning_rate": 0.0705332032124299,
      "loss": 0.2457,
      "num_input_tokens_seen": 23478880,
      "step": 27110
    },
    {
      "epoch": 12.784064120697785,
      "grad_norm": 0.0025456161238253117,
      "learning_rate": 0.0704832500098871,
      "loss": 0.1994,
      "num_input_tokens_seen": 23483744,
      "step": 27115
    },
    {
      "epoch": 12.786421499292786,
      "grad_norm": 0.0023624091409146786,
      "learning_rate": 0.07043330906982641,
      "loss": 0.1926,
      "num_input_tokens_seen": 23487248,
      "step": 27120
    },
    {
      "epoch": 12.788778877887788,
      "grad_norm": 0.0023349819239228964,
      "learning_rate": 0.07038338039994936,
      "loss": 0.205,
      "num_input_tokens_seen": 23490976,
      "step": 27125
    },
    {
      "epoch": 12.79113625648279,
      "grad_norm": 0.0022836208809167147,
      "learning_rate": 0.07033346400795562,
      "loss": 0.2208,
      "num_input_tokens_seen": 23495472,
      "step": 27130
    },
    {
      "epoch": 12.793493635077793,
      "grad_norm": 0.002792704850435257,
      "learning_rate": 0.07028355990154282,
      "loss": 0.2486,
      "num_input_tokens_seen": 23498832,
      "step": 27135
    },
    {
      "epoch": 12.795851013672795,
      "grad_norm": 0.002568983007222414,
      "learning_rate": 0.07023366808840685,
      "loss": 0.2386,
      "num_input_tokens_seen": 23503216,
      "step": 27140
    },
    {
      "epoch": 12.798208392267798,
      "grad_norm": 0.0036334767937660217,
      "learning_rate": 0.07018378857624172,
      "loss": 0.2123,
      "num_input_tokens_seen": 23507504,
      "step": 27145
    },
    {
      "epoch": 12.8005657708628,
      "grad_norm": 0.0019087977707386017,
      "learning_rate": 0.0701339213727394,
      "loss": 0.2175,
      "num_input_tokens_seen": 23511968,
      "step": 27150
    },
    {
      "epoch": 12.802923149457802,
      "grad_norm": 0.0031531427521258593,
      "learning_rate": 0.07008406648559008,
      "loss": 0.1895,
      "num_input_tokens_seen": 23516080,
      "step": 27155
    },
    {
      "epoch": 12.805280528052805,
      "grad_norm": 0.002929659327492118,
      "learning_rate": 0.07003422392248196,
      "loss": 0.2235,
      "num_input_tokens_seen": 23520768,
      "step": 27160
    },
    {
      "epoch": 12.807637906647807,
      "grad_norm": 0.0028042495250701904,
      "learning_rate": 0.06998439369110142,
      "loss": 0.2215,
      "num_input_tokens_seen": 23524736,
      "step": 27165
    },
    {
      "epoch": 12.80999528524281,
      "grad_norm": 0.002739723538979888,
      "learning_rate": 0.06993457579913295,
      "loss": 0.2181,
      "num_input_tokens_seen": 23529264,
      "step": 27170
    },
    {
      "epoch": 12.812352663837812,
      "grad_norm": 0.00191792706027627,
      "learning_rate": 0.06988477025425903,
      "loss": 0.178,
      "num_input_tokens_seen": 23533248,
      "step": 27175
    },
    {
      "epoch": 12.814710042432814,
      "grad_norm": 0.002396202879026532,
      "learning_rate": 0.06983497706416032,
      "loss": 0.1985,
      "num_input_tokens_seen": 23537536,
      "step": 27180
    },
    {
      "epoch": 12.817067421027817,
      "grad_norm": 0.001986737595871091,
      "learning_rate": 0.0697851962365156,
      "loss": 0.1983,
      "num_input_tokens_seen": 23541824,
      "step": 27185
    },
    {
      "epoch": 12.81942479962282,
      "grad_norm": 0.0017289062961935997,
      "learning_rate": 0.06973542777900163,
      "loss": 0.1755,
      "num_input_tokens_seen": 23545600,
      "step": 27190
    },
    {
      "epoch": 12.821782178217822,
      "grad_norm": 0.002296967664733529,
      "learning_rate": 0.06968567169929342,
      "loss": 0.2391,
      "num_input_tokens_seen": 23549280,
      "step": 27195
    },
    {
      "epoch": 12.824139556812824,
      "grad_norm": 0.0018754933262243867,
      "learning_rate": 0.06963592800506392,
      "loss": 0.2103,
      "num_input_tokens_seen": 23553008,
      "step": 27200
    },
    {
      "epoch": 12.824139556812824,
      "eval_loss": 0.21196311712265015,
      "eval_runtime": 21.8781,
      "eval_samples_per_second": 43.102,
      "eval_steps_per_second": 21.574,
      "num_input_tokens_seen": 23553008,
      "step": 27200
    },
    {
      "epoch": 12.826496935407826,
      "grad_norm": 0.0032809290569275618,
      "learning_rate": 0.06958619670398417,
      "loss": 0.1927,
      "num_input_tokens_seen": 23556704,
      "step": 27205
    },
    {
      "epoch": 12.828854314002829,
      "grad_norm": 0.0019201061222702265,
      "learning_rate": 0.0695364778037235,
      "loss": 0.2059,
      "num_input_tokens_seen": 23561360,
      "step": 27210
    },
    {
      "epoch": 12.831211692597831,
      "grad_norm": 0.0026406196411699057,
      "learning_rate": 0.06948677131194907,
      "loss": 0.233,
      "num_input_tokens_seen": 23565888,
      "step": 27215
    },
    {
      "epoch": 12.833569071192834,
      "grad_norm": 0.0021884581074118614,
      "learning_rate": 0.06943707723632629,
      "loss": 0.2519,
      "num_input_tokens_seen": 23570400,
      "step": 27220
    },
    {
      "epoch": 12.835926449787836,
      "grad_norm": 0.0023663865868002176,
      "learning_rate": 0.06938739558451867,
      "loss": 0.198,
      "num_input_tokens_seen": 23574432,
      "step": 27225
    },
    {
      "epoch": 12.838283828382838,
      "grad_norm": 0.002473373431712389,
      "learning_rate": 0.06933772636418763,
      "loss": 0.1936,
      "num_input_tokens_seen": 23578592,
      "step": 27230
    },
    {
      "epoch": 12.84064120697784,
      "grad_norm": 0.0026748350355774164,
      "learning_rate": 0.06928806958299293,
      "loss": 0.2172,
      "num_input_tokens_seen": 23582784,
      "step": 27235
    },
    {
      "epoch": 12.842998585572843,
      "grad_norm": 0.0023759088944643736,
      "learning_rate": 0.06923842524859211,
      "loss": 0.192,
      "num_input_tokens_seen": 23586944,
      "step": 27240
    },
    {
      "epoch": 12.845355964167846,
      "grad_norm": 0.0029762990307062864,
      "learning_rate": 0.06918879336864105,
      "loss": 0.2277,
      "num_input_tokens_seen": 23591328,
      "step": 27245
    },
    {
      "epoch": 12.847713342762848,
      "grad_norm": 0.0017382703954353929,
      "learning_rate": 0.06913917395079362,
      "loss": 0.1775,
      "num_input_tokens_seen": 23595264,
      "step": 27250
    },
    {
      "epoch": 12.85007072135785,
      "grad_norm": 0.0016341244336217642,
      "learning_rate": 0.0690895670027017,
      "loss": 0.197,
      "num_input_tokens_seen": 23598960,
      "step": 27255
    },
    {
      "epoch": 12.852428099952853,
      "grad_norm": 0.0019064786611124873,
      "learning_rate": 0.06903997253201531,
      "loss": 0.2717,
      "num_input_tokens_seen": 23603184,
      "step": 27260
    },
    {
      "epoch": 12.854785478547855,
      "grad_norm": 0.0023457540664821863,
      "learning_rate": 0.06899039054638263,
      "loss": 0.2053,
      "num_input_tokens_seen": 23606944,
      "step": 27265
    },
    {
      "epoch": 12.857142857142858,
      "grad_norm": 0.00204528053291142,
      "learning_rate": 0.06894082105344976,
      "loss": 0.2006,
      "num_input_tokens_seen": 23611808,
      "step": 27270
    },
    {
      "epoch": 12.85950023573786,
      "grad_norm": 0.0022474732249975204,
      "learning_rate": 0.06889126406086087,
      "loss": 0.2229,
      "num_input_tokens_seen": 23616368,
      "step": 27275
    },
    {
      "epoch": 12.861857614332862,
      "grad_norm": 0.0023429959546774626,
      "learning_rate": 0.0688417195762584,
      "loss": 0.2112,
      "num_input_tokens_seen": 23620800,
      "step": 27280
    },
    {
      "epoch": 12.864214992927865,
      "grad_norm": 0.002810097299516201,
      "learning_rate": 0.06879218760728262,
      "loss": 0.1987,
      "num_input_tokens_seen": 23625136,
      "step": 27285
    },
    {
      "epoch": 12.866572371522867,
      "grad_norm": 0.0024655533488839865,
      "learning_rate": 0.06874266816157207,
      "loss": 0.2319,
      "num_input_tokens_seen": 23628720,
      "step": 27290
    },
    {
      "epoch": 12.86892975011787,
      "grad_norm": 0.0019841566681861877,
      "learning_rate": 0.06869316124676321,
      "loss": 0.2009,
      "num_input_tokens_seen": 23632992,
      "step": 27295
    },
    {
      "epoch": 12.871287128712872,
      "grad_norm": 0.0019208498997613788,
      "learning_rate": 0.06864366687049062,
      "loss": 0.1955,
      "num_input_tokens_seen": 23636832,
      "step": 27300
    },
    {
      "epoch": 12.873644507307874,
      "grad_norm": 0.0022023788187652826,
      "learning_rate": 0.06859418504038704,
      "loss": 0.1895,
      "num_input_tokens_seen": 23640752,
      "step": 27305
    },
    {
      "epoch": 12.876001885902877,
      "grad_norm": 0.0022295555099844933,
      "learning_rate": 0.06854471576408311,
      "loss": 0.2238,
      "num_input_tokens_seen": 23645184,
      "step": 27310
    },
    {
      "epoch": 12.878359264497877,
      "grad_norm": 0.0022809377405792475,
      "learning_rate": 0.06849525904920767,
      "loss": 0.2113,
      "num_input_tokens_seen": 23649392,
      "step": 27315
    },
    {
      "epoch": 12.88071664309288,
      "grad_norm": 0.002749457722529769,
      "learning_rate": 0.06844581490338748,
      "loss": 0.257,
      "num_input_tokens_seen": 23653520,
      "step": 27320
    },
    {
      "epoch": 12.883074021687882,
      "grad_norm": 0.0019319417187944055,
      "learning_rate": 0.06839638333424752,
      "loss": 0.2058,
      "num_input_tokens_seen": 23657696,
      "step": 27325
    },
    {
      "epoch": 12.885431400282885,
      "grad_norm": 0.0022181631065905094,
      "learning_rate": 0.06834696434941082,
      "loss": 0.2195,
      "num_input_tokens_seen": 23661856,
      "step": 27330
    },
    {
      "epoch": 12.887788778877887,
      "grad_norm": 0.001990054501220584,
      "learning_rate": 0.06829755795649824,
      "loss": 0.1871,
      "num_input_tokens_seen": 23665904,
      "step": 27335
    },
    {
      "epoch": 12.89014615747289,
      "grad_norm": 0.0023155089002102613,
      "learning_rate": 0.06824816416312904,
      "loss": 0.2242,
      "num_input_tokens_seen": 23670528,
      "step": 27340
    },
    {
      "epoch": 12.892503536067892,
      "grad_norm": 0.002733101835474372,
      "learning_rate": 0.06819878297692027,
      "loss": 0.2156,
      "num_input_tokens_seen": 23674544,
      "step": 27345
    },
    {
      "epoch": 12.894860914662894,
      "grad_norm": 0.0035316413268446922,
      "learning_rate": 0.0681494144054871,
      "loss": 0.22,
      "num_input_tokens_seen": 23679680,
      "step": 27350
    },
    {
      "epoch": 12.897218293257897,
      "grad_norm": 0.0023094178177416325,
      "learning_rate": 0.06810005845644286,
      "loss": 0.1782,
      "num_input_tokens_seen": 23683472,
      "step": 27355
    },
    {
      "epoch": 12.899575671852899,
      "grad_norm": 0.0029660502914339304,
      "learning_rate": 0.06805071513739878,
      "loss": 0.2245,
      "num_input_tokens_seen": 23687216,
      "step": 27360
    },
    {
      "epoch": 12.901933050447902,
      "grad_norm": 0.001864682068116963,
      "learning_rate": 0.06800138445596428,
      "loss": 0.1536,
      "num_input_tokens_seen": 23691392,
      "step": 27365
    },
    {
      "epoch": 12.904290429042904,
      "grad_norm": 0.002606457332149148,
      "learning_rate": 0.06795206641974678,
      "loss": 0.223,
      "num_input_tokens_seen": 23695776,
      "step": 27370
    },
    {
      "epoch": 12.906647807637906,
      "grad_norm": 0.0019833019468933344,
      "learning_rate": 0.06790276103635169,
      "loss": 0.2224,
      "num_input_tokens_seen": 23700640,
      "step": 27375
    },
    {
      "epoch": 12.909005186232909,
      "grad_norm": 0.003156360937282443,
      "learning_rate": 0.0678534683133826,
      "loss": 0.1868,
      "num_input_tokens_seen": 23705168,
      "step": 27380
    },
    {
      "epoch": 12.911362564827911,
      "grad_norm": 0.002458041999489069,
      "learning_rate": 0.06780418825844095,
      "loss": 0.1863,
      "num_input_tokens_seen": 23709104,
      "step": 27385
    },
    {
      "epoch": 12.913719943422914,
      "grad_norm": 0.0029289769008755684,
      "learning_rate": 0.0677549208791264,
      "loss": 0.2004,
      "num_input_tokens_seen": 23713520,
      "step": 27390
    },
    {
      "epoch": 12.916077322017916,
      "grad_norm": 0.002219975693151355,
      "learning_rate": 0.06770566618303668,
      "loss": 0.1759,
      "num_input_tokens_seen": 23718160,
      "step": 27395
    },
    {
      "epoch": 12.918434700612918,
      "grad_norm": 0.0023491783067584038,
      "learning_rate": 0.06765642417776736,
      "loss": 0.2114,
      "num_input_tokens_seen": 23722608,
      "step": 27400
    },
    {
      "epoch": 12.918434700612918,
      "eval_loss": 0.2126309722661972,
      "eval_runtime": 21.9486,
      "eval_samples_per_second": 42.964,
      "eval_steps_per_second": 21.505,
      "num_input_tokens_seen": 23722608,
      "step": 27400
    },
    {
      "epoch": 12.92079207920792,
      "grad_norm": 0.002085567219182849,
      "learning_rate": 0.0676071948709122,
      "loss": 0.165,
      "num_input_tokens_seen": 23727392,
      "step": 27405
    },
    {
      "epoch": 12.923149457802923,
      "grad_norm": 0.0024900645948946476,
      "learning_rate": 0.06755797827006307,
      "loss": 0.1768,
      "num_input_tokens_seen": 23733056,
      "step": 27410
    },
    {
      "epoch": 12.925506836397926,
      "grad_norm": 0.0029378936160355806,
      "learning_rate": 0.06750877438280974,
      "loss": 0.1891,
      "num_input_tokens_seen": 23737488,
      "step": 27415
    },
    {
      "epoch": 12.927864214992928,
      "grad_norm": 0.0023847802076488733,
      "learning_rate": 0.06745958321673998,
      "loss": 0.185,
      "num_input_tokens_seen": 23741792,
      "step": 27420
    },
    {
      "epoch": 12.93022159358793,
      "grad_norm": 0.003579799784347415,
      "learning_rate": 0.0674104047794398,
      "loss": 0.1517,
      "num_input_tokens_seen": 23746512,
      "step": 27425
    },
    {
      "epoch": 12.932578972182933,
      "grad_norm": 0.003797441255301237,
      "learning_rate": 0.06736123907849303,
      "loss": 0.2027,
      "num_input_tokens_seen": 23751040,
      "step": 27430
    },
    {
      "epoch": 12.934936350777935,
      "grad_norm": 0.0036071850918233395,
      "learning_rate": 0.06731208612148178,
      "loss": 0.2427,
      "num_input_tokens_seen": 23755440,
      "step": 27435
    },
    {
      "epoch": 12.937293729372938,
      "grad_norm": 0.0027280524373054504,
      "learning_rate": 0.0672629459159859,
      "loss": 0.1759,
      "num_input_tokens_seen": 23759200,
      "step": 27440
    },
    {
      "epoch": 12.93965110796794,
      "grad_norm": 0.0033215878065675497,
      "learning_rate": 0.0672138184695835,
      "loss": 0.2033,
      "num_input_tokens_seen": 23763392,
      "step": 27445
    },
    {
      "epoch": 12.942008486562942,
      "grad_norm": 0.0025842224713414907,
      "learning_rate": 0.0671647037898507,
      "loss": 0.1926,
      "num_input_tokens_seen": 23767696,
      "step": 27450
    },
    {
      "epoch": 12.944365865157945,
      "grad_norm": 0.0027955754194408655,
      "learning_rate": 0.0671156018843615,
      "loss": 0.2323,
      "num_input_tokens_seen": 23772096,
      "step": 27455
    },
    {
      "epoch": 12.946723243752947,
      "grad_norm": 0.002187287900596857,
      "learning_rate": 0.06706651276068812,
      "loss": 0.1715,
      "num_input_tokens_seen": 23775344,
      "step": 27460
    },
    {
      "epoch": 12.94908062234795,
      "grad_norm": 0.0036013885401189327,
      "learning_rate": 0.06701743642640064,
      "loss": 0.2236,
      "num_input_tokens_seen": 23779792,
      "step": 27465
    },
    {
      "epoch": 12.951438000942952,
      "grad_norm": 0.0035697331186383963,
      "learning_rate": 0.06696837288906729,
      "loss": 0.2243,
      "num_input_tokens_seen": 23783744,
      "step": 27470
    },
    {
      "epoch": 12.953795379537954,
      "grad_norm": 0.004157577641308308,
      "learning_rate": 0.06691932215625432,
      "loss": 0.2654,
      "num_input_tokens_seen": 23788208,
      "step": 27475
    },
    {
      "epoch": 12.956152758132957,
      "grad_norm": 0.004324604757130146,
      "learning_rate": 0.06687028423552589,
      "loss": 0.1986,
      "num_input_tokens_seen": 23792496,
      "step": 27480
    },
    {
      "epoch": 12.95851013672796,
      "grad_norm": 0.002359729493036866,
      "learning_rate": 0.06682125913444435,
      "loss": 0.2119,
      "num_input_tokens_seen": 23797072,
      "step": 27485
    },
    {
      "epoch": 12.960867515322962,
      "grad_norm": 0.00412893807515502,
      "learning_rate": 0.0667722468605699,
      "loss": 0.1946,
      "num_input_tokens_seen": 23800976,
      "step": 27490
    },
    {
      "epoch": 12.963224893917964,
      "grad_norm": 0.0035797799937427044,
      "learning_rate": 0.06672324742146094,
      "loss": 0.2202,
      "num_input_tokens_seen": 23805072,
      "step": 27495
    },
    {
      "epoch": 12.965582272512966,
      "grad_norm": 0.0029910383746027946,
      "learning_rate": 0.06667426082467373,
      "loss": 0.2304,
      "num_input_tokens_seen": 23809424,
      "step": 27500
    },
    {
      "epoch": 12.967939651107969,
      "grad_norm": 0.0034758036490529776,
      "learning_rate": 0.0666252870777626,
      "loss": 0.1952,
      "num_input_tokens_seen": 23813184,
      "step": 27505
    },
    {
      "epoch": 12.97029702970297,
      "grad_norm": 0.003667723387479782,
      "learning_rate": 0.06657632618827995,
      "loss": 0.2135,
      "num_input_tokens_seen": 23817008,
      "step": 27510
    },
    {
      "epoch": 12.972654408297974,
      "grad_norm": 0.002123629441484809,
      "learning_rate": 0.06652737816377623,
      "loss": 0.2019,
      "num_input_tokens_seen": 23821856,
      "step": 27515
    },
    {
      "epoch": 12.975011786892974,
      "grad_norm": 0.002284109126776457,
      "learning_rate": 0.06647844301179971,
      "loss": 0.2108,
      "num_input_tokens_seen": 23826368,
      "step": 27520
    },
    {
      "epoch": 12.977369165487977,
      "grad_norm": 0.003243680577725172,
      "learning_rate": 0.06642952073989689,
      "loss": 0.1777,
      "num_input_tokens_seen": 23830928,
      "step": 27525
    },
    {
      "epoch": 12.979726544082979,
      "grad_norm": 0.003789369948208332,
      "learning_rate": 0.06638061135561223,
      "loss": 0.2729,
      "num_input_tokens_seen": 23835248,
      "step": 27530
    },
    {
      "epoch": 12.982083922677981,
      "grad_norm": 0.002655148971825838,
      "learning_rate": 0.06633171486648808,
      "loss": 0.2047,
      "num_input_tokens_seen": 23838992,
      "step": 27535
    },
    {
      "epoch": 12.984441301272984,
      "grad_norm": 0.0018445730675011873,
      "learning_rate": 0.06628283128006499,
      "loss": 0.1943,
      "num_input_tokens_seen": 23843072,
      "step": 27540
    },
    {
      "epoch": 12.986798679867986,
      "grad_norm": 0.0027047304902225733,
      "learning_rate": 0.0662339606038813,
      "loss": 0.2306,
      "num_input_tokens_seen": 23847296,
      "step": 27545
    },
    {
      "epoch": 12.989156058462989,
      "grad_norm": 0.0015737038338556886,
      "learning_rate": 0.06618510284547358,
      "loss": 0.2255,
      "num_input_tokens_seen": 23851488,
      "step": 27550
    },
    {
      "epoch": 12.991513437057991,
      "grad_norm": 0.0033369131851941347,
      "learning_rate": 0.06613625801237633,
      "loss": 0.2074,
      "num_input_tokens_seen": 23855248,
      "step": 27555
    },
    {
      "epoch": 12.993870815652993,
      "grad_norm": 0.002806410426273942,
      "learning_rate": 0.066087426112122,
      "loss": 0.1795,
      "num_input_tokens_seen": 23860064,
      "step": 27560
    },
    {
      "epoch": 12.996228194247996,
      "grad_norm": 0.002588457427918911,
      "learning_rate": 0.06603860715224101,
      "loss": 0.2237,
      "num_input_tokens_seen": 23864064,
      "step": 27565
    },
    {
      "epoch": 12.998585572842998,
      "grad_norm": 0.0031592040322721004,
      "learning_rate": 0.06598980114026198,
      "loss": 0.1716,
      "num_input_tokens_seen": 23867856,
      "step": 27570
    },
    {
      "epoch": 13.000942951438,
      "grad_norm": 0.0021865188609808683,
      "learning_rate": 0.06594100808371128,
      "loss": 0.1977,
      "num_input_tokens_seen": 23872560,
      "step": 27575
    },
    {
      "epoch": 13.003300330033003,
      "grad_norm": 0.0021558599546551704,
      "learning_rate": 0.06589222799011357,
      "loss": 0.1991,
      "num_input_tokens_seen": 23877344,
      "step": 27580
    },
    {
      "epoch": 13.005657708628005,
      "grad_norm": 0.0025004108902066946,
      "learning_rate": 0.0658434608669912,
      "loss": 0.2588,
      "num_input_tokens_seen": 23881312,
      "step": 27585
    },
    {
      "epoch": 13.008015087223008,
      "grad_norm": 0.001766430796124041,
      "learning_rate": 0.06579470672186473,
      "loss": 0.2005,
      "num_input_tokens_seen": 23885248,
      "step": 27590
    },
    {
      "epoch": 13.01037246581801,
      "grad_norm": 0.003001140197739005,
      "learning_rate": 0.06574596556225275,
      "loss": 0.2548,
      "num_input_tokens_seen": 23889312,
      "step": 27595
    },
    {
      "epoch": 13.012729844413013,
      "grad_norm": 0.0022847128566354513,
      "learning_rate": 0.06569723739567161,
      "loss": 0.1868,
      "num_input_tokens_seen": 23892928,
      "step": 27600
    },
    {
      "epoch": 13.012729844413013,
      "eval_loss": 0.21349774301052094,
      "eval_runtime": 21.9029,
      "eval_samples_per_second": 43.054,
      "eval_steps_per_second": 21.55,
      "num_input_tokens_seen": 23892928,
      "step": 27600
    },
    {
      "epoch": 13.015087223008015,
      "grad_norm": 0.003682458773255348,
      "learning_rate": 0.06564852222963588,
      "loss": 0.2169,
      "num_input_tokens_seen": 23897472,
      "step": 27605
    },
    {
      "epoch": 13.017444601603017,
      "grad_norm": 0.0031703754793852568,
      "learning_rate": 0.06559982007165813,
      "loss": 0.1611,
      "num_input_tokens_seen": 23901392,
      "step": 27610
    },
    {
      "epoch": 13.01980198019802,
      "grad_norm": 0.0021520324517041445,
      "learning_rate": 0.06555113092924868,
      "loss": 0.1584,
      "num_input_tokens_seen": 23905616,
      "step": 27615
    },
    {
      "epoch": 13.022159358793022,
      "grad_norm": 0.003042110474780202,
      "learning_rate": 0.06550245480991615,
      "loss": 0.2312,
      "num_input_tokens_seen": 23909008,
      "step": 27620
    },
    {
      "epoch": 13.024516737388025,
      "grad_norm": 0.0022913780994713306,
      "learning_rate": 0.0654537917211669,
      "loss": 0.2369,
      "num_input_tokens_seen": 23912496,
      "step": 27625
    },
    {
      "epoch": 13.026874115983027,
      "grad_norm": 0.004723566118627787,
      "learning_rate": 0.0654051416705055,
      "loss": 0.2197,
      "num_input_tokens_seen": 23916320,
      "step": 27630
    },
    {
      "epoch": 13.02923149457803,
      "grad_norm": 0.0023805885575711727,
      "learning_rate": 0.06535650466543427,
      "loss": 0.1884,
      "num_input_tokens_seen": 23920992,
      "step": 27635
    },
    {
      "epoch": 13.031588873173032,
      "grad_norm": 0.002575993537902832,
      "learning_rate": 0.0653078807134538,
      "loss": 0.1624,
      "num_input_tokens_seen": 23925792,
      "step": 27640
    },
    {
      "epoch": 13.033946251768034,
      "grad_norm": 0.002734814304858446,
      "learning_rate": 0.06525926982206236,
      "loss": 0.1832,
      "num_input_tokens_seen": 23930672,
      "step": 27645
    },
    {
      "epoch": 13.036303630363037,
      "grad_norm": 0.002183645498007536,
      "learning_rate": 0.06521067199875648,
      "loss": 0.2052,
      "num_input_tokens_seen": 23934848,
      "step": 27650
    },
    {
      "epoch": 13.038661008958039,
      "grad_norm": 0.0024088930804282427,
      "learning_rate": 0.06516208725103047,
      "loss": 0.192,
      "num_input_tokens_seen": 23939488,
      "step": 27655
    },
    {
      "epoch": 13.041018387553041,
      "grad_norm": 0.003235182724893093,
      "learning_rate": 0.06511351558637678,
      "loss": 0.1942,
      "num_input_tokens_seen": 23943120,
      "step": 27660
    },
    {
      "epoch": 13.043375766148044,
      "grad_norm": 0.0020766339730471373,
      "learning_rate": 0.06506495701228569,
      "loss": 0.1974,
      "num_input_tokens_seen": 23947728,
      "step": 27665
    },
    {
      "epoch": 13.045733144743046,
      "grad_norm": 0.0026147235184907913,
      "learning_rate": 0.06501641153624559,
      "loss": 0.2065,
      "num_input_tokens_seen": 23952336,
      "step": 27670
    },
    {
      "epoch": 13.048090523338049,
      "grad_norm": 0.004071047995239496,
      "learning_rate": 0.06496787916574286,
      "loss": 0.2149,
      "num_input_tokens_seen": 23956560,
      "step": 27675
    },
    {
      "epoch": 13.050447901933051,
      "grad_norm": 0.002962120808660984,
      "learning_rate": 0.06491935990826168,
      "loss": 0.2225,
      "num_input_tokens_seen": 23961408,
      "step": 27680
    },
    {
      "epoch": 13.052805280528053,
      "grad_norm": 0.0036243193317204714,
      "learning_rate": 0.0648708537712844,
      "loss": 0.2179,
      "num_input_tokens_seen": 23965504,
      "step": 27685
    },
    {
      "epoch": 13.055162659123056,
      "grad_norm": 0.0031917677260935307,
      "learning_rate": 0.06482236076229132,
      "loss": 0.1602,
      "num_input_tokens_seen": 23969472,
      "step": 27690
    },
    {
      "epoch": 13.057520037718058,
      "grad_norm": 0.002462289994582534,
      "learning_rate": 0.06477388088876056,
      "loss": 0.183,
      "num_input_tokens_seen": 23973680,
      "step": 27695
    },
    {
      "epoch": 13.05987741631306,
      "grad_norm": 0.004937973339110613,
      "learning_rate": 0.06472541415816846,
      "loss": 0.1957,
      "num_input_tokens_seen": 23978592,
      "step": 27700
    },
    {
      "epoch": 13.062234794908063,
      "grad_norm": 0.0024522114545106888,
      "learning_rate": 0.06467696057798909,
      "loss": 0.1769,
      "num_input_tokens_seen": 23983520,
      "step": 27705
    },
    {
      "epoch": 13.064592173503065,
      "grad_norm": 0.004625607747584581,
      "learning_rate": 0.0646285201556946,
      "loss": 0.2528,
      "num_input_tokens_seen": 23987232,
      "step": 27710
    },
    {
      "epoch": 13.066949552098066,
      "grad_norm": 0.004562231712043285,
      "learning_rate": 0.06458009289875521,
      "loss": 0.249,
      "num_input_tokens_seen": 23991632,
      "step": 27715
    },
    {
      "epoch": 13.069306930693068,
      "grad_norm": 0.0035593973007053137,
      "learning_rate": 0.0645316788146389,
      "loss": 0.1765,
      "num_input_tokens_seen": 23996000,
      "step": 27720
    },
    {
      "epoch": 13.07166430928807,
      "grad_norm": 0.0035469927825033665,
      "learning_rate": 0.06448327791081175,
      "loss": 0.2435,
      "num_input_tokens_seen": 23999632,
      "step": 27725
    },
    {
      "epoch": 13.074021687883073,
      "grad_norm": 0.0033889133483171463,
      "learning_rate": 0.0644348901947379,
      "loss": 0.1905,
      "num_input_tokens_seen": 24004368,
      "step": 27730
    },
    {
      "epoch": 13.076379066478076,
      "grad_norm": 0.0035366276279091835,
      "learning_rate": 0.06438651567387917,
      "loss": 0.1634,
      "num_input_tokens_seen": 24008608,
      "step": 27735
    },
    {
      "epoch": 13.078736445073078,
      "grad_norm": 0.002637267578393221,
      "learning_rate": 0.0643381543556957,
      "loss": 0.1627,
      "num_input_tokens_seen": 24013280,
      "step": 27740
    },
    {
      "epoch": 13.08109382366808,
      "grad_norm": 0.003385243471711874,
      "learning_rate": 0.06428980624764526,
      "loss": 0.1905,
      "num_input_tokens_seen": 24017664,
      "step": 27745
    },
    {
      "epoch": 13.083451202263083,
      "grad_norm": 0.002465470228344202,
      "learning_rate": 0.06424147135718378,
      "loss": 0.2304,
      "num_input_tokens_seen": 24022112,
      "step": 27750
    },
    {
      "epoch": 13.085808580858085,
      "grad_norm": 0.0031016876455396414,
      "learning_rate": 0.06419314969176519,
      "loss": 0.2024,
      "num_input_tokens_seen": 24026032,
      "step": 27755
    },
    {
      "epoch": 13.088165959453088,
      "grad_norm": 0.002490459708496928,
      "learning_rate": 0.06414484125884118,
      "loss": 0.2081,
      "num_input_tokens_seen": 24030272,
      "step": 27760
    },
    {
      "epoch": 13.09052333804809,
      "grad_norm": 0.0033876849338412285,
      "learning_rate": 0.06409654606586157,
      "loss": 0.1862,
      "num_input_tokens_seen": 24034352,
      "step": 27765
    },
    {
      "epoch": 13.092880716643092,
      "grad_norm": 0.002968772314488888,
      "learning_rate": 0.06404826412027415,
      "loss": 0.1723,
      "num_input_tokens_seen": 24038336,
      "step": 27770
    },
    {
      "epoch": 13.095238095238095,
      "grad_norm": 0.0035668055061250925,
      "learning_rate": 0.06399999542952453,
      "loss": 0.2262,
      "num_input_tokens_seen": 24042240,
      "step": 27775
    },
    {
      "epoch": 13.097595473833097,
      "grad_norm": 0.0032752917613834143,
      "learning_rate": 0.0639517400010563,
      "loss": 0.1841,
      "num_input_tokens_seen": 24046496,
      "step": 27780
    },
    {
      "epoch": 13.0999528524281,
      "grad_norm": 0.004176370333880186,
      "learning_rate": 0.06390349784231118,
      "loss": 0.1936,
      "num_input_tokens_seen": 24050800,
      "step": 27785
    },
    {
      "epoch": 13.102310231023102,
      "grad_norm": 0.0024933184031397104,
      "learning_rate": 0.06385526896072859,
      "loss": 0.1646,
      "num_input_tokens_seen": 24055104,
      "step": 27790
    },
    {
      "epoch": 13.104667609618105,
      "grad_norm": 0.0037771931383758783,
      "learning_rate": 0.06380705336374613,
      "loss": 0.2437,
      "num_input_tokens_seen": 24059456,
      "step": 27795
    },
    {
      "epoch": 13.107024988213107,
      "grad_norm": 0.004191447515040636,
      "learning_rate": 0.06375885105879918,
      "loss": 0.225,
      "num_input_tokens_seen": 24063632,
      "step": 27800
    },
    {
      "epoch": 13.107024988213107,
      "eval_loss": 0.21396879851818085,
      "eval_runtime": 21.9007,
      "eval_samples_per_second": 43.058,
      "eval_steps_per_second": 21.552,
      "num_input_tokens_seen": 24063632,
      "step": 27800
    },
    {
      "epoch": 13.10938236680811,
      "grad_norm": 0.0027595381252467632,
      "learning_rate": 0.06371066205332115,
      "loss": 0.1885,
      "num_input_tokens_seen": 24068320,
      "step": 27805
    },
    {
      "epoch": 13.111739745403112,
      "grad_norm": 0.004130181856453419,
      "learning_rate": 0.06366248635474347,
      "loss": 0.186,
      "num_input_tokens_seen": 24073200,
      "step": 27810
    },
    {
      "epoch": 13.114097123998114,
      "grad_norm": 0.0022752874065190554,
      "learning_rate": 0.06361432397049532,
      "loss": 0.1609,
      "num_input_tokens_seen": 24077280,
      "step": 27815
    },
    {
      "epoch": 13.116454502593117,
      "grad_norm": 0.00259313709102571,
      "learning_rate": 0.06356617490800408,
      "loss": 0.1879,
      "num_input_tokens_seen": 24080848,
      "step": 27820
    },
    {
      "epoch": 13.118811881188119,
      "grad_norm": 0.0036827493458986282,
      "learning_rate": 0.06351803917469478,
      "loss": 0.2411,
      "num_input_tokens_seen": 24085520,
      "step": 27825
    },
    {
      "epoch": 13.121169259783121,
      "grad_norm": 0.004294184502214193,
      "learning_rate": 0.06346991677799067,
      "loss": 0.2086,
      "num_input_tokens_seen": 24090064,
      "step": 27830
    },
    {
      "epoch": 13.123526638378124,
      "grad_norm": 0.004798646084964275,
      "learning_rate": 0.06342180772531283,
      "loss": 0.2032,
      "num_input_tokens_seen": 24093952,
      "step": 27835
    },
    {
      "epoch": 13.125884016973126,
      "grad_norm": 0.002713978523388505,
      "learning_rate": 0.06337371202408021,
      "loss": 0.1442,
      "num_input_tokens_seen": 24098176,
      "step": 27840
    },
    {
      "epoch": 13.128241395568129,
      "grad_norm": 0.0034760653506964445,
      "learning_rate": 0.06332562968170984,
      "loss": 0.2039,
      "num_input_tokens_seen": 24102176,
      "step": 27845
    },
    {
      "epoch": 13.130598774163131,
      "grad_norm": 0.004568598233163357,
      "learning_rate": 0.06327756070561656,
      "loss": 0.2418,
      "num_input_tokens_seen": 24106960,
      "step": 27850
    },
    {
      "epoch": 13.132956152758133,
      "grad_norm": 0.004888992756605148,
      "learning_rate": 0.06322950510321329,
      "loss": 0.182,
      "num_input_tokens_seen": 24110720,
      "step": 27855
    },
    {
      "epoch": 13.135313531353136,
      "grad_norm": 0.0038136441726237535,
      "learning_rate": 0.06318146288191076,
      "loss": 0.2349,
      "num_input_tokens_seen": 24115232,
      "step": 27860
    },
    {
      "epoch": 13.137670909948138,
      "grad_norm": 0.005578142590820789,
      "learning_rate": 0.06313343404911763,
      "loss": 0.2506,
      "num_input_tokens_seen": 24119712,
      "step": 27865
    },
    {
      "epoch": 13.14002828854314,
      "grad_norm": 0.003042817348614335,
      "learning_rate": 0.0630854186122406,
      "loss": 0.1904,
      "num_input_tokens_seen": 24123504,
      "step": 27870
    },
    {
      "epoch": 13.142385667138143,
      "grad_norm": 0.003032529028132558,
      "learning_rate": 0.06303741657868431,
      "loss": 0.1865,
      "num_input_tokens_seen": 24127856,
      "step": 27875
    },
    {
      "epoch": 13.144743045733145,
      "grad_norm": 0.003658212022855878,
      "learning_rate": 0.06298942795585115,
      "loss": 0.1937,
      "num_input_tokens_seen": 24132704,
      "step": 27880
    },
    {
      "epoch": 13.147100424328148,
      "grad_norm": 0.003972828853875399,
      "learning_rate": 0.06294145275114167,
      "loss": 0.2712,
      "num_input_tokens_seen": 24137584,
      "step": 27885
    },
    {
      "epoch": 13.14945780292315,
      "grad_norm": 0.002773654880002141,
      "learning_rate": 0.06289349097195428,
      "loss": 0.2402,
      "num_input_tokens_seen": 24141840,
      "step": 27890
    },
    {
      "epoch": 13.151815181518153,
      "grad_norm": 0.003078120294958353,
      "learning_rate": 0.06284554262568516,
      "loss": 0.211,
      "num_input_tokens_seen": 24146512,
      "step": 27895
    },
    {
      "epoch": 13.154172560113155,
      "grad_norm": 0.0034743689466267824,
      "learning_rate": 0.06279760771972868,
      "loss": 0.1935,
      "num_input_tokens_seen": 24150976,
      "step": 27900
    },
    {
      "epoch": 13.156529938708157,
      "grad_norm": 0.0023883068934082985,
      "learning_rate": 0.06274968626147688,
      "loss": 0.1976,
      "num_input_tokens_seen": 24156128,
      "step": 27905
    },
    {
      "epoch": 13.15888731730316,
      "grad_norm": 0.004443046171218157,
      "learning_rate": 0.06270177825831993,
      "loss": 0.235,
      "num_input_tokens_seen": 24160096,
      "step": 27910
    },
    {
      "epoch": 13.16124469589816,
      "grad_norm": 0.0033913280349224806,
      "learning_rate": 0.06265388371764587,
      "loss": 0.2305,
      "num_input_tokens_seen": 24164624,
      "step": 27915
    },
    {
      "epoch": 13.163602074493163,
      "grad_norm": 0.0029678100254386663,
      "learning_rate": 0.0626060026468406,
      "loss": 0.1693,
      "num_input_tokens_seen": 24168416,
      "step": 27920
    },
    {
      "epoch": 13.165959453088165,
      "grad_norm": 0.003149575786665082,
      "learning_rate": 0.06255813505328794,
      "loss": 0.1726,
      "num_input_tokens_seen": 24173216,
      "step": 27925
    },
    {
      "epoch": 13.168316831683168,
      "grad_norm": 0.0029644242022186518,
      "learning_rate": 0.06251028094436978,
      "loss": 0.2217,
      "num_input_tokens_seen": 24177040,
      "step": 27930
    },
    {
      "epoch": 13.17067421027817,
      "grad_norm": 0.003028053091838956,
      "learning_rate": 0.06246244032746568,
      "loss": 0.1933,
      "num_input_tokens_seen": 24180880,
      "step": 27935
    },
    {
      "epoch": 13.173031588873172,
      "grad_norm": 0.004019772633910179,
      "learning_rate": 0.06241461320995342,
      "loss": 0.1773,
      "num_input_tokens_seen": 24185136,
      "step": 27940
    },
    {
      "epoch": 13.175388967468175,
      "grad_norm": 0.002202889183536172,
      "learning_rate": 0.062366799599208426,
      "loss": 0.1963,
      "num_input_tokens_seen": 24188800,
      "step": 27945
    },
    {
      "epoch": 13.177746346063177,
      "grad_norm": 0.0024415520019829273,
      "learning_rate": 0.06231899950260418,
      "loss": 0.2057,
      "num_input_tokens_seen": 24192832,
      "step": 27950
    },
    {
      "epoch": 13.18010372465818,
      "grad_norm": 0.00513939606025815,
      "learning_rate": 0.06227121292751214,
      "loss": 0.198,
      "num_input_tokens_seen": 24197440,
      "step": 27955
    },
    {
      "epoch": 13.182461103253182,
      "grad_norm": 0.004126903135329485,
      "learning_rate": 0.062223439881301496,
      "loss": 0.1869,
      "num_input_tokens_seen": 24202592,
      "step": 27960
    },
    {
      "epoch": 13.184818481848184,
      "grad_norm": 0.0019482758361846209,
      "learning_rate": 0.06217568037133948,
      "loss": 0.1939,
      "num_input_tokens_seen": 24207568,
      "step": 27965
    },
    {
      "epoch": 13.187175860443187,
      "grad_norm": 0.0032388712279498577,
      "learning_rate": 0.06212793440499126,
      "loss": 0.2126,
      "num_input_tokens_seen": 24211664,
      "step": 27970
    },
    {
      "epoch": 13.18953323903819,
      "grad_norm": 0.006384497974067926,
      "learning_rate": 0.062080201989619783,
      "loss": 0.1952,
      "num_input_tokens_seen": 24215904,
      "step": 27975
    },
    {
      "epoch": 13.191890617633192,
      "grad_norm": 0.006659139413386583,
      "learning_rate": 0.062032483132586094,
      "loss": 0.188,
      "num_input_tokens_seen": 24220016,
      "step": 27980
    },
    {
      "epoch": 13.194247996228194,
      "grad_norm": 0.004265317227691412,
      "learning_rate": 0.0619847778412489,
      "loss": 0.2083,
      "num_input_tokens_seen": 24224016,
      "step": 27985
    },
    {
      "epoch": 13.196605374823196,
      "grad_norm": 0.004381841514259577,
      "learning_rate": 0.06193708612296509,
      "loss": 0.1962,
      "num_input_tokens_seen": 24228736,
      "step": 27990
    },
    {
      "epoch": 13.198962753418199,
      "grad_norm": 0.0052251312881708145,
      "learning_rate": 0.06188940798508923,
      "loss": 0.1922,
      "num_input_tokens_seen": 24232400,
      "step": 27995
    },
    {
      "epoch": 13.201320132013201,
      "grad_norm": 0.0033367096912115812,
      "learning_rate": 0.06184174343497397,
      "loss": 0.2023,
      "num_input_tokens_seen": 24237248,
      "step": 28000
    },
    {
      "epoch": 13.201320132013201,
      "eval_loss": 0.2140951305627823,
      "eval_runtime": 21.9522,
      "eval_samples_per_second": 42.957,
      "eval_steps_per_second": 21.501,
      "num_input_tokens_seen": 24237248,
      "step": 28000
    },
    {
      "epoch": 13.203677510608204,
      "grad_norm": 0.005400930531322956,
      "learning_rate": 0.061794092479969726,
      "loss": 0.1785,
      "num_input_tokens_seen": 24241456,
      "step": 28005
    },
    {
      "epoch": 13.206034889203206,
      "grad_norm": 0.003697616746649146,
      "learning_rate": 0.06174645512742485,
      "loss": 0.2231,
      "num_input_tokens_seen": 24245712,
      "step": 28010
    },
    {
      "epoch": 13.208392267798208,
      "grad_norm": 0.005141960922628641,
      "learning_rate": 0.06169883138468565,
      "loss": 0.2152,
      "num_input_tokens_seen": 24249904,
      "step": 28015
    },
    {
      "epoch": 13.21074964639321,
      "grad_norm": 0.004695049021393061,
      "learning_rate": 0.06165122125909637,
      "loss": 0.1604,
      "num_input_tokens_seen": 24253808,
      "step": 28020
    },
    {
      "epoch": 13.213107024988213,
      "grad_norm": 0.004301515407860279,
      "learning_rate": 0.061603624757998965,
      "loss": 0.2452,
      "num_input_tokens_seen": 24257472,
      "step": 28025
    },
    {
      "epoch": 13.215464403583216,
      "grad_norm": 0.0022376300767064095,
      "learning_rate": 0.0615560418887335,
      "loss": 0.1725,
      "num_input_tokens_seen": 24261040,
      "step": 28030
    },
    {
      "epoch": 13.217821782178218,
      "grad_norm": 0.005454407539218664,
      "learning_rate": 0.06150847265863787,
      "loss": 0.2209,
      "num_input_tokens_seen": 24264976,
      "step": 28035
    },
    {
      "epoch": 13.22017916077322,
      "grad_norm": 0.0018482495797798038,
      "learning_rate": 0.061460917075047757,
      "loss": 0.1314,
      "num_input_tokens_seen": 24269920,
      "step": 28040
    },
    {
      "epoch": 13.222536539368223,
      "grad_norm": 0.004151421133428812,
      "learning_rate": 0.06141337514529694,
      "loss": 0.2004,
      "num_input_tokens_seen": 24276144,
      "step": 28045
    },
    {
      "epoch": 13.224893917963225,
      "grad_norm": 0.004850315395742655,
      "learning_rate": 0.06136584687671687,
      "loss": 0.1713,
      "num_input_tokens_seen": 24280688,
      "step": 28050
    },
    {
      "epoch": 13.227251296558228,
      "grad_norm": 0.004380761180073023,
      "learning_rate": 0.061318332276637064,
      "loss": 0.2078,
      "num_input_tokens_seen": 24285424,
      "step": 28055
    },
    {
      "epoch": 13.22960867515323,
      "grad_norm": 0.0027994848787784576,
      "learning_rate": 0.06127083135238491,
      "loss": 0.1871,
      "num_input_tokens_seen": 24289136,
      "step": 28060
    },
    {
      "epoch": 13.231966053748232,
      "grad_norm": 0.005169219803065062,
      "learning_rate": 0.06122334411128555,
      "loss": 0.2711,
      "num_input_tokens_seen": 24293072,
      "step": 28065
    },
    {
      "epoch": 13.234323432343235,
      "grad_norm": 0.00351754785515368,
      "learning_rate": 0.06117587056066223,
      "loss": 0.2171,
      "num_input_tokens_seen": 24296752,
      "step": 28070
    },
    {
      "epoch": 13.236680810938237,
      "grad_norm": 0.004001711029559374,
      "learning_rate": 0.06112841070783589,
      "loss": 0.2184,
      "num_input_tokens_seen": 24301920,
      "step": 28075
    },
    {
      "epoch": 13.23903818953324,
      "grad_norm": 0.005521354731172323,
      "learning_rate": 0.061080964560125406,
      "loss": 0.2301,
      "num_input_tokens_seen": 24305952,
      "step": 28080
    },
    {
      "epoch": 13.241395568128242,
      "grad_norm": 0.002503706607967615,
      "learning_rate": 0.06103353212484766,
      "loss": 0.2148,
      "num_input_tokens_seen": 24310016,
      "step": 28085
    },
    {
      "epoch": 13.243752946723244,
      "grad_norm": 0.0036319224163889885,
      "learning_rate": 0.06098611340931722,
      "loss": 0.1896,
      "num_input_tokens_seen": 24314544,
      "step": 28090
    },
    {
      "epoch": 13.246110325318247,
      "grad_norm": 0.0036453919019550085,
      "learning_rate": 0.06093870842084672,
      "loss": 0.1524,
      "num_input_tokens_seen": 24319616,
      "step": 28095
    },
    {
      "epoch": 13.24846770391325,
      "grad_norm": 0.0028188067954033613,
      "learning_rate": 0.06089131716674666,
      "loss": 0.1547,
      "num_input_tokens_seen": 24323568,
      "step": 28100
    },
    {
      "epoch": 13.250825082508252,
      "grad_norm": 0.002170430961996317,
      "learning_rate": 0.060843939654325226,
      "loss": 0.1521,
      "num_input_tokens_seen": 24328016,
      "step": 28105
    },
    {
      "epoch": 13.253182461103254,
      "grad_norm": 0.0026750110555440187,
      "learning_rate": 0.06079657589088873,
      "loss": 0.1555,
      "num_input_tokens_seen": 24331904,
      "step": 28110
    },
    {
      "epoch": 13.255539839698255,
      "grad_norm": 0.0031522426288574934,
      "learning_rate": 0.06074922588374126,
      "loss": 0.2194,
      "num_input_tokens_seen": 24336368,
      "step": 28115
    },
    {
      "epoch": 13.257897218293257,
      "grad_norm": 0.0028848464135080576,
      "learning_rate": 0.06070188964018472,
      "loss": 0.2181,
      "num_input_tokens_seen": 24341824,
      "step": 28120
    },
    {
      "epoch": 13.26025459688826,
      "grad_norm": 0.0029969143215566874,
      "learning_rate": 0.06065456716751902,
      "loss": 0.2,
      "num_input_tokens_seen": 24346336,
      "step": 28125
    },
    {
      "epoch": 13.262611975483262,
      "grad_norm": 0.00850124005228281,
      "learning_rate": 0.06060725847304182,
      "loss": 0.1883,
      "num_input_tokens_seen": 24350960,
      "step": 28130
    },
    {
      "epoch": 13.264969354078264,
      "grad_norm": 0.0036414163187146187,
      "learning_rate": 0.06055996356404877,
      "loss": 0.2027,
      "num_input_tokens_seen": 24355344,
      "step": 28135
    },
    {
      "epoch": 13.267326732673267,
      "grad_norm": 0.0027886638417840004,
      "learning_rate": 0.06051268244783327,
      "loss": 0.206,
      "num_input_tokens_seen": 24359360,
      "step": 28140
    },
    {
      "epoch": 13.269684111268269,
      "grad_norm": 0.0029075411148369312,
      "learning_rate": 0.06046541513168676,
      "loss": 0.1943,
      "num_input_tokens_seen": 24363360,
      "step": 28145
    },
    {
      "epoch": 13.272041489863271,
      "grad_norm": 0.005886119324713945,
      "learning_rate": 0.060418161622898356,
      "loss": 0.2288,
      "num_input_tokens_seen": 24367792,
      "step": 28150
    },
    {
      "epoch": 13.274398868458274,
      "grad_norm": 0.006161311641335487,
      "learning_rate": 0.06037092192875521,
      "loss": 0.2044,
      "num_input_tokens_seen": 24372528,
      "step": 28155
    },
    {
      "epoch": 13.276756247053276,
      "grad_norm": 0.006770299281924963,
      "learning_rate": 0.060323696056542225,
      "loss": 0.2341,
      "num_input_tokens_seen": 24376768,
      "step": 28160
    },
    {
      "epoch": 13.279113625648279,
      "grad_norm": 0.0034961046185344458,
      "learning_rate": 0.06027648401354229,
      "loss": 0.1839,
      "num_input_tokens_seen": 24380896,
      "step": 28165
    },
    {
      "epoch": 13.281471004243281,
      "grad_norm": 0.0040716324001550674,
      "learning_rate": 0.06022928580703601,
      "loss": 0.1922,
      "num_input_tokens_seen": 24384912,
      "step": 28170
    },
    {
      "epoch": 13.283828382838283,
      "grad_norm": 0.002871529432013631,
      "learning_rate": 0.060182101444301986,
      "loss": 0.2,
      "num_input_tokens_seen": 24389264,
      "step": 28175
    },
    {
      "epoch": 13.286185761433286,
      "grad_norm": 0.004829874727874994,
      "learning_rate": 0.06013493093261669,
      "loss": 0.1876,
      "num_input_tokens_seen": 24393584,
      "step": 28180
    },
    {
      "epoch": 13.288543140028288,
      "grad_norm": 0.002820091089233756,
      "learning_rate": 0.06008777427925432,
      "loss": 0.1997,
      "num_input_tokens_seen": 24398896,
      "step": 28185
    },
    {
      "epoch": 13.29090051862329,
      "grad_norm": 0.0027086748741567135,
      "learning_rate": 0.06004063149148705,
      "loss": 0.1499,
      "num_input_tokens_seen": 24403440,
      "step": 28190
    },
    {
      "epoch": 13.293257897218293,
      "grad_norm": 0.006282461807131767,
      "learning_rate": 0.05999350257658497,
      "loss": 0.2398,
      "num_input_tokens_seen": 24408000,
      "step": 28195
    },
    {
      "epoch": 13.295615275813295,
      "grad_norm": 0.0027803899720311165,
      "learning_rate": 0.05994638754181582,
      "loss": 0.2117,
      "num_input_tokens_seen": 24411712,
      "step": 28200
    },
    {
      "epoch": 13.295615275813295,
      "eval_loss": 0.21384276449680328,
      "eval_runtime": 21.9345,
      "eval_samples_per_second": 42.992,
      "eval_steps_per_second": 21.519,
      "num_input_tokens_seen": 24411712,
      "step": 28200
    },
    {
      "epoch": 13.297972654408298,
      "grad_norm": 0.005892493762075901,
      "learning_rate": 0.059899286394445445,
      "loss": 0.2265,
      "num_input_tokens_seen": 24415824,
      "step": 28205
    },
    {
      "epoch": 13.3003300330033,
      "grad_norm": 0.0029511824250221252,
      "learning_rate": 0.059852199141737346,
      "loss": 0.1965,
      "num_input_tokens_seen": 24420096,
      "step": 28210
    },
    {
      "epoch": 13.302687411598303,
      "grad_norm": 0.003923149313777685,
      "learning_rate": 0.05980512579095304,
      "loss": 0.2149,
      "num_input_tokens_seen": 24424336,
      "step": 28215
    },
    {
      "epoch": 13.305044790193305,
      "grad_norm": 0.00433417409658432,
      "learning_rate": 0.05975806634935181,
      "loss": 0.1748,
      "num_input_tokens_seen": 24427904,
      "step": 28220
    },
    {
      "epoch": 13.307402168788308,
      "grad_norm": 0.003342958865687251,
      "learning_rate": 0.05971102082419076,
      "loss": 0.2433,
      "num_input_tokens_seen": 24432480,
      "step": 28225
    },
    {
      "epoch": 13.30975954738331,
      "grad_norm": 0.0038674804382026196,
      "learning_rate": 0.05966398922272492,
      "loss": 0.1798,
      "num_input_tokens_seen": 24436688,
      "step": 28230
    },
    {
      "epoch": 13.312116925978312,
      "grad_norm": 0.003663122421130538,
      "learning_rate": 0.059616971552207236,
      "loss": 0.1997,
      "num_input_tokens_seen": 24440032,
      "step": 28235
    },
    {
      "epoch": 13.314474304573315,
      "grad_norm": 0.003171068150550127,
      "learning_rate": 0.059569967819888305,
      "loss": 0.1746,
      "num_input_tokens_seen": 24444480,
      "step": 28240
    },
    {
      "epoch": 13.316831683168317,
      "grad_norm": 0.0039473711512982845,
      "learning_rate": 0.05952297803301681,
      "loss": 0.1768,
      "num_input_tokens_seen": 24449104,
      "step": 28245
    },
    {
      "epoch": 13.31918906176332,
      "grad_norm": 0.004195316694676876,
      "learning_rate": 0.059476002198839056,
      "loss": 0.1942,
      "num_input_tokens_seen": 24452800,
      "step": 28250
    },
    {
      "epoch": 13.321546440358322,
      "grad_norm": 0.0033729146234691143,
      "learning_rate": 0.05942904032459935,
      "loss": 0.1543,
      "num_input_tokens_seen": 24457200,
      "step": 28255
    },
    {
      "epoch": 13.323903818953324,
      "grad_norm": 0.004285888280719519,
      "learning_rate": 0.05938209241753987,
      "loss": 0.2051,
      "num_input_tokens_seen": 24461776,
      "step": 28260
    },
    {
      "epoch": 13.326261197548327,
      "grad_norm": 0.006336894351989031,
      "learning_rate": 0.05933515848490046,
      "loss": 0.2457,
      "num_input_tokens_seen": 24465504,
      "step": 28265
    },
    {
      "epoch": 13.32861857614333,
      "grad_norm": 0.004565963987261057,
      "learning_rate": 0.059288238533918985,
      "loss": 0.1661,
      "num_input_tokens_seen": 24470240,
      "step": 28270
    },
    {
      "epoch": 13.330975954738332,
      "grad_norm": 0.00478568347170949,
      "learning_rate": 0.05924133257183113,
      "loss": 0.1917,
      "num_input_tokens_seen": 24474624,
      "step": 28275
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 0.002604943234473467,
      "learning_rate": 0.059194440605870285,
      "loss": 0.2095,
      "num_input_tokens_seen": 24479024,
      "step": 28280
    },
    {
      "epoch": 13.335690711928336,
      "grad_norm": 0.006089653354138136,
      "learning_rate": 0.059147562643267884,
      "loss": 0.2378,
      "num_input_tokens_seen": 24483584,
      "step": 28285
    },
    {
      "epoch": 13.338048090523339,
      "grad_norm": 0.0036345927510410547,
      "learning_rate": 0.059100698691253055,
      "loss": 0.1829,
      "num_input_tokens_seen": 24488208,
      "step": 28290
    },
    {
      "epoch": 13.340405469118341,
      "grad_norm": 0.0033221275079995394,
      "learning_rate": 0.05905384875705273,
      "loss": 0.2075,
      "num_input_tokens_seen": 24492240,
      "step": 28295
    },
    {
      "epoch": 13.342762847713344,
      "grad_norm": 0.0029119213577359915,
      "learning_rate": 0.05900701284789189,
      "loss": 0.192,
      "num_input_tokens_seen": 24496848,
      "step": 28300
    },
    {
      "epoch": 13.345120226308346,
      "grad_norm": 0.008414680138230324,
      "learning_rate": 0.058960190970993115,
      "loss": 0.1876,
      "num_input_tokens_seen": 24502480,
      "step": 28305
    },
    {
      "epoch": 13.347477604903348,
      "grad_norm": 0.0036090167704969645,
      "learning_rate": 0.058913383133576955,
      "loss": 0.2627,
      "num_input_tokens_seen": 24506736,
      "step": 28310
    },
    {
      "epoch": 13.34983498349835,
      "grad_norm": 0.0033711292780935764,
      "learning_rate": 0.05886658934286185,
      "loss": 0.2141,
      "num_input_tokens_seen": 24511040,
      "step": 28315
    },
    {
      "epoch": 13.352192362093351,
      "grad_norm": 0.0029844765085726976,
      "learning_rate": 0.058819809606063846,
      "loss": 0.1817,
      "num_input_tokens_seen": 24514608,
      "step": 28320
    },
    {
      "epoch": 13.354549740688354,
      "grad_norm": 0.0028412824030965567,
      "learning_rate": 0.05877304393039711,
      "loss": 0.1681,
      "num_input_tokens_seen": 24518368,
      "step": 28325
    },
    {
      "epoch": 13.356907119283356,
      "grad_norm": 0.00374275678768754,
      "learning_rate": 0.05872629232307338,
      "loss": 0.179,
      "num_input_tokens_seen": 24522752,
      "step": 28330
    },
    {
      "epoch": 13.359264497878359,
      "grad_norm": 0.003863735357299447,
      "learning_rate": 0.05867955479130239,
      "loss": 0.1837,
      "num_input_tokens_seen": 24526688,
      "step": 28335
    },
    {
      "epoch": 13.361621876473361,
      "grad_norm": 0.0031909197568893433,
      "learning_rate": 0.058632831342291705,
      "loss": 0.1433,
      "num_input_tokens_seen": 24531984,
      "step": 28340
    },
    {
      "epoch": 13.363979255068363,
      "grad_norm": 0.0034158031921833754,
      "learning_rate": 0.05858612198324655,
      "loss": 0.2103,
      "num_input_tokens_seen": 24536320,
      "step": 28345
    },
    {
      "epoch": 13.366336633663366,
      "grad_norm": 0.008751826360821724,
      "learning_rate": 0.05853942672137025,
      "loss": 0.2085,
      "num_input_tokens_seen": 24541600,
      "step": 28350
    },
    {
      "epoch": 13.368694012258368,
      "grad_norm": 0.004063539672642946,
      "learning_rate": 0.05849274556386363,
      "loss": 0.2129,
      "num_input_tokens_seen": 24545600,
      "step": 28355
    },
    {
      "epoch": 13.37105139085337,
      "grad_norm": 0.005061716306954622,
      "learning_rate": 0.05844607851792567,
      "loss": 0.1496,
      "num_input_tokens_seen": 24549952,
      "step": 28360
    },
    {
      "epoch": 13.373408769448373,
      "grad_norm": 0.003114063758403063,
      "learning_rate": 0.058399425590752924,
      "loss": 0.1658,
      "num_input_tokens_seen": 24554560,
      "step": 28365
    },
    {
      "epoch": 13.375766148043375,
      "grad_norm": 0.003976886626332998,
      "learning_rate": 0.05835278678953985,
      "loss": 0.1747,
      "num_input_tokens_seen": 24559040,
      "step": 28370
    },
    {
      "epoch": 13.378123526638378,
      "grad_norm": 0.004157899878919125,
      "learning_rate": 0.05830616212147874,
      "loss": 0.1832,
      "num_input_tokens_seen": 24563408,
      "step": 28375
    },
    {
      "epoch": 13.38048090523338,
      "grad_norm": 0.004519881214946508,
      "learning_rate": 0.058259551593759784,
      "loss": 0.2118,
      "num_input_tokens_seen": 24567072,
      "step": 28380
    },
    {
      "epoch": 13.382838283828383,
      "grad_norm": 0.003807040164247155,
      "learning_rate": 0.058212955213570804,
      "loss": 0.1562,
      "num_input_tokens_seen": 24571824,
      "step": 28385
    },
    {
      "epoch": 13.385195662423385,
      "grad_norm": 0.004244002979248762,
      "learning_rate": 0.0581663729880976,
      "loss": 0.233,
      "num_input_tokens_seen": 24576272,
      "step": 28390
    },
    {
      "epoch": 13.387553041018387,
      "grad_norm": 0.002989956410601735,
      "learning_rate": 0.05811980492452379,
      "loss": 0.2049,
      "num_input_tokens_seen": 24580304,
      "step": 28395
    },
    {
      "epoch": 13.38991041961339,
      "grad_norm": 0.00439246604219079,
      "learning_rate": 0.058073251030030644,
      "loss": 0.1836,
      "num_input_tokens_seen": 24584800,
      "step": 28400
    },
    {
      "epoch": 13.38991041961339,
      "eval_loss": 0.21217310428619385,
      "eval_runtime": 21.927,
      "eval_samples_per_second": 43.006,
      "eval_steps_per_second": 21.526,
      "num_input_tokens_seen": 24584800,
      "step": 28400
    },
    {
      "epoch": 13.392267798208392,
      "grad_norm": 0.00468235369771719,
      "learning_rate": 0.05802671131179747,
      "loss": 0.1912,
      "num_input_tokens_seen": 24589376,
      "step": 28405
    },
    {
      "epoch": 13.394625176803395,
      "grad_norm": 0.0034818497952073812,
      "learning_rate": 0.057980185777001154,
      "loss": 0.2414,
      "num_input_tokens_seen": 24593840,
      "step": 28410
    },
    {
      "epoch": 13.396982555398397,
      "grad_norm": 0.0028383713215589523,
      "learning_rate": 0.057933674432816606,
      "loss": 0.2027,
      "num_input_tokens_seen": 24598752,
      "step": 28415
    },
    {
      "epoch": 13.3993399339934,
      "grad_norm": 0.005863791797310114,
      "learning_rate": 0.05788717728641648,
      "loss": 0.2596,
      "num_input_tokens_seen": 24602368,
      "step": 28420
    },
    {
      "epoch": 13.401697312588402,
      "grad_norm": 0.005047095473855734,
      "learning_rate": 0.057840694344971126,
      "loss": 0.1757,
      "num_input_tokens_seen": 24606480,
      "step": 28425
    },
    {
      "epoch": 13.404054691183404,
      "grad_norm": 0.0047446079552173615,
      "learning_rate": 0.0577942256156489,
      "loss": 0.1959,
      "num_input_tokens_seen": 24610544,
      "step": 28430
    },
    {
      "epoch": 13.406412069778407,
      "grad_norm": 0.003061014460399747,
      "learning_rate": 0.057747771105615804,
      "loss": 0.1846,
      "num_input_tokens_seen": 24615680,
      "step": 28435
    },
    {
      "epoch": 13.408769448373409,
      "grad_norm": 0.004367829766124487,
      "learning_rate": 0.05770133082203568,
      "loss": 0.2115,
      "num_input_tokens_seen": 24620128,
      "step": 28440
    },
    {
      "epoch": 13.411126826968411,
      "grad_norm": 0.004928905516862869,
      "learning_rate": 0.0576549047720703,
      "loss": 0.1973,
      "num_input_tokens_seen": 24625072,
      "step": 28445
    },
    {
      "epoch": 13.413484205563414,
      "grad_norm": 0.0037302598357200623,
      "learning_rate": 0.05760849296287902,
      "loss": 0.1898,
      "num_input_tokens_seen": 24628704,
      "step": 28450
    },
    {
      "epoch": 13.415841584158416,
      "grad_norm": 0.004397667478770018,
      "learning_rate": 0.05756209540161919,
      "loss": 0.2185,
      "num_input_tokens_seen": 24633424,
      "step": 28455
    },
    {
      "epoch": 13.418198962753419,
      "grad_norm": 0.004364490509033203,
      "learning_rate": 0.05751571209544595,
      "loss": 0.1882,
      "num_input_tokens_seen": 24637136,
      "step": 28460
    },
    {
      "epoch": 13.420556341348421,
      "grad_norm": 0.0018602063646540046,
      "learning_rate": 0.057469343051512085,
      "loss": 0.188,
      "num_input_tokens_seen": 24641072,
      "step": 28465
    },
    {
      "epoch": 13.422913719943423,
      "grad_norm": 0.0025981825310736895,
      "learning_rate": 0.057422988276968324,
      "loss": 0.1357,
      "num_input_tokens_seen": 24644912,
      "step": 28470
    },
    {
      "epoch": 13.425271098538426,
      "grad_norm": 0.00603348296135664,
      "learning_rate": 0.05737664777896323,
      "loss": 0.1793,
      "num_input_tokens_seen": 24649344,
      "step": 28475
    },
    {
      "epoch": 13.427628477133428,
      "grad_norm": 0.0021637571044266224,
      "learning_rate": 0.057330321564642975,
      "loss": 0.2287,
      "num_input_tokens_seen": 24653808,
      "step": 28480
    },
    {
      "epoch": 13.42998585572843,
      "grad_norm": 0.0034166877157986164,
      "learning_rate": 0.05728400964115174,
      "loss": 0.168,
      "num_input_tokens_seen": 24658528,
      "step": 28485
    },
    {
      "epoch": 13.432343234323433,
      "grad_norm": 0.003960008267313242,
      "learning_rate": 0.057237712015631305,
      "loss": 0.1896,
      "num_input_tokens_seen": 24662848,
      "step": 28490
    },
    {
      "epoch": 13.434700612918435,
      "grad_norm": 0.005399961955845356,
      "learning_rate": 0.057191428695221425,
      "loss": 0.1934,
      "num_input_tokens_seen": 24667056,
      "step": 28495
    },
    {
      "epoch": 13.437057991513438,
      "grad_norm": 0.00431847246363759,
      "learning_rate": 0.05714515968705958,
      "loss": 0.224,
      "num_input_tokens_seen": 24671360,
      "step": 28500
    },
    {
      "epoch": 13.43941537010844,
      "grad_norm": 0.003903711447492242,
      "learning_rate": 0.05709890499828099,
      "loss": 0.174,
      "num_input_tokens_seen": 24676480,
      "step": 28505
    },
    {
      "epoch": 13.441772748703443,
      "grad_norm": 0.0037968994583934546,
      "learning_rate": 0.05705266463601868,
      "loss": 0.1788,
      "num_input_tokens_seen": 24680864,
      "step": 28510
    },
    {
      "epoch": 13.444130127298443,
      "grad_norm": 0.0038604261353611946,
      "learning_rate": 0.057006438607403565,
      "loss": 0.1397,
      "num_input_tokens_seen": 24684944,
      "step": 28515
    },
    {
      "epoch": 13.446487505893446,
      "grad_norm": 0.004196444526314735,
      "learning_rate": 0.056960226919564205,
      "loss": 0.1942,
      "num_input_tokens_seen": 24688864,
      "step": 28520
    },
    {
      "epoch": 13.448844884488448,
      "grad_norm": 0.005792573094367981,
      "learning_rate": 0.05691402957962713,
      "loss": 0.2285,
      "num_input_tokens_seen": 24693632,
      "step": 28525
    },
    {
      "epoch": 13.45120226308345,
      "grad_norm": 0.002516391221433878,
      "learning_rate": 0.05686784659471642,
      "loss": 0.2288,
      "num_input_tokens_seen": 24697456,
      "step": 28530
    },
    {
      "epoch": 13.453559641678453,
      "grad_norm": 0.003014878137037158,
      "learning_rate": 0.056821677971954136,
      "loss": 0.1935,
      "num_input_tokens_seen": 24701920,
      "step": 28535
    },
    {
      "epoch": 13.455917020273455,
      "grad_norm": 0.002779040252789855,
      "learning_rate": 0.05677552371846012,
      "loss": 0.1753,
      "num_input_tokens_seen": 24706080,
      "step": 28540
    },
    {
      "epoch": 13.458274398868458,
      "grad_norm": 0.0035162148997187614,
      "learning_rate": 0.05672938384135182,
      "loss": 0.2106,
      "num_input_tokens_seen": 24710160,
      "step": 28545
    },
    {
      "epoch": 13.46063177746346,
      "grad_norm": 0.004334775265306234,
      "learning_rate": 0.05668325834774465,
      "loss": 0.1738,
      "num_input_tokens_seen": 24714256,
      "step": 28550
    },
    {
      "epoch": 13.462989156058462,
      "grad_norm": 0.0028555768076330423,
      "learning_rate": 0.05663714724475177,
      "loss": 0.1302,
      "num_input_tokens_seen": 24717872,
      "step": 28555
    },
    {
      "epoch": 13.465346534653465,
      "grad_norm": 0.004129506181925535,
      "learning_rate": 0.05659105053948403,
      "loss": 0.1898,
      "num_input_tokens_seen": 24723472,
      "step": 28560
    },
    {
      "epoch": 13.467703913248467,
      "grad_norm": 0.0075521161779761314,
      "learning_rate": 0.056544968239050176,
      "loss": 0.2837,
      "num_input_tokens_seen": 24729120,
      "step": 28565
    },
    {
      "epoch": 13.47006129184347,
      "grad_norm": 0.005326953250914812,
      "learning_rate": 0.056498900350556616,
      "loss": 0.2394,
      "num_input_tokens_seen": 24733552,
      "step": 28570
    },
    {
      "epoch": 13.472418670438472,
      "grad_norm": 0.002482346259057522,
      "learning_rate": 0.05645284688110766,
      "loss": 0.1607,
      "num_input_tokens_seen": 24738496,
      "step": 28575
    },
    {
      "epoch": 13.474776049033474,
      "grad_norm": 0.0029092859476804733,
      "learning_rate": 0.05640680783780532,
      "loss": 0.1684,
      "num_input_tokens_seen": 24742256,
      "step": 28580
    },
    {
      "epoch": 13.477133427628477,
      "grad_norm": 0.0027815192006528378,
      "learning_rate": 0.056360783227749324,
      "loss": 0.2098,
      "num_input_tokens_seen": 24746768,
      "step": 28585
    },
    {
      "epoch": 13.47949080622348,
      "grad_norm": 0.003038255264982581,
      "learning_rate": 0.05631477305803728,
      "loss": 0.1545,
      "num_input_tokens_seen": 24750960,
      "step": 28590
    },
    {
      "epoch": 13.481848184818482,
      "grad_norm": 0.006580726243555546,
      "learning_rate": 0.05626877733576462,
      "loss": 0.1363,
      "num_input_tokens_seen": 24755120,
      "step": 28595
    },
    {
      "epoch": 13.484205563413484,
      "grad_norm": 0.0019359199795871973,
      "learning_rate": 0.05622279606802435,
      "loss": 0.1836,
      "num_input_tokens_seen": 24759888,
      "step": 28600
    },
    {
      "epoch": 13.484205563413484,
      "eval_loss": 0.22224083542823792,
      "eval_runtime": 21.9158,
      "eval_samples_per_second": 43.028,
      "eval_steps_per_second": 21.537,
      "num_input_tokens_seen": 24759888,
      "step": 28600
    },
    {
      "epoch": 13.486562942008486,
      "grad_norm": 0.003124724142253399,
      "learning_rate": 0.05617682926190744,
      "loss": 0.1894,
      "num_input_tokens_seen": 24764560,
      "step": 28605
    },
    {
      "epoch": 13.488920320603489,
      "grad_norm": 0.003626782912760973,
      "learning_rate": 0.05613087692450248,
      "loss": 0.214,
      "num_input_tokens_seen": 24769184,
      "step": 28610
    },
    {
      "epoch": 13.491277699198491,
      "grad_norm": 0.003989465069025755,
      "learning_rate": 0.05608493906289592,
      "loss": 0.2224,
      "num_input_tokens_seen": 24773088,
      "step": 28615
    },
    {
      "epoch": 13.493635077793494,
      "grad_norm": 0.003706319723278284,
      "learning_rate": 0.05603901568417201,
      "loss": 0.1828,
      "num_input_tokens_seen": 24777520,
      "step": 28620
    },
    {
      "epoch": 13.495992456388496,
      "grad_norm": 0.004123826045542955,
      "learning_rate": 0.055993106795412625,
      "loss": 0.2338,
      "num_input_tokens_seen": 24781360,
      "step": 28625
    },
    {
      "epoch": 13.498349834983498,
      "grad_norm": 0.0038644177839159966,
      "learning_rate": 0.05594721240369759,
      "loss": 0.1592,
      "num_input_tokens_seen": 24785648,
      "step": 28630
    },
    {
      "epoch": 13.500707213578501,
      "grad_norm": 0.003213512944057584,
      "learning_rate": 0.055901332516104296,
      "loss": 0.1908,
      "num_input_tokens_seen": 24790304,
      "step": 28635
    },
    {
      "epoch": 13.503064592173503,
      "grad_norm": 0.005704285576939583,
      "learning_rate": 0.05585546713970804,
      "loss": 0.2105,
      "num_input_tokens_seen": 24794048,
      "step": 28640
    },
    {
      "epoch": 13.505421970768506,
      "grad_norm": 0.003139129839837551,
      "learning_rate": 0.05580961628158189,
      "loss": 0.225,
      "num_input_tokens_seen": 24799040,
      "step": 28645
    },
    {
      "epoch": 13.507779349363508,
      "grad_norm": 0.003374840598553419,
      "learning_rate": 0.05576377994879659,
      "loss": 0.2037,
      "num_input_tokens_seen": 24802992,
      "step": 28650
    },
    {
      "epoch": 13.51013672795851,
      "grad_norm": 0.0038126667495816946,
      "learning_rate": 0.05571795814842063,
      "loss": 0.184,
      "num_input_tokens_seen": 24807248,
      "step": 28655
    },
    {
      "epoch": 13.512494106553513,
      "grad_norm": 0.002653178060427308,
      "learning_rate": 0.05567215088752037,
      "loss": 0.2166,
      "num_input_tokens_seen": 24811696,
      "step": 28660
    },
    {
      "epoch": 13.514851485148515,
      "grad_norm": 0.0051523963920772076,
      "learning_rate": 0.05562635817315981,
      "loss": 0.2072,
      "num_input_tokens_seen": 24815376,
      "step": 28665
    },
    {
      "epoch": 13.517208863743518,
      "grad_norm": 0.002797238063067198,
      "learning_rate": 0.05558058001240083,
      "loss": 0.1601,
      "num_input_tokens_seen": 24819712,
      "step": 28670
    },
    {
      "epoch": 13.51956624233852,
      "grad_norm": 0.005402186419814825,
      "learning_rate": 0.055534816412302915,
      "loss": 0.133,
      "num_input_tokens_seen": 24824768,
      "step": 28675
    },
    {
      "epoch": 13.521923620933523,
      "grad_norm": 0.004436859395354986,
      "learning_rate": 0.055489067379923436,
      "loss": 0.2048,
      "num_input_tokens_seen": 24829104,
      "step": 28680
    },
    {
      "epoch": 13.524280999528525,
      "grad_norm": 0.004116388037800789,
      "learning_rate": 0.055443332922317505,
      "loss": 0.1965,
      "num_input_tokens_seen": 24832864,
      "step": 28685
    },
    {
      "epoch": 13.526638378123527,
      "grad_norm": 0.0035979272797703743,
      "learning_rate": 0.055397613046537876,
      "loss": 0.1729,
      "num_input_tokens_seen": 24836656,
      "step": 28690
    },
    {
      "epoch": 13.52899575671853,
      "grad_norm": 0.0059866104274988174,
      "learning_rate": 0.055351907759635145,
      "loss": 0.2924,
      "num_input_tokens_seen": 24841120,
      "step": 28695
    },
    {
      "epoch": 13.531353135313532,
      "grad_norm": 0.004189211875200272,
      "learning_rate": 0.05530621706865772,
      "loss": 0.1932,
      "num_input_tokens_seen": 24845792,
      "step": 28700
    },
    {
      "epoch": 13.533710513908535,
      "grad_norm": 0.004581711255013943,
      "learning_rate": 0.055260540980651564,
      "loss": 0.183,
      "num_input_tokens_seen": 24850704,
      "step": 28705
    },
    {
      "epoch": 13.536067892503535,
      "grad_norm": 0.005896758288145065,
      "learning_rate": 0.05521487950266062,
      "loss": 0.1509,
      "num_input_tokens_seen": 24855200,
      "step": 28710
    },
    {
      "epoch": 13.53842527109854,
      "grad_norm": 0.003892758162692189,
      "learning_rate": 0.055169232641726344,
      "loss": 0.1931,
      "num_input_tokens_seen": 24859696,
      "step": 28715
    },
    {
      "epoch": 13.54078264969354,
      "grad_norm": 0.0030240598134696484,
      "learning_rate": 0.055123600404888166,
      "loss": 0.1884,
      "num_input_tokens_seen": 24864000,
      "step": 28720
    },
    {
      "epoch": 13.543140028288542,
      "grad_norm": 0.004869920667260885,
      "learning_rate": 0.05507798279918309,
      "loss": 0.1765,
      "num_input_tokens_seen": 24868720,
      "step": 28725
    },
    {
      "epoch": 13.545497406883545,
      "grad_norm": 0.004318545572459698,
      "learning_rate": 0.0550323798316459,
      "loss": 0.1821,
      "num_input_tokens_seen": 24873056,
      "step": 28730
    },
    {
      "epoch": 13.547854785478547,
      "grad_norm": 0.00289716525003314,
      "learning_rate": 0.05498679150930916,
      "loss": 0.2107,
      "num_input_tokens_seen": 24877200,
      "step": 28735
    },
    {
      "epoch": 13.55021216407355,
      "grad_norm": 0.00298098917119205,
      "learning_rate": 0.05494121783920323,
      "loss": 0.2495,
      "num_input_tokens_seen": 24881584,
      "step": 28740
    },
    {
      "epoch": 13.552569542668552,
      "grad_norm": 0.0060999151319265366,
      "learning_rate": 0.05489565882835605,
      "loss": 0.1711,
      "num_input_tokens_seen": 24886544,
      "step": 28745
    },
    {
      "epoch": 13.554926921263554,
      "grad_norm": 0.003732851706445217,
      "learning_rate": 0.05485011448379348,
      "loss": 0.1503,
      "num_input_tokens_seen": 24891664,
      "step": 28750
    },
    {
      "epoch": 13.557284299858557,
      "grad_norm": 0.007610046770423651,
      "learning_rate": 0.05480458481253893,
      "loss": 0.1851,
      "num_input_tokens_seen": 24895920,
      "step": 28755
    },
    {
      "epoch": 13.55964167845356,
      "grad_norm": 0.003554454306140542,
      "learning_rate": 0.054759069821613715,
      "loss": 0.1549,
      "num_input_tokens_seen": 24900304,
      "step": 28760
    },
    {
      "epoch": 13.561999057048562,
      "grad_norm": 0.002551827346906066,
      "learning_rate": 0.05471356951803683,
      "loss": 0.2156,
      "num_input_tokens_seen": 24904432,
      "step": 28765
    },
    {
      "epoch": 13.564356435643564,
      "grad_norm": 0.0034580405335873365,
      "learning_rate": 0.054668083908824945,
      "loss": 0.2575,
      "num_input_tokens_seen": 24909568,
      "step": 28770
    },
    {
      "epoch": 13.566713814238566,
      "grad_norm": 0.004853108432143927,
      "learning_rate": 0.054622613000992526,
      "loss": 0.25,
      "num_input_tokens_seen": 24913616,
      "step": 28775
    },
    {
      "epoch": 13.569071192833569,
      "grad_norm": 0.005195214878767729,
      "learning_rate": 0.05457715680155182,
      "loss": 0.1675,
      "num_input_tokens_seen": 24918048,
      "step": 28780
    },
    {
      "epoch": 13.571428571428571,
      "grad_norm": 0.005279366858303547,
      "learning_rate": 0.05453171531751265,
      "loss": 0.1853,
      "num_input_tokens_seen": 24922464,
      "step": 28785
    },
    {
      "epoch": 13.573785950023574,
      "grad_norm": 0.00409263651818037,
      "learning_rate": 0.05448628855588276,
      "loss": 0.2076,
      "num_input_tokens_seen": 24926800,
      "step": 28790
    },
    {
      "epoch": 13.576143328618576,
      "grad_norm": 0.0037708377931267023,
      "learning_rate": 0.05444087652366746,
      "loss": 0.2264,
      "num_input_tokens_seen": 24932192,
      "step": 28795
    },
    {
      "epoch": 13.578500707213578,
      "grad_norm": 0.0039263032376766205,
      "learning_rate": 0.05439547922786984,
      "loss": 0.2203,
      "num_input_tokens_seen": 24936720,
      "step": 28800
    },
    {
      "epoch": 13.578500707213578,
      "eval_loss": 0.22089065611362457,
      "eval_runtime": 21.9223,
      "eval_samples_per_second": 43.016,
      "eval_steps_per_second": 21.531,
      "num_input_tokens_seen": 24936720,
      "step": 28800
    },
    {
      "epoch": 13.58085808580858,
      "grad_norm": 0.006409460678696632,
      "learning_rate": 0.0543500966754908,
      "loss": 0.1963,
      "num_input_tokens_seen": 24941280,
      "step": 28805
    },
    {
      "epoch": 13.583215464403583,
      "grad_norm": 0.003566692117601633,
      "learning_rate": 0.05430472887352882,
      "loss": 0.1872,
      "num_input_tokens_seen": 24945088,
      "step": 28810
    },
    {
      "epoch": 13.585572842998586,
      "grad_norm": 0.0038699000142514706,
      "learning_rate": 0.05425937582898023,
      "loss": 0.2181,
      "num_input_tokens_seen": 24949456,
      "step": 28815
    },
    {
      "epoch": 13.587930221593588,
      "grad_norm": 0.0036918758414685726,
      "learning_rate": 0.054214037548839085,
      "loss": 0.1601,
      "num_input_tokens_seen": 24954336,
      "step": 28820
    },
    {
      "epoch": 13.59028760018859,
      "grad_norm": 0.004864132963120937,
      "learning_rate": 0.05416871404009703,
      "loss": 0.215,
      "num_input_tokens_seen": 24958752,
      "step": 28825
    },
    {
      "epoch": 13.592644978783593,
      "grad_norm": 0.0038606938906013966,
      "learning_rate": 0.054123405309743605,
      "loss": 0.1631,
      "num_input_tokens_seen": 24963296,
      "step": 28830
    },
    {
      "epoch": 13.595002357378595,
      "grad_norm": 0.004901327192783356,
      "learning_rate": 0.0540781113647659,
      "loss": 0.2024,
      "num_input_tokens_seen": 24967728,
      "step": 28835
    },
    {
      "epoch": 13.597359735973598,
      "grad_norm": 0.005510315764695406,
      "learning_rate": 0.054032832212148836,
      "loss": 0.2424,
      "num_input_tokens_seen": 24972544,
      "step": 28840
    },
    {
      "epoch": 13.5997171145686,
      "grad_norm": 0.004320807289332151,
      "learning_rate": 0.0539875678588751,
      "loss": 0.2456,
      "num_input_tokens_seen": 24977168,
      "step": 28845
    },
    {
      "epoch": 13.602074493163602,
      "grad_norm": 0.00375325046479702,
      "learning_rate": 0.05394231831192492,
      "loss": 0.1934,
      "num_input_tokens_seen": 24981632,
      "step": 28850
    },
    {
      "epoch": 13.604431871758605,
      "grad_norm": 0.007146845106035471,
      "learning_rate": 0.05389708357827639,
      "loss": 0.2199,
      "num_input_tokens_seen": 24985552,
      "step": 28855
    },
    {
      "epoch": 13.606789250353607,
      "grad_norm": 0.004711825400590897,
      "learning_rate": 0.05385186366490533,
      "loss": 0.1986,
      "num_input_tokens_seen": 24989712,
      "step": 28860
    },
    {
      "epoch": 13.60914662894861,
      "grad_norm": 0.005951453931629658,
      "learning_rate": 0.053806658578785166,
      "loss": 0.2152,
      "num_input_tokens_seen": 24994224,
      "step": 28865
    },
    {
      "epoch": 13.611504007543612,
      "grad_norm": 0.0059767975471913815,
      "learning_rate": 0.05376146832688705,
      "loss": 0.1974,
      "num_input_tokens_seen": 24998928,
      "step": 28870
    },
    {
      "epoch": 13.613861386138614,
      "grad_norm": 0.0045145186595618725,
      "learning_rate": 0.053716292916179964,
      "loss": 0.2441,
      "num_input_tokens_seen": 25003504,
      "step": 28875
    },
    {
      "epoch": 13.616218764733617,
      "grad_norm": 0.0036954889073967934,
      "learning_rate": 0.05367113235363045,
      "loss": 0.1986,
      "num_input_tokens_seen": 25008064,
      "step": 28880
    },
    {
      "epoch": 13.61857614332862,
      "grad_norm": 0.00467325933277607,
      "learning_rate": 0.05362598664620289,
      "loss": 0.164,
      "num_input_tokens_seen": 25011840,
      "step": 28885
    },
    {
      "epoch": 13.620933521923622,
      "grad_norm": 0.0039536841213703156,
      "learning_rate": 0.053580855800859285,
      "loss": 0.1471,
      "num_input_tokens_seen": 25015680,
      "step": 28890
    },
    {
      "epoch": 13.623290900518624,
      "grad_norm": 0.0029755316209048033,
      "learning_rate": 0.05353573982455938,
      "loss": 0.1855,
      "num_input_tokens_seen": 25020464,
      "step": 28895
    },
    {
      "epoch": 13.625648279113626,
      "grad_norm": 0.006905414164066315,
      "learning_rate": 0.053490638724260686,
      "loss": 0.2005,
      "num_input_tokens_seen": 25025296,
      "step": 28900
    },
    {
      "epoch": 13.628005657708629,
      "grad_norm": 0.004159688483923674,
      "learning_rate": 0.05344555250691827,
      "loss": 0.2131,
      "num_input_tokens_seen": 25029696,
      "step": 28905
    },
    {
      "epoch": 13.630363036303631,
      "grad_norm": 0.004368811380118132,
      "learning_rate": 0.053400481179485086,
      "loss": 0.1855,
      "num_input_tokens_seen": 25034592,
      "step": 28910
    },
    {
      "epoch": 13.632720414898632,
      "grad_norm": 0.004641393665224314,
      "learning_rate": 0.05335542474891159,
      "loss": 0.1797,
      "num_input_tokens_seen": 25039120,
      "step": 28915
    },
    {
      "epoch": 13.635077793493634,
      "grad_norm": 0.004671767819672823,
      "learning_rate": 0.053310383222146124,
      "loss": 0.2283,
      "num_input_tokens_seen": 25043648,
      "step": 28920
    },
    {
      "epoch": 13.637435172088637,
      "grad_norm": 0.0031480840407311916,
      "learning_rate": 0.053265356606134684,
      "loss": 0.2592,
      "num_input_tokens_seen": 25047728,
      "step": 28925
    },
    {
      "epoch": 13.639792550683639,
      "grad_norm": 0.002788742072880268,
      "learning_rate": 0.053220344907820856,
      "loss": 0.1512,
      "num_input_tokens_seen": 25052112,
      "step": 28930
    },
    {
      "epoch": 13.642149929278641,
      "grad_norm": 0.0038656184915453196,
      "learning_rate": 0.05317534813414608,
      "loss": 0.1938,
      "num_input_tokens_seen": 25056416,
      "step": 28935
    },
    {
      "epoch": 13.644507307873644,
      "grad_norm": 0.004766746889799833,
      "learning_rate": 0.05313036629204942,
      "loss": 0.1909,
      "num_input_tokens_seen": 25060096,
      "step": 28940
    },
    {
      "epoch": 13.646864686468646,
      "grad_norm": 0.004360992927104235,
      "learning_rate": 0.05308539938846756,
      "loss": 0.1666,
      "num_input_tokens_seen": 25064640,
      "step": 28945
    },
    {
      "epoch": 13.649222065063649,
      "grad_norm": 0.006198831368237734,
      "learning_rate": 0.05304044743033507,
      "loss": 0.2402,
      "num_input_tokens_seen": 25068736,
      "step": 28950
    },
    {
      "epoch": 13.651579443658651,
      "grad_norm": 0.0019321818836033344,
      "learning_rate": 0.05299551042458401,
      "loss": 0.151,
      "num_input_tokens_seen": 25072736,
      "step": 28955
    },
    {
      "epoch": 13.653936822253653,
      "grad_norm": 0.006511553190648556,
      "learning_rate": 0.052950588378144266,
      "loss": 0.218,
      "num_input_tokens_seen": 25077712,
      "step": 28960
    },
    {
      "epoch": 13.656294200848656,
      "grad_norm": 0.003966438118368387,
      "learning_rate": 0.052905681297943465,
      "loss": 0.186,
      "num_input_tokens_seen": 25082208,
      "step": 28965
    },
    {
      "epoch": 13.658651579443658,
      "grad_norm": 0.004760412033647299,
      "learning_rate": 0.0528607891909067,
      "loss": 0.2952,
      "num_input_tokens_seen": 25086864,
      "step": 28970
    },
    {
      "epoch": 13.66100895803866,
      "grad_norm": 0.004523079376667738,
      "learning_rate": 0.05281591206395697,
      "loss": 0.2032,
      "num_input_tokens_seen": 25091376,
      "step": 28975
    },
    {
      "epoch": 13.663366336633663,
      "grad_norm": 0.0040833428502082825,
      "learning_rate": 0.05277104992401496,
      "loss": 0.1861,
      "num_input_tokens_seen": 25095424,
      "step": 28980
    },
    {
      "epoch": 13.665723715228665,
      "grad_norm": 0.005317801609635353,
      "learning_rate": 0.05272620277799884,
      "loss": 0.2445,
      "num_input_tokens_seen": 25099360,
      "step": 28985
    },
    {
      "epoch": 13.668081093823668,
      "grad_norm": 0.0036685646045953035,
      "learning_rate": 0.05268137063282473,
      "loss": 0.1817,
      "num_input_tokens_seen": 25103376,
      "step": 28990
    },
    {
      "epoch": 13.67043847241867,
      "grad_norm": 0.0037000251468271017,
      "learning_rate": 0.0526365534954062,
      "loss": 0.2213,
      "num_input_tokens_seen": 25107440,
      "step": 28995
    },
    {
      "epoch": 13.672795851013673,
      "grad_norm": 0.004412584938108921,
      "learning_rate": 0.052591751372654656,
      "loss": 0.1911,
      "num_input_tokens_seen": 25110864,
      "step": 29000
    },
    {
      "epoch": 13.672795851013673,
      "eval_loss": 0.21386408805847168,
      "eval_runtime": 21.9346,
      "eval_samples_per_second": 42.991,
      "eval_steps_per_second": 21.519,
      "num_input_tokens_seen": 25110864,
      "step": 29000
    },
    {
      "epoch": 13.675153229608675,
      "grad_norm": 0.006533700041472912,
      "learning_rate": 0.05254696427147921,
      "loss": 0.2067,
      "num_input_tokens_seen": 25115472,
      "step": 29005
    },
    {
      "epoch": 13.677510608203677,
      "grad_norm": 0.0022053380962461233,
      "learning_rate": 0.052502192198786546,
      "loss": 0.1909,
      "num_input_tokens_seen": 25119696,
      "step": 29010
    },
    {
      "epoch": 13.67986798679868,
      "grad_norm": 0.003370769554749131,
      "learning_rate": 0.05245743516148103,
      "loss": 0.1861,
      "num_input_tokens_seen": 25123824,
      "step": 29015
    },
    {
      "epoch": 13.682225365393682,
      "grad_norm": 0.002879176288843155,
      "learning_rate": 0.05241269316646486,
      "loss": 0.1729,
      "num_input_tokens_seen": 25127536,
      "step": 29020
    },
    {
      "epoch": 13.684582743988685,
      "grad_norm": 0.0028984853997826576,
      "learning_rate": 0.052367966220637725,
      "loss": 0.1849,
      "num_input_tokens_seen": 25131536,
      "step": 29025
    },
    {
      "epoch": 13.686940122583687,
      "grad_norm": 0.004233913496136665,
      "learning_rate": 0.05232325433089716,
      "loss": 0.1804,
      "num_input_tokens_seen": 25136128,
      "step": 29030
    },
    {
      "epoch": 13.68929750117869,
      "grad_norm": 0.00387250492349267,
      "learning_rate": 0.052278557504138214,
      "loss": 0.192,
      "num_input_tokens_seen": 25140768,
      "step": 29035
    },
    {
      "epoch": 13.691654879773692,
      "grad_norm": 0.002160158473998308,
      "learning_rate": 0.05223387574725372,
      "loss": 0.1724,
      "num_input_tokens_seen": 25144608,
      "step": 29040
    },
    {
      "epoch": 13.694012258368694,
      "grad_norm": 0.004393074195832014,
      "learning_rate": 0.05218920906713428,
      "loss": 0.1912,
      "num_input_tokens_seen": 25148688,
      "step": 29045
    },
    {
      "epoch": 13.696369636963697,
      "grad_norm": 0.005205902736634016,
      "learning_rate": 0.05214455747066789,
      "loss": 0.221,
      "num_input_tokens_seen": 25154336,
      "step": 29050
    },
    {
      "epoch": 13.698727015558699,
      "grad_norm": 0.004779312759637833,
      "learning_rate": 0.05209992096474048,
      "loss": 0.2348,
      "num_input_tokens_seen": 25159040,
      "step": 29055
    },
    {
      "epoch": 13.701084394153701,
      "grad_norm": 0.003442067187279463,
      "learning_rate": 0.05205529955623559,
      "loss": 0.21,
      "num_input_tokens_seen": 25162768,
      "step": 29060
    },
    {
      "epoch": 13.703441772748704,
      "grad_norm": 0.003546150168403983,
      "learning_rate": 0.052010693252034314,
      "loss": 0.2063,
      "num_input_tokens_seen": 25166560,
      "step": 29065
    },
    {
      "epoch": 13.705799151343706,
      "grad_norm": 0.0046943132765591145,
      "learning_rate": 0.0519661020590156,
      "loss": 0.1803,
      "num_input_tokens_seen": 25170272,
      "step": 29070
    },
    {
      "epoch": 13.708156529938709,
      "grad_norm": 0.004767566453665495,
      "learning_rate": 0.05192152598405586,
      "loss": 0.1957,
      "num_input_tokens_seen": 25174272,
      "step": 29075
    },
    {
      "epoch": 13.710513908533711,
      "grad_norm": 0.003599602961912751,
      "learning_rate": 0.05187696503402941,
      "loss": 0.171,
      "num_input_tokens_seen": 25178688,
      "step": 29080
    },
    {
      "epoch": 13.712871287128714,
      "grad_norm": 0.004343706648796797,
      "learning_rate": 0.05183241921580798,
      "loss": 0.2229,
      "num_input_tokens_seen": 25182752,
      "step": 29085
    },
    {
      "epoch": 13.715228665723716,
      "grad_norm": 0.003426008392125368,
      "learning_rate": 0.051787888536261206,
      "loss": 0.2249,
      "num_input_tokens_seen": 25187344,
      "step": 29090
    },
    {
      "epoch": 13.717586044318718,
      "grad_norm": 0.004933008924126625,
      "learning_rate": 0.051743373002256184,
      "loss": 0.2249,
      "num_input_tokens_seen": 25191920,
      "step": 29095
    },
    {
      "epoch": 13.71994342291372,
      "grad_norm": 0.004095638170838356,
      "learning_rate": 0.05169887262065787,
      "loss": 0.1591,
      "num_input_tokens_seen": 25196368,
      "step": 29100
    },
    {
      "epoch": 13.722300801508723,
      "grad_norm": 0.0018408370669931173,
      "learning_rate": 0.051654387398328665,
      "loss": 0.1712,
      "num_input_tokens_seen": 25201264,
      "step": 29105
    },
    {
      "epoch": 13.724658180103724,
      "grad_norm": 0.003776341676712036,
      "learning_rate": 0.05160991734212888,
      "loss": 0.1595,
      "num_input_tokens_seen": 25205952,
      "step": 29110
    },
    {
      "epoch": 13.727015558698728,
      "grad_norm": 0.005363414995372295,
      "learning_rate": 0.051565462458916224,
      "loss": 0.1743,
      "num_input_tokens_seen": 25210208,
      "step": 29115
    },
    {
      "epoch": 13.729372937293729,
      "grad_norm": 0.003551613073796034,
      "learning_rate": 0.05152102275554627,
      "loss": 0.1999,
      "num_input_tokens_seen": 25214720,
      "step": 29120
    },
    {
      "epoch": 13.731730315888731,
      "grad_norm": 0.003412707010284066,
      "learning_rate": 0.05147659823887222,
      "loss": 0.2016,
      "num_input_tokens_seen": 25219040,
      "step": 29125
    },
    {
      "epoch": 13.734087694483733,
      "grad_norm": 0.004067674279212952,
      "learning_rate": 0.05143218891574479,
      "loss": 0.2051,
      "num_input_tokens_seen": 25222992,
      "step": 29130
    },
    {
      "epoch": 13.736445073078736,
      "grad_norm": 0.004270528443157673,
      "learning_rate": 0.0513877947930125,
      "loss": 0.2326,
      "num_input_tokens_seen": 25227472,
      "step": 29135
    },
    {
      "epoch": 13.738802451673738,
      "grad_norm": 0.003574499860405922,
      "learning_rate": 0.051343415877521566,
      "loss": 0.1979,
      "num_input_tokens_seen": 25232400,
      "step": 29140
    },
    {
      "epoch": 13.74115983026874,
      "grad_norm": 0.0026290391106158495,
      "learning_rate": 0.051299052176115634,
      "loss": 0.1689,
      "num_input_tokens_seen": 25236192,
      "step": 29145
    },
    {
      "epoch": 13.743517208863743,
      "grad_norm": 0.004557745531201363,
      "learning_rate": 0.051254703695636256,
      "loss": 0.2433,
      "num_input_tokens_seen": 25240640,
      "step": 29150
    },
    {
      "epoch": 13.745874587458745,
      "grad_norm": 0.005136494059115648,
      "learning_rate": 0.05121037044292249,
      "loss": 0.2148,
      "num_input_tokens_seen": 25244752,
      "step": 29155
    },
    {
      "epoch": 13.748231966053748,
      "grad_norm": 0.0033568604849278927,
      "learning_rate": 0.05116605242481101,
      "loss": 0.2303,
      "num_input_tokens_seen": 25249632,
      "step": 29160
    },
    {
      "epoch": 13.75058934464875,
      "grad_norm": 0.004388198256492615,
      "learning_rate": 0.05112174964813634,
      "loss": 0.224,
      "num_input_tokens_seen": 25254384,
      "step": 29165
    },
    {
      "epoch": 13.752946723243753,
      "grad_norm": 0.0033751139417290688,
      "learning_rate": 0.05107746211973038,
      "loss": 0.17,
      "num_input_tokens_seen": 25258208,
      "step": 29170
    },
    {
      "epoch": 13.755304101838755,
      "grad_norm": 0.0032822161447256804,
      "learning_rate": 0.05103318984642291,
      "loss": 0.1786,
      "num_input_tokens_seen": 25261840,
      "step": 29175
    },
    {
      "epoch": 13.757661480433757,
      "grad_norm": 0.004384009633213282,
      "learning_rate": 0.05098893283504131,
      "loss": 0.2024,
      "num_input_tokens_seen": 25266096,
      "step": 29180
    },
    {
      "epoch": 13.76001885902876,
      "grad_norm": 0.0037242916878312826,
      "learning_rate": 0.050944691092410475,
      "loss": 0.2024,
      "num_input_tokens_seen": 25270720,
      "step": 29185
    },
    {
      "epoch": 13.762376237623762,
      "grad_norm": 0.003847974119707942,
      "learning_rate": 0.05090046462535313,
      "loss": 0.218,
      "num_input_tokens_seen": 25275344,
      "step": 29190
    },
    {
      "epoch": 13.764733616218765,
      "grad_norm": 0.005201333202421665,
      "learning_rate": 0.050856253440689454,
      "loss": 0.2303,
      "num_input_tokens_seen": 25279920,
      "step": 29195
    },
    {
      "epoch": 13.767090994813767,
      "grad_norm": 0.004798797890543938,
      "learning_rate": 0.050812057545237405,
      "loss": 0.2123,
      "num_input_tokens_seen": 25284944,
      "step": 29200
    },
    {
      "epoch": 13.767090994813767,
      "eval_loss": 0.2117646038532257,
      "eval_runtime": 21.9302,
      "eval_samples_per_second": 43.0,
      "eval_steps_per_second": 21.523,
      "num_input_tokens_seen": 25284944,
      "step": 29200
    },
    {
      "epoch": 13.76944837340877,
      "grad_norm": 0.0035623477306216955,
      "learning_rate": 0.0507678769458126,
      "loss": 0.1987,
      "num_input_tokens_seen": 25288480,
      "step": 29205
    },
    {
      "epoch": 13.771805752003772,
      "grad_norm": 0.003694303799420595,
      "learning_rate": 0.050723711649228155,
      "loss": 0.1778,
      "num_input_tokens_seen": 25292736,
      "step": 29210
    },
    {
      "epoch": 13.774163130598774,
      "grad_norm": 0.004534947220236063,
      "learning_rate": 0.05067956166229496,
      "loss": 0.2103,
      "num_input_tokens_seen": 25297280,
      "step": 29215
    },
    {
      "epoch": 13.776520509193777,
      "grad_norm": 0.002619493054226041,
      "learning_rate": 0.05063542699182155,
      "loss": 0.2182,
      "num_input_tokens_seen": 25302320,
      "step": 29220
    },
    {
      "epoch": 13.778877887788779,
      "grad_norm": 0.0028981624636799097,
      "learning_rate": 0.050591307644613996,
      "loss": 0.1687,
      "num_input_tokens_seen": 25306416,
      "step": 29225
    },
    {
      "epoch": 13.781235266383781,
      "grad_norm": 0.004888403695076704,
      "learning_rate": 0.05054720362747599,
      "loss": 0.2358,
      "num_input_tokens_seen": 25310768,
      "step": 29230
    },
    {
      "epoch": 13.783592644978784,
      "grad_norm": 0.004227655474096537,
      "learning_rate": 0.050503114947209035,
      "loss": 0.2017,
      "num_input_tokens_seen": 25315056,
      "step": 29235
    },
    {
      "epoch": 13.785950023573786,
      "grad_norm": 0.0024586431682109833,
      "learning_rate": 0.05045904161061207,
      "loss": 0.1586,
      "num_input_tokens_seen": 25319584,
      "step": 29240
    },
    {
      "epoch": 13.788307402168789,
      "grad_norm": 0.004200305789709091,
      "learning_rate": 0.05041498362448185,
      "loss": 0.2146,
      "num_input_tokens_seen": 25323376,
      "step": 29245
    },
    {
      "epoch": 13.790664780763791,
      "grad_norm": 0.004687074571847916,
      "learning_rate": 0.05037094099561256,
      "loss": 0.1756,
      "num_input_tokens_seen": 25328160,
      "step": 29250
    },
    {
      "epoch": 13.793022159358793,
      "grad_norm": 0.004202248528599739,
      "learning_rate": 0.05032691373079624,
      "loss": 0.243,
      "num_input_tokens_seen": 25332464,
      "step": 29255
    },
    {
      "epoch": 13.795379537953796,
      "grad_norm": 0.004029267933219671,
      "learning_rate": 0.05028290183682234,
      "loss": 0.1876,
      "num_input_tokens_seen": 25336896,
      "step": 29260
    },
    {
      "epoch": 13.797736916548798,
      "grad_norm": 0.004172188229858875,
      "learning_rate": 0.050238905320478096,
      "loss": 0.1879,
      "num_input_tokens_seen": 25341312,
      "step": 29265
    },
    {
      "epoch": 13.8000942951438,
      "grad_norm": 0.002621263498440385,
      "learning_rate": 0.05019492418854838,
      "loss": 0.1602,
      "num_input_tokens_seen": 25345968,
      "step": 29270
    },
    {
      "epoch": 13.802451673738803,
      "grad_norm": 0.013429324142634869,
      "learning_rate": 0.05015095844781554,
      "loss": 0.2404,
      "num_input_tokens_seen": 25349728,
      "step": 29275
    },
    {
      "epoch": 13.804809052333805,
      "grad_norm": 0.0051779006607830524,
      "learning_rate": 0.05010700810505968,
      "loss": 0.2411,
      "num_input_tokens_seen": 25354032,
      "step": 29280
    },
    {
      "epoch": 13.807166430928808,
      "grad_norm": 0.006728032138198614,
      "learning_rate": 0.05006307316705856,
      "loss": 0.2325,
      "num_input_tokens_seen": 25357792,
      "step": 29285
    },
    {
      "epoch": 13.80952380952381,
      "grad_norm": 0.004422380588948727,
      "learning_rate": 0.0500191536405874,
      "loss": 0.2063,
      "num_input_tokens_seen": 25362128,
      "step": 29290
    },
    {
      "epoch": 13.811881188118813,
      "grad_norm": 0.0037285154685378075,
      "learning_rate": 0.04997524953241922,
      "loss": 0.2111,
      "num_input_tokens_seen": 25365600,
      "step": 29295
    },
    {
      "epoch": 13.814238566713815,
      "grad_norm": 0.003240319201722741,
      "learning_rate": 0.049931360849324556,
      "loss": 0.2122,
      "num_input_tokens_seen": 25369952,
      "step": 29300
    },
    {
      "epoch": 13.816595945308817,
      "grad_norm": 0.00333136273548007,
      "learning_rate": 0.04988748759807155,
      "loss": 0.1483,
      "num_input_tokens_seen": 25374816,
      "step": 29305
    },
    {
      "epoch": 13.81895332390382,
      "grad_norm": 0.003069463651627302,
      "learning_rate": 0.0498436297854261,
      "loss": 0.1511,
      "num_input_tokens_seen": 25380176,
      "step": 29310
    },
    {
      "epoch": 13.82131070249882,
      "grad_norm": 0.006015172693878412,
      "learning_rate": 0.04979978741815152,
      "loss": 0.1996,
      "num_input_tokens_seen": 25383856,
      "step": 29315
    },
    {
      "epoch": 13.823668081093825,
      "grad_norm": 0.00577114662155509,
      "learning_rate": 0.04975596050300891,
      "loss": 0.2117,
      "num_input_tokens_seen": 25387312,
      "step": 29320
    },
    {
      "epoch": 13.826025459688825,
      "grad_norm": 0.005388208199292421,
      "learning_rate": 0.049712149046757005,
      "loss": 0.1892,
      "num_input_tokens_seen": 25392192,
      "step": 29325
    },
    {
      "epoch": 13.828382838283828,
      "grad_norm": 0.004310495685786009,
      "learning_rate": 0.04966835305615194,
      "loss": 0.2302,
      "num_input_tokens_seen": 25396240,
      "step": 29330
    },
    {
      "epoch": 13.83074021687883,
      "grad_norm": 0.0033561259042471647,
      "learning_rate": 0.049624572537947755,
      "loss": 0.1944,
      "num_input_tokens_seen": 25400768,
      "step": 29335
    },
    {
      "epoch": 13.833097595473832,
      "grad_norm": 0.004013275261968374,
      "learning_rate": 0.04958080749889582,
      "loss": 0.2007,
      "num_input_tokens_seen": 25404960,
      "step": 29340
    },
    {
      "epoch": 13.835454974068835,
      "grad_norm": 0.005120350979268551,
      "learning_rate": 0.049537057945745304,
      "loss": 0.2266,
      "num_input_tokens_seen": 25408768,
      "step": 29345
    },
    {
      "epoch": 13.837812352663837,
      "grad_norm": 0.005392777267843485,
      "learning_rate": 0.049493323885243,
      "loss": 0.2128,
      "num_input_tokens_seen": 25413712,
      "step": 29350
    },
    {
      "epoch": 13.84016973125884,
      "grad_norm": 0.004702049307525158,
      "learning_rate": 0.04944960532413318,
      "loss": 0.1986,
      "num_input_tokens_seen": 25417888,
      "step": 29355
    },
    {
      "epoch": 13.842527109853842,
      "grad_norm": 0.0027801289688795805,
      "learning_rate": 0.049405902269157774,
      "loss": 0.1904,
      "num_input_tokens_seen": 25422176,
      "step": 29360
    },
    {
      "epoch": 13.844884488448844,
      "grad_norm": 0.0043770126067101955,
      "learning_rate": 0.04936221472705646,
      "loss": 0.2521,
      "num_input_tokens_seen": 25427216,
      "step": 29365
    },
    {
      "epoch": 13.847241867043847,
      "grad_norm": 0.00483175553381443,
      "learning_rate": 0.04931854270456632,
      "loss": 0.1828,
      "num_input_tokens_seen": 25431376,
      "step": 29370
    },
    {
      "epoch": 13.84959924563885,
      "grad_norm": 0.004217939916998148,
      "learning_rate": 0.049274886208422075,
      "loss": 0.2154,
      "num_input_tokens_seen": 25435680,
      "step": 29375
    },
    {
      "epoch": 13.851956624233852,
      "grad_norm": 0.0030798183288425207,
      "learning_rate": 0.049231245245356235,
      "loss": 0.2105,
      "num_input_tokens_seen": 25439936,
      "step": 29380
    },
    {
      "epoch": 13.854314002828854,
      "grad_norm": 0.004276847466826439,
      "learning_rate": 0.049187619822098655,
      "loss": 0.1837,
      "num_input_tokens_seen": 25444720,
      "step": 29385
    },
    {
      "epoch": 13.856671381423856,
      "grad_norm": 0.0035279253497719765,
      "learning_rate": 0.04914400994537705,
      "loss": 0.2108,
      "num_input_tokens_seen": 25448480,
      "step": 29390
    },
    {
      "epoch": 13.859028760018859,
      "grad_norm": 0.003942518029361963,
      "learning_rate": 0.049100415621916485,
      "loss": 0.1876,
      "num_input_tokens_seen": 25452464,
      "step": 29395
    },
    {
      "epoch": 13.861386138613861,
      "grad_norm": 0.002274898113682866,
      "learning_rate": 0.04905683685843981,
      "loss": 0.159,
      "num_input_tokens_seen": 25456816,
      "step": 29400
    },
    {
      "epoch": 13.861386138613861,
      "eval_loss": 0.2113555669784546,
      "eval_runtime": 21.9227,
      "eval_samples_per_second": 43.015,
      "eval_steps_per_second": 21.53,
      "num_input_tokens_seen": 25456816,
      "step": 29400
    },
    {
      "epoch": 13.863743517208864,
      "grad_norm": 0.003278976771980524,
      "learning_rate": 0.049013273661667495,
      "loss": 0.1819,
      "num_input_tokens_seen": 25461280,
      "step": 29405
    },
    {
      "epoch": 13.866100895803866,
      "grad_norm": 0.006098782643675804,
      "learning_rate": 0.048969726038317396,
      "loss": 0.2239,
      "num_input_tokens_seen": 25465664,
      "step": 29410
    },
    {
      "epoch": 13.868458274398868,
      "grad_norm": 0.005837530363351107,
      "learning_rate": 0.048926193995105206,
      "loss": 0.2407,
      "num_input_tokens_seen": 25469504,
      "step": 29415
    },
    {
      "epoch": 13.87081565299387,
      "grad_norm": 0.003387365723028779,
      "learning_rate": 0.048882677538744035,
      "loss": 0.1793,
      "num_input_tokens_seen": 25473072,
      "step": 29420
    },
    {
      "epoch": 13.873173031588873,
      "grad_norm": 0.0037284009158611298,
      "learning_rate": 0.048839176675944715,
      "loss": 0.2205,
      "num_input_tokens_seen": 25476624,
      "step": 29425
    },
    {
      "epoch": 13.875530410183876,
      "grad_norm": 0.004622723441570997,
      "learning_rate": 0.04879569141341566,
      "loss": 0.2055,
      "num_input_tokens_seen": 25482160,
      "step": 29430
    },
    {
      "epoch": 13.877887788778878,
      "grad_norm": 0.003325487021356821,
      "learning_rate": 0.04875222175786274,
      "loss": 0.1792,
      "num_input_tokens_seen": 25487296,
      "step": 29435
    },
    {
      "epoch": 13.88024516737388,
      "grad_norm": 0.0018232128350064158,
      "learning_rate": 0.04870876771598966,
      "loss": 0.2057,
      "num_input_tokens_seen": 25491056,
      "step": 29440
    },
    {
      "epoch": 13.882602545968883,
      "grad_norm": 0.002652060240507126,
      "learning_rate": 0.04866532929449744,
      "loss": 0.1739,
      "num_input_tokens_seen": 25495520,
      "step": 29445
    },
    {
      "epoch": 13.884959924563885,
      "grad_norm": 0.0031212049070745707,
      "learning_rate": 0.048621906500084945,
      "loss": 0.216,
      "num_input_tokens_seen": 25500352,
      "step": 29450
    },
    {
      "epoch": 13.887317303158888,
      "grad_norm": 0.0034292039927095175,
      "learning_rate": 0.04857849933944845,
      "loss": 0.1613,
      "num_input_tokens_seen": 25505200,
      "step": 29455
    },
    {
      "epoch": 13.88967468175389,
      "grad_norm": 0.005079376976937056,
      "learning_rate": 0.048535107819281866,
      "loss": 0.2591,
      "num_input_tokens_seen": 25509552,
      "step": 29460
    },
    {
      "epoch": 13.892032060348892,
      "grad_norm": 0.003466268302872777,
      "learning_rate": 0.04849173194627675,
      "loss": 0.1917,
      "num_input_tokens_seen": 25513840,
      "step": 29465
    },
    {
      "epoch": 13.894389438943895,
      "grad_norm": 0.002527260221540928,
      "learning_rate": 0.04844837172712223,
      "loss": 0.187,
      "num_input_tokens_seen": 25518016,
      "step": 29470
    },
    {
      "epoch": 13.896746817538897,
      "grad_norm": 0.0036825696006417274,
      "learning_rate": 0.04840502716850494,
      "loss": 0.181,
      "num_input_tokens_seen": 25522688,
      "step": 29475
    },
    {
      "epoch": 13.8991041961339,
      "grad_norm": 0.003721988992765546,
      "learning_rate": 0.04836169827710916,
      "loss": 0.1906,
      "num_input_tokens_seen": 25528400,
      "step": 29480
    },
    {
      "epoch": 13.901461574728902,
      "grad_norm": 0.004758345428854227,
      "learning_rate": 0.04831838505961684,
      "loss": 0.192,
      "num_input_tokens_seen": 25532496,
      "step": 29485
    },
    {
      "epoch": 13.903818953323904,
      "grad_norm": 0.003015707479789853,
      "learning_rate": 0.048275087522707295,
      "loss": 0.1698,
      "num_input_tokens_seen": 25536736,
      "step": 29490
    },
    {
      "epoch": 13.906176331918907,
      "grad_norm": 0.003071174258366227,
      "learning_rate": 0.04823180567305766,
      "loss": 0.1944,
      "num_input_tokens_seen": 25540912,
      "step": 29495
    },
    {
      "epoch": 13.90853371051391,
      "grad_norm": 0.004068693146109581,
      "learning_rate": 0.04818853951734244,
      "loss": 0.1948,
      "num_input_tokens_seen": 25544768,
      "step": 29500
    },
    {
      "epoch": 13.910891089108912,
      "grad_norm": 0.0025439232122153044,
      "learning_rate": 0.04814528906223387,
      "loss": 0.201,
      "num_input_tokens_seen": 25549200,
      "step": 29505
    },
    {
      "epoch": 13.913248467703912,
      "grad_norm": 0.0048871999606490135,
      "learning_rate": 0.04810205431440177,
      "loss": 0.1717,
      "num_input_tokens_seen": 25552816,
      "step": 29510
    },
    {
      "epoch": 13.915605846298917,
      "grad_norm": 0.0030690126586705446,
      "learning_rate": 0.04805883528051341,
      "loss": 0.1776,
      "num_input_tokens_seen": 25557744,
      "step": 29515
    },
    {
      "epoch": 13.917963224893917,
      "grad_norm": 0.004626222420483828,
      "learning_rate": 0.048015631967233685,
      "loss": 0.1628,
      "num_input_tokens_seen": 25561344,
      "step": 29520
    },
    {
      "epoch": 13.92032060348892,
      "grad_norm": 0.002297256840392947,
      "learning_rate": 0.04797244438122517,
      "loss": 0.1723,
      "num_input_tokens_seen": 25566496,
      "step": 29525
    },
    {
      "epoch": 13.922677982083922,
      "grad_norm": 0.004911480471491814,
      "learning_rate": 0.04792927252914784,
      "loss": 0.1994,
      "num_input_tokens_seen": 25570800,
      "step": 29530
    },
    {
      "epoch": 13.925035360678924,
      "grad_norm": 0.00404330063611269,
      "learning_rate": 0.04788611641765944,
      "loss": 0.2078,
      "num_input_tokens_seen": 25574816,
      "step": 29535
    },
    {
      "epoch": 13.927392739273927,
      "grad_norm": 0.0038285688497126102,
      "learning_rate": 0.04784297605341508,
      "loss": 0.1808,
      "num_input_tokens_seen": 25579936,
      "step": 29540
    },
    {
      "epoch": 13.92975011786893,
      "grad_norm": 0.005920145660638809,
      "learning_rate": 0.04779985144306761,
      "loss": 0.2246,
      "num_input_tokens_seen": 25584032,
      "step": 29545
    },
    {
      "epoch": 13.932107496463932,
      "grad_norm": 0.0023077167570590973,
      "learning_rate": 0.047756742593267405,
      "loss": 0.1561,
      "num_input_tokens_seen": 25588160,
      "step": 29550
    },
    {
      "epoch": 13.934464875058934,
      "grad_norm": 0.004504472482949495,
      "learning_rate": 0.047713649510662315,
      "loss": 0.2124,
      "num_input_tokens_seen": 25592272,
      "step": 29555
    },
    {
      "epoch": 13.936822253653936,
      "grad_norm": 0.004626700188964605,
      "learning_rate": 0.04767057220189789,
      "loss": 0.1976,
      "num_input_tokens_seen": 25596544,
      "step": 29560
    },
    {
      "epoch": 13.939179632248939,
      "grad_norm": 0.004087931476533413,
      "learning_rate": 0.04762751067361722,
      "loss": 0.1797,
      "num_input_tokens_seen": 25600512,
      "step": 29565
    },
    {
      "epoch": 13.941537010843941,
      "grad_norm": 0.0039518242701888084,
      "learning_rate": 0.04758446493246086,
      "loss": 0.1348,
      "num_input_tokens_seen": 25604240,
      "step": 29570
    },
    {
      "epoch": 13.943894389438944,
      "grad_norm": 0.0029246718622744083,
      "learning_rate": 0.047541434985067084,
      "loss": 0.1359,
      "num_input_tokens_seen": 25609344,
      "step": 29575
    },
    {
      "epoch": 13.946251768033946,
      "grad_norm": 0.005112757440656424,
      "learning_rate": 0.047498420838071556,
      "loss": 0.1648,
      "num_input_tokens_seen": 25613728,
      "step": 29580
    },
    {
      "epoch": 13.948609146628948,
      "grad_norm": 0.00202812347561121,
      "learning_rate": 0.04745542249810772,
      "loss": 0.1811,
      "num_input_tokens_seen": 25618240,
      "step": 29585
    },
    {
      "epoch": 13.95096652522395,
      "grad_norm": 0.003631796222180128,
      "learning_rate": 0.047412439971806324,
      "loss": 0.1674,
      "num_input_tokens_seen": 25622240,
      "step": 29590
    },
    {
      "epoch": 13.953323903818953,
      "grad_norm": 0.003187691094353795,
      "learning_rate": 0.04736947326579592,
      "loss": 0.192,
      "num_input_tokens_seen": 25627232,
      "step": 29595
    },
    {
      "epoch": 13.955681282413956,
      "grad_norm": 0.006024655885994434,
      "learning_rate": 0.04732652238670245,
      "loss": 0.1791,
      "num_input_tokens_seen": 25631728,
      "step": 29600
    },
    {
      "epoch": 13.955681282413956,
      "eval_loss": 0.2154432088136673,
      "eval_runtime": 21.8926,
      "eval_samples_per_second": 43.074,
      "eval_steps_per_second": 21.56,
      "num_input_tokens_seen": 25631728,
      "step": 29600
    },
    {
      "epoch": 13.958038661008958,
      "grad_norm": 0.004916062578558922,
      "learning_rate": 0.04728358734114952,
      "loss": 0.179,
      "num_input_tokens_seen": 25635200,
      "step": 29605
    },
    {
      "epoch": 13.96039603960396,
      "grad_norm": 0.004516012966632843,
      "learning_rate": 0.04724066813575821,
      "loss": 0.1609,
      "num_input_tokens_seen": 25639456,
      "step": 29610
    },
    {
      "epoch": 13.962753418198963,
      "grad_norm": 0.0034733896609395742,
      "learning_rate": 0.04719776477714729,
      "loss": 0.2103,
      "num_input_tokens_seen": 25643824,
      "step": 29615
    },
    {
      "epoch": 13.965110796793965,
      "grad_norm": 0.004846430383622646,
      "learning_rate": 0.047154877271932856,
      "loss": 0.2005,
      "num_input_tokens_seen": 25648688,
      "step": 29620
    },
    {
      "epoch": 13.967468175388968,
      "grad_norm": 0.005264287814497948,
      "learning_rate": 0.0471120056267288,
      "loss": 0.1913,
      "num_input_tokens_seen": 25652464,
      "step": 29625
    },
    {
      "epoch": 13.96982555398397,
      "grad_norm": 0.005964715965092182,
      "learning_rate": 0.047069149848146495,
      "loss": 0.2207,
      "num_input_tokens_seen": 25657248,
      "step": 29630
    },
    {
      "epoch": 13.972182932578972,
      "grad_norm": 0.006545036565512419,
      "learning_rate": 0.04702630994279473,
      "loss": 0.1589,
      "num_input_tokens_seen": 25661472,
      "step": 29635
    },
    {
      "epoch": 13.974540311173975,
      "grad_norm": 0.004161347169429064,
      "learning_rate": 0.046983485917280035,
      "loss": 0.2047,
      "num_input_tokens_seen": 25665584,
      "step": 29640
    },
    {
      "epoch": 13.976897689768977,
      "grad_norm": 0.005011442583054304,
      "learning_rate": 0.04694067777820644,
      "loss": 0.1542,
      "num_input_tokens_seen": 25669536,
      "step": 29645
    },
    {
      "epoch": 13.97925506836398,
      "grad_norm": 0.0025283421855419874,
      "learning_rate": 0.046897885532175415,
      "loss": 0.1339,
      "num_input_tokens_seen": 25673584,
      "step": 29650
    },
    {
      "epoch": 13.981612446958982,
      "grad_norm": 0.004765721503645182,
      "learning_rate": 0.04685510918578613,
      "loss": 0.1873,
      "num_input_tokens_seen": 25678592,
      "step": 29655
    },
    {
      "epoch": 13.983969825553984,
      "grad_norm": 0.005070932675153017,
      "learning_rate": 0.04681234874563519,
      "loss": 0.1632,
      "num_input_tokens_seen": 25682640,
      "step": 29660
    },
    {
      "epoch": 13.986327204148987,
      "grad_norm": 0.003561283927410841,
      "learning_rate": 0.046769604218316836,
      "loss": 0.1579,
      "num_input_tokens_seen": 25686704,
      "step": 29665
    },
    {
      "epoch": 13.98868458274399,
      "grad_norm": 0.0030871948692947626,
      "learning_rate": 0.04672687561042279,
      "loss": 0.1538,
      "num_input_tokens_seen": 25690832,
      "step": 29670
    },
    {
      "epoch": 13.991041961338992,
      "grad_norm": 0.004767861682921648,
      "learning_rate": 0.046684162928542286,
      "loss": 0.2175,
      "num_input_tokens_seen": 25694864,
      "step": 29675
    },
    {
      "epoch": 13.993399339933994,
      "grad_norm": 0.0059538334608078,
      "learning_rate": 0.04664146617926222,
      "loss": 0.2235,
      "num_input_tokens_seen": 25699120,
      "step": 29680
    },
    {
      "epoch": 13.995756718528996,
      "grad_norm": 0.004485993180423975,
      "learning_rate": 0.046598785369167,
      "loss": 0.2127,
      "num_input_tokens_seen": 25704080,
      "step": 29685
    },
    {
      "epoch": 13.998114097123999,
      "grad_norm": 0.0056234849616885185,
      "learning_rate": 0.046556120504838434,
      "loss": 0.1866,
      "num_input_tokens_seen": 25708384,
      "step": 29690
    },
    {
      "epoch": 14.000471475719001,
      "grad_norm": 0.0055489493533968925,
      "learning_rate": 0.04651347159285609,
      "loss": 0.183,
      "num_input_tokens_seen": 25712288,
      "step": 29695
    },
    {
      "epoch": 14.002828854314004,
      "grad_norm": 0.005752101074904203,
      "learning_rate": 0.04647083863979688,
      "loss": 0.2187,
      "num_input_tokens_seen": 25716256,
      "step": 29700
    },
    {
      "epoch": 14.005186232909006,
      "grad_norm": 0.005097091663628817,
      "learning_rate": 0.04642822165223538,
      "loss": 0.1802,
      "num_input_tokens_seen": 25720176,
      "step": 29705
    },
    {
      "epoch": 14.007543611504008,
      "grad_norm": 0.003930123522877693,
      "learning_rate": 0.046385620636743716,
      "loss": 0.1534,
      "num_input_tokens_seen": 25724432,
      "step": 29710
    },
    {
      "epoch": 14.009900990099009,
      "grad_norm": 0.0047265589237213135,
      "learning_rate": 0.04634303559989141,
      "loss": 0.1617,
      "num_input_tokens_seen": 25729664,
      "step": 29715
    },
    {
      "epoch": 14.012258368694011,
      "grad_norm": 0.004092021379619837,
      "learning_rate": 0.046300466548245635,
      "loss": 0.1692,
      "num_input_tokens_seen": 25734256,
      "step": 29720
    },
    {
      "epoch": 14.014615747289014,
      "grad_norm": 0.005378470290452242,
      "learning_rate": 0.04625791348837114,
      "loss": 0.1481,
      "num_input_tokens_seen": 25738176,
      "step": 29725
    },
    {
      "epoch": 14.016973125884016,
      "grad_norm": 0.003937562461942434,
      "learning_rate": 0.046215376426830095,
      "loss": 0.1804,
      "num_input_tokens_seen": 25741680,
      "step": 29730
    },
    {
      "epoch": 14.019330504479019,
      "grad_norm": 0.007620553951710463,
      "learning_rate": 0.04617285537018219,
      "loss": 0.1974,
      "num_input_tokens_seen": 25745440,
      "step": 29735
    },
    {
      "epoch": 14.021687883074021,
      "grad_norm": 0.00842153187841177,
      "learning_rate": 0.046130350324984803,
      "loss": 0.2265,
      "num_input_tokens_seen": 25750128,
      "step": 29740
    },
    {
      "epoch": 14.024045261669023,
      "grad_norm": 0.0057876682840287685,
      "learning_rate": 0.046087861297792666,
      "loss": 0.1647,
      "num_input_tokens_seen": 25755088,
      "step": 29745
    },
    {
      "epoch": 14.026402640264026,
      "grad_norm": 0.004547099117189646,
      "learning_rate": 0.0460453882951582,
      "loss": 0.1448,
      "num_input_tokens_seen": 25759072,
      "step": 29750
    },
    {
      "epoch": 14.028760018859028,
      "grad_norm": 0.004431622568517923,
      "learning_rate": 0.04600293132363119,
      "loss": 0.1455,
      "num_input_tokens_seen": 25763312,
      "step": 29755
    },
    {
      "epoch": 14.03111739745403,
      "grad_norm": 0.0062558613717556,
      "learning_rate": 0.045960490389759086,
      "loss": 0.1928,
      "num_input_tokens_seen": 25767984,
      "step": 29760
    },
    {
      "epoch": 14.033474776049033,
      "grad_norm": 0.0032056611962616444,
      "learning_rate": 0.04591806550008685,
      "loss": 0.1589,
      "num_input_tokens_seen": 25771824,
      "step": 29765
    },
    {
      "epoch": 14.035832154644035,
      "grad_norm": 0.004509766586124897,
      "learning_rate": 0.045875656661156825,
      "loss": 0.2244,
      "num_input_tokens_seen": 25776048,
      "step": 29770
    },
    {
      "epoch": 14.038189533239038,
      "grad_norm": 0.006259105633944273,
      "learning_rate": 0.04583326387950911,
      "loss": 0.1994,
      "num_input_tokens_seen": 25780816,
      "step": 29775
    },
    {
      "epoch": 14.04054691183404,
      "grad_norm": 0.004545127507299185,
      "learning_rate": 0.0457908871616811,
      "loss": 0.1499,
      "num_input_tokens_seen": 25784992,
      "step": 29780
    },
    {
      "epoch": 14.042904290429043,
      "grad_norm": 0.005558746866881847,
      "learning_rate": 0.04574852651420786,
      "loss": 0.2006,
      "num_input_tokens_seen": 25788352,
      "step": 29785
    },
    {
      "epoch": 14.045261669024045,
      "grad_norm": 0.007731474470347166,
      "learning_rate": 0.045706181943621985,
      "loss": 0.1529,
      "num_input_tokens_seen": 25792400,
      "step": 29790
    },
    {
      "epoch": 14.047619047619047,
      "grad_norm": 0.007570208050310612,
      "learning_rate": 0.04566385345645344,
      "loss": 0.17,
      "num_input_tokens_seen": 25796800,
      "step": 29795
    },
    {
      "epoch": 14.04997642621405,
      "grad_norm": 0.004104857798665762,
      "learning_rate": 0.04562154105922993,
      "loss": 0.1437,
      "num_input_tokens_seen": 25801056,
      "step": 29800
    },
    {
      "epoch": 14.04997642621405,
      "eval_loss": 0.21508099138736725,
      "eval_runtime": 21.9567,
      "eval_samples_per_second": 42.948,
      "eval_steps_per_second": 21.497,
      "num_input_tokens_seen": 25801056,
      "step": 29800
    },
    {
      "epoch": 14.052333804809052,
      "grad_norm": 0.0056895907036960125,
      "learning_rate": 0.04557924475847642,
      "loss": 0.129,
      "num_input_tokens_seen": 25805520,
      "step": 29805
    },
    {
      "epoch": 14.054691183404055,
      "grad_norm": 0.004744086880236864,
      "learning_rate": 0.04553696456071567,
      "loss": 0.2101,
      "num_input_tokens_seen": 25810032,
      "step": 29810
    },
    {
      "epoch": 14.057048561999057,
      "grad_norm": 0.007415502332150936,
      "learning_rate": 0.045494700472467724,
      "loss": 0.1935,
      "num_input_tokens_seen": 25814080,
      "step": 29815
    },
    {
      "epoch": 14.05940594059406,
      "grad_norm": 0.008217088878154755,
      "learning_rate": 0.04545245250025024,
      "loss": 0.192,
      "num_input_tokens_seen": 25818144,
      "step": 29820
    },
    {
      "epoch": 14.061763319189062,
      "grad_norm": 0.004311414435505867,
      "learning_rate": 0.045410220650578384,
      "loss": 0.1486,
      "num_input_tokens_seen": 25822496,
      "step": 29825
    },
    {
      "epoch": 14.064120697784064,
      "grad_norm": 0.0026819631457328796,
      "learning_rate": 0.04536800492996492,
      "loss": 0.1126,
      "num_input_tokens_seen": 25826768,
      "step": 29830
    },
    {
      "epoch": 14.066478076379067,
      "grad_norm": 0.005270141176879406,
      "learning_rate": 0.04532580534491994,
      "loss": 0.1503,
      "num_input_tokens_seen": 25831408,
      "step": 29835
    },
    {
      "epoch": 14.068835454974069,
      "grad_norm": 0.00592161575332284,
      "learning_rate": 0.045283621901951183,
      "loss": 0.1774,
      "num_input_tokens_seen": 25835744,
      "step": 29840
    },
    {
      "epoch": 14.071192833569071,
      "grad_norm": 0.005093283019959927,
      "learning_rate": 0.04524145460756393,
      "loss": 0.1582,
      "num_input_tokens_seen": 25840144,
      "step": 29845
    },
    {
      "epoch": 14.073550212164074,
      "grad_norm": 0.005124999210238457,
      "learning_rate": 0.045199303468260794,
      "loss": 0.1714,
      "num_input_tokens_seen": 25844736,
      "step": 29850
    },
    {
      "epoch": 14.075907590759076,
      "grad_norm": 0.00470344303175807,
      "learning_rate": 0.04515716849054214,
      "loss": 0.1079,
      "num_input_tokens_seen": 25849056,
      "step": 29855
    },
    {
      "epoch": 14.078264969354079,
      "grad_norm": 0.006090222857892513,
      "learning_rate": 0.04511504968090558,
      "loss": 0.1282,
      "num_input_tokens_seen": 25852992,
      "step": 29860
    },
    {
      "epoch": 14.080622347949081,
      "grad_norm": 0.005547447130084038,
      "learning_rate": 0.04507294704584644,
      "loss": 0.1802,
      "num_input_tokens_seen": 25856880,
      "step": 29865
    },
    {
      "epoch": 14.082979726544083,
      "grad_norm": 0.00720406835898757,
      "learning_rate": 0.04503086059185749,
      "loss": 0.2297,
      "num_input_tokens_seen": 25860864,
      "step": 29870
    },
    {
      "epoch": 14.085337105139086,
      "grad_norm": 0.009011568501591682,
      "learning_rate": 0.04498879032542893,
      "loss": 0.1842,
      "num_input_tokens_seen": 25865616,
      "step": 29875
    },
    {
      "epoch": 14.087694483734088,
      "grad_norm": 0.0037466883659362793,
      "learning_rate": 0.0449467362530486,
      "loss": 0.1521,
      "num_input_tokens_seen": 25869520,
      "step": 29880
    },
    {
      "epoch": 14.09005186232909,
      "grad_norm": 0.007506763096898794,
      "learning_rate": 0.04490469838120171,
      "loss": 0.2256,
      "num_input_tokens_seen": 25873776,
      "step": 29885
    },
    {
      "epoch": 14.092409240924093,
      "grad_norm": 0.004818316549062729,
      "learning_rate": 0.04486267671637101,
      "loss": 0.1382,
      "num_input_tokens_seen": 25878528,
      "step": 29890
    },
    {
      "epoch": 14.094766619519095,
      "grad_norm": 0.00964922271668911,
      "learning_rate": 0.04482067126503683,
      "loss": 0.1883,
      "num_input_tokens_seen": 25883376,
      "step": 29895
    },
    {
      "epoch": 14.097123998114098,
      "grad_norm": 0.00368486437946558,
      "learning_rate": 0.04477868203367687,
      "loss": 0.1315,
      "num_input_tokens_seen": 25887552,
      "step": 29900
    },
    {
      "epoch": 14.0994813767091,
      "grad_norm": 0.00603363336995244,
      "learning_rate": 0.044736709028766426,
      "loss": 0.1542,
      "num_input_tokens_seen": 25891664,
      "step": 29905
    },
    {
      "epoch": 14.101838755304103,
      "grad_norm": 0.011149140074849129,
      "learning_rate": 0.04469475225677832,
      "loss": 0.153,
      "num_input_tokens_seen": 25895584,
      "step": 29910
    },
    {
      "epoch": 14.104196133899103,
      "grad_norm": 0.00668503949418664,
      "learning_rate": 0.04465281172418273,
      "loss": 0.16,
      "num_input_tokens_seen": 25899584,
      "step": 29915
    },
    {
      "epoch": 14.106553512494106,
      "grad_norm": 0.004469035658985376,
      "learning_rate": 0.044610887437447476,
      "loss": 0.1903,
      "num_input_tokens_seen": 25904576,
      "step": 29920
    },
    {
      "epoch": 14.108910891089108,
      "grad_norm": 0.0034630061127245426,
      "learning_rate": 0.044568979403037744,
      "loss": 0.1406,
      "num_input_tokens_seen": 25909168,
      "step": 29925
    },
    {
      "epoch": 14.11126826968411,
      "grad_norm": 0.006076529622077942,
      "learning_rate": 0.04452708762741631,
      "loss": 0.2503,
      "num_input_tokens_seen": 25914672,
      "step": 29930
    },
    {
      "epoch": 14.113625648279113,
      "grad_norm": 0.013170834630727768,
      "learning_rate": 0.044485212117043475,
      "loss": 0.2124,
      "num_input_tokens_seen": 25918960,
      "step": 29935
    },
    {
      "epoch": 14.115983026874115,
      "grad_norm": 0.004110877402126789,
      "learning_rate": 0.04444335287837687,
      "loss": 0.1282,
      "num_input_tokens_seen": 25923808,
      "step": 29940
    },
    {
      "epoch": 14.118340405469118,
      "grad_norm": 0.007103634998202324,
      "learning_rate": 0.04440150991787179,
      "loss": 0.1413,
      "num_input_tokens_seen": 25928528,
      "step": 29945
    },
    {
      "epoch": 14.12069778406412,
      "grad_norm": 0.004805678967386484,
      "learning_rate": 0.04435968324198088,
      "loss": 0.161,
      "num_input_tokens_seen": 25933408,
      "step": 29950
    },
    {
      "epoch": 14.123055162659123,
      "grad_norm": 0.010558979585766792,
      "learning_rate": 0.04431787285715442,
      "loss": 0.1442,
      "num_input_tokens_seen": 25938464,
      "step": 29955
    },
    {
      "epoch": 14.125412541254125,
      "grad_norm": 0.007052351254969835,
      "learning_rate": 0.04427607876984004,
      "loss": 0.2052,
      "num_input_tokens_seen": 25943168,
      "step": 29960
    },
    {
      "epoch": 14.127769919849127,
      "grad_norm": 0.007279901299625635,
      "learning_rate": 0.044234300986482886,
      "loss": 0.1669,
      "num_input_tokens_seen": 25947872,
      "step": 29965
    },
    {
      "epoch": 14.13012729844413,
      "grad_norm": 0.004778577946126461,
      "learning_rate": 0.04419253951352566,
      "loss": 0.1448,
      "num_input_tokens_seen": 25952608,
      "step": 29970
    },
    {
      "epoch": 14.132484677039132,
      "grad_norm": 0.006930164061486721,
      "learning_rate": 0.044150794357408533,
      "loss": 0.1422,
      "num_input_tokens_seen": 25956704,
      "step": 29975
    },
    {
      "epoch": 14.134842055634135,
      "grad_norm": 0.002455734880641103,
      "learning_rate": 0.044109065524569065,
      "loss": 0.1383,
      "num_input_tokens_seen": 25962080,
      "step": 29980
    },
    {
      "epoch": 14.137199434229137,
      "grad_norm": 0.007291420828551054,
      "learning_rate": 0.0440673530214424,
      "loss": 0.2469,
      "num_input_tokens_seen": 25966352,
      "step": 29985
    },
    {
      "epoch": 14.13955681282414,
      "grad_norm": 0.009213839657604694,
      "learning_rate": 0.04402565685446117,
      "loss": 0.2125,
      "num_input_tokens_seen": 25970224,
      "step": 29990
    },
    {
      "epoch": 14.141914191419142,
      "grad_norm": 0.00779935996979475,
      "learning_rate": 0.04398397703005536,
      "loss": 0.1142,
      "num_input_tokens_seen": 25973968,
      "step": 29995
    },
    {
      "epoch": 14.144271570014144,
      "grad_norm": 0.008030143566429615,
      "learning_rate": 0.043942313554652626,
      "loss": 0.2108,
      "num_input_tokens_seen": 25978896,
      "step": 30000
    },
    {
      "epoch": 14.144271570014144,
      "eval_loss": 0.2241128385066986,
      "eval_runtime": 21.9301,
      "eval_samples_per_second": 43.0,
      "eval_steps_per_second": 21.523,
      "num_input_tokens_seen": 25978896,
      "step": 30000
    },
    {
      "epoch": 14.146628948609147,
      "grad_norm": 0.006169900763779879,
      "learning_rate": 0.0439006664346779,
      "loss": 0.1378,
      "num_input_tokens_seen": 25983040,
      "step": 30005
    },
    {
      "epoch": 14.148986327204149,
      "grad_norm": 0.0065635498613119125,
      "learning_rate": 0.043859035676553755,
      "loss": 0.1813,
      "num_input_tokens_seen": 25987488,
      "step": 30010
    },
    {
      "epoch": 14.151343705799151,
      "grad_norm": 0.0036043680738657713,
      "learning_rate": 0.043817421286700194,
      "loss": 0.1615,
      "num_input_tokens_seen": 25992016,
      "step": 30015
    },
    {
      "epoch": 14.153701084394154,
      "grad_norm": 0.003938963636755943,
      "learning_rate": 0.043775823271534585,
      "loss": 0.1707,
      "num_input_tokens_seen": 25996160,
      "step": 30020
    },
    {
      "epoch": 14.156058462989156,
      "grad_norm": 0.004694386385381222,
      "learning_rate": 0.04373424163747197,
      "loss": 0.186,
      "num_input_tokens_seen": 26000816,
      "step": 30025
    },
    {
      "epoch": 14.158415841584159,
      "grad_norm": 0.008192849345505238,
      "learning_rate": 0.04369267639092473,
      "loss": 0.1356,
      "num_input_tokens_seen": 26005232,
      "step": 30030
    },
    {
      "epoch": 14.160773220179161,
      "grad_norm": 0.007724672090262175,
      "learning_rate": 0.04365112753830268,
      "loss": 0.2168,
      "num_input_tokens_seen": 26009088,
      "step": 30035
    },
    {
      "epoch": 14.163130598774163,
      "grad_norm": 0.007274783682078123,
      "learning_rate": 0.04360959508601327,
      "loss": 0.1388,
      "num_input_tokens_seen": 26013232,
      "step": 30040
    },
    {
      "epoch": 14.165487977369166,
      "grad_norm": 0.014119734056293964,
      "learning_rate": 0.04356807904046123,
      "loss": 0.1811,
      "num_input_tokens_seen": 26016976,
      "step": 30045
    },
    {
      "epoch": 14.167845355964168,
      "grad_norm": 0.004459151532500982,
      "learning_rate": 0.04352657940804892,
      "loss": 0.2109,
      "num_input_tokens_seen": 26021872,
      "step": 30050
    },
    {
      "epoch": 14.17020273455917,
      "grad_norm": 0.0042907241731882095,
      "learning_rate": 0.04348509619517613,
      "loss": 0.1266,
      "num_input_tokens_seen": 26026352,
      "step": 30055
    },
    {
      "epoch": 14.172560113154173,
      "grad_norm": 0.012276563793420792,
      "learning_rate": 0.04344362940824002,
      "loss": 0.242,
      "num_input_tokens_seen": 26031728,
      "step": 30060
    },
    {
      "epoch": 14.174917491749175,
      "grad_norm": 0.006126211956143379,
      "learning_rate": 0.04340217905363533,
      "loss": 0.1632,
      "num_input_tokens_seen": 26035840,
      "step": 30065
    },
    {
      "epoch": 14.177274870344178,
      "grad_norm": 0.0074376813136041164,
      "learning_rate": 0.04336074513775425,
      "loss": 0.1767,
      "num_input_tokens_seen": 26039360,
      "step": 30070
    },
    {
      "epoch": 14.17963224893918,
      "grad_norm": 0.0062879156321287155,
      "learning_rate": 0.04331932766698636,
      "loss": 0.1471,
      "num_input_tokens_seen": 26043696,
      "step": 30075
    },
    {
      "epoch": 14.181989627534183,
      "grad_norm": 0.005722375586628914,
      "learning_rate": 0.0432779266477188,
      "loss": 0.1465,
      "num_input_tokens_seen": 26047680,
      "step": 30080
    },
    {
      "epoch": 14.184347006129185,
      "grad_norm": 0.0045995512045919895,
      "learning_rate": 0.04323654208633607,
      "loss": 0.153,
      "num_input_tokens_seen": 26052208,
      "step": 30085
    },
    {
      "epoch": 14.186704384724187,
      "grad_norm": 0.007991827093064785,
      "learning_rate": 0.04319517398922024,
      "loss": 0.122,
      "num_input_tokens_seen": 26057408,
      "step": 30090
    },
    {
      "epoch": 14.18906176331919,
      "grad_norm": 0.0077020348981022835,
      "learning_rate": 0.04315382236275079,
      "loss": 0.1831,
      "num_input_tokens_seen": 26062224,
      "step": 30095
    },
    {
      "epoch": 14.191419141914192,
      "grad_norm": 0.0075521995313465595,
      "learning_rate": 0.043112487213304664,
      "loss": 0.2506,
      "num_input_tokens_seen": 26067664,
      "step": 30100
    },
    {
      "epoch": 14.193776520509195,
      "grad_norm": 0.010488913394510746,
      "learning_rate": 0.04307116854725618,
      "loss": 0.1798,
      "num_input_tokens_seen": 26071520,
      "step": 30105
    },
    {
      "epoch": 14.196133899104197,
      "grad_norm": 0.005848318804055452,
      "learning_rate": 0.043029866370977325,
      "loss": 0.1824,
      "num_input_tokens_seen": 26076064,
      "step": 30110
    },
    {
      "epoch": 14.198491277699198,
      "grad_norm": 0.0107017382979393,
      "learning_rate": 0.04298858069083728,
      "loss": 0.1961,
      "num_input_tokens_seen": 26080656,
      "step": 30115
    },
    {
      "epoch": 14.2008486562942,
      "grad_norm": 0.007990701124072075,
      "learning_rate": 0.04294731151320295,
      "loss": 0.1797,
      "num_input_tokens_seen": 26085328,
      "step": 30120
    },
    {
      "epoch": 14.203206034889202,
      "grad_norm": 0.008625784888863564,
      "learning_rate": 0.04290605884443841,
      "loss": 0.1824,
      "num_input_tokens_seen": 26089568,
      "step": 30125
    },
    {
      "epoch": 14.205563413484205,
      "grad_norm": 0.006376832723617554,
      "learning_rate": 0.04286482269090545,
      "loss": 0.1769,
      "num_input_tokens_seen": 26094128,
      "step": 30130
    },
    {
      "epoch": 14.207920792079207,
      "grad_norm": 0.004033069591969252,
      "learning_rate": 0.04282360305896323,
      "loss": 0.1243,
      "num_input_tokens_seen": 26098976,
      "step": 30135
    },
    {
      "epoch": 14.21027817067421,
      "grad_norm": 0.0039036825764924288,
      "learning_rate": 0.04278239995496822,
      "loss": 0.1649,
      "num_input_tokens_seen": 26103744,
      "step": 30140
    },
    {
      "epoch": 14.212635549269212,
      "grad_norm": 0.008768069557845592,
      "learning_rate": 0.042741213385274514,
      "loss": 0.1505,
      "num_input_tokens_seen": 26107776,
      "step": 30145
    },
    {
      "epoch": 14.214992927864214,
      "grad_norm": 0.00647443812340498,
      "learning_rate": 0.04270004335623366,
      "loss": 0.1826,
      "num_input_tokens_seen": 26112032,
      "step": 30150
    },
    {
      "epoch": 14.217350306459217,
      "grad_norm": 0.007099487353116274,
      "learning_rate": 0.04265888987419448,
      "loss": 0.2343,
      "num_input_tokens_seen": 26116384,
      "step": 30155
    },
    {
      "epoch": 14.21970768505422,
      "grad_norm": 0.007134127430617809,
      "learning_rate": 0.04261775294550346,
      "loss": 0.1467,
      "num_input_tokens_seen": 26120640,
      "step": 30160
    },
    {
      "epoch": 14.222065063649222,
      "grad_norm": 0.003391087520867586,
      "learning_rate": 0.042576632576504354,
      "loss": 0.1374,
      "num_input_tokens_seen": 26125216,
      "step": 30165
    },
    {
      "epoch": 14.224422442244224,
      "grad_norm": 0.004971641581505537,
      "learning_rate": 0.0425355287735385,
      "loss": 0.2146,
      "num_input_tokens_seen": 26129376,
      "step": 30170
    },
    {
      "epoch": 14.226779820839226,
      "grad_norm": 0.006574847269803286,
      "learning_rate": 0.0424944415429446,
      "loss": 0.1633,
      "num_input_tokens_seen": 26133616,
      "step": 30175
    },
    {
      "epoch": 14.229137199434229,
      "grad_norm": 0.007146715652197599,
      "learning_rate": 0.04245337089105877,
      "loss": 0.1989,
      "num_input_tokens_seen": 26137392,
      "step": 30180
    },
    {
      "epoch": 14.231494578029231,
      "grad_norm": 0.005999289453029633,
      "learning_rate": 0.04241231682421467,
      "loss": 0.1603,
      "num_input_tokens_seen": 26141616,
      "step": 30185
    },
    {
      "epoch": 14.233851956624234,
      "grad_norm": 0.007023707497864962,
      "learning_rate": 0.04237127934874337,
      "loss": 0.1752,
      "num_input_tokens_seen": 26146592,
      "step": 30190
    },
    {
      "epoch": 14.236209335219236,
      "grad_norm": 0.0046423571184277534,
      "learning_rate": 0.042330258470973305,
      "loss": 0.1473,
      "num_input_tokens_seen": 26150960,
      "step": 30195
    },
    {
      "epoch": 14.238566713814238,
      "grad_norm": 0.005026846192777157,
      "learning_rate": 0.042289254197230515,
      "loss": 0.192,
      "num_input_tokens_seen": 26156672,
      "step": 30200
    },
    {
      "epoch": 14.238566713814238,
      "eval_loss": 0.22202041745185852,
      "eval_runtime": 21.8938,
      "eval_samples_per_second": 43.072,
      "eval_steps_per_second": 21.559,
      "num_input_tokens_seen": 26156672,
      "step": 30200
    },
    {
      "epoch": 14.24092409240924,
      "grad_norm": 0.007946528494358063,
      "learning_rate": 0.04224826653383823,
      "loss": 0.2082,
      "num_input_tokens_seen": 26161312,
      "step": 30205
    },
    {
      "epoch": 14.243281471004243,
      "grad_norm": 0.0075263893231749535,
      "learning_rate": 0.04220729548711735,
      "loss": 0.1782,
      "num_input_tokens_seen": 26165568,
      "step": 30210
    },
    {
      "epoch": 14.245638849599246,
      "grad_norm": 0.002777648624032736,
      "learning_rate": 0.04216634106338616,
      "loss": 0.1293,
      "num_input_tokens_seen": 26169584,
      "step": 30215
    },
    {
      "epoch": 14.247996228194248,
      "grad_norm": 0.004664845298975706,
      "learning_rate": 0.04212540326896025,
      "loss": 0.1864,
      "num_input_tokens_seen": 26174208,
      "step": 30220
    },
    {
      "epoch": 14.25035360678925,
      "grad_norm": 0.004095825832337141,
      "learning_rate": 0.0420844821101528,
      "loss": 0.1872,
      "num_input_tokens_seen": 26179088,
      "step": 30225
    },
    {
      "epoch": 14.252710985384253,
      "grad_norm": 0.00579098192974925,
      "learning_rate": 0.04204357759327441,
      "loss": 0.2115,
      "num_input_tokens_seen": 26182768,
      "step": 30230
    },
    {
      "epoch": 14.255068363979255,
      "grad_norm": 0.005154747981578112,
      "learning_rate": 0.042002689724632954,
      "loss": 0.1998,
      "num_input_tokens_seen": 26186880,
      "step": 30235
    },
    {
      "epoch": 14.257425742574258,
      "grad_norm": 0.005254998337477446,
      "learning_rate": 0.04196181851053398,
      "loss": 0.1485,
      "num_input_tokens_seen": 26191472,
      "step": 30240
    },
    {
      "epoch": 14.25978312116926,
      "grad_norm": 0.004900936037302017,
      "learning_rate": 0.041920963957280295,
      "loss": 0.1421,
      "num_input_tokens_seen": 26197504,
      "step": 30245
    },
    {
      "epoch": 14.262140499764262,
      "grad_norm": 0.006543243769556284,
      "learning_rate": 0.04188012607117212,
      "loss": 0.2116,
      "num_input_tokens_seen": 26201056,
      "step": 30250
    },
    {
      "epoch": 14.264497878359265,
      "grad_norm": 0.011217565275728703,
      "learning_rate": 0.04183930485850725,
      "loss": 0.2502,
      "num_input_tokens_seen": 26205952,
      "step": 30255
    },
    {
      "epoch": 14.266855256954267,
      "grad_norm": 0.004521931055933237,
      "learning_rate": 0.04179850032558078,
      "loss": 0.1897,
      "num_input_tokens_seen": 26209904,
      "step": 30260
    },
    {
      "epoch": 14.26921263554927,
      "grad_norm": 0.005287121050059795,
      "learning_rate": 0.041757712478685295,
      "loss": 0.1911,
      "num_input_tokens_seen": 26214720,
      "step": 30265
    },
    {
      "epoch": 14.271570014144272,
      "grad_norm": 0.005865566898137331,
      "learning_rate": 0.04171694132411085,
      "loss": 0.1773,
      "num_input_tokens_seen": 26219120,
      "step": 30270
    },
    {
      "epoch": 14.273927392739274,
      "grad_norm": 0.0062105427496135235,
      "learning_rate": 0.04167618686814479,
      "loss": 0.1474,
      "num_input_tokens_seen": 26223776,
      "step": 30275
    },
    {
      "epoch": 14.276284771334277,
      "grad_norm": 0.004904657602310181,
      "learning_rate": 0.041635449117072024,
      "loss": 0.1408,
      "num_input_tokens_seen": 26228448,
      "step": 30280
    },
    {
      "epoch": 14.27864214992928,
      "grad_norm": 0.010240467265248299,
      "learning_rate": 0.04159472807717477,
      "loss": 0.1152,
      "num_input_tokens_seen": 26233024,
      "step": 30285
    },
    {
      "epoch": 14.280999528524282,
      "grad_norm": 0.007853321731090546,
      "learning_rate": 0.041554023754732744,
      "loss": 0.1457,
      "num_input_tokens_seen": 26238400,
      "step": 30290
    },
    {
      "epoch": 14.283356907119284,
      "grad_norm": 0.0038395572919398546,
      "learning_rate": 0.04151333615602311,
      "loss": 0.1365,
      "num_input_tokens_seen": 26242304,
      "step": 30295
    },
    {
      "epoch": 14.285714285714286,
      "grad_norm": 0.005419599357992411,
      "learning_rate": 0.04147266528732034,
      "loss": 0.1455,
      "num_input_tokens_seen": 26246752,
      "step": 30300
    },
    {
      "epoch": 14.288071664309289,
      "grad_norm": 0.006951130460947752,
      "learning_rate": 0.0414320111548964,
      "loss": 0.1205,
      "num_input_tokens_seen": 26251440,
      "step": 30305
    },
    {
      "epoch": 14.290429042904291,
      "grad_norm": 0.008912274613976479,
      "learning_rate": 0.04139137376502076,
      "loss": 0.1555,
      "num_input_tokens_seen": 26255776,
      "step": 30310
    },
    {
      "epoch": 14.292786421499294,
      "grad_norm": 0.0076253036968410015,
      "learning_rate": 0.04135075312396014,
      "loss": 0.2262,
      "num_input_tokens_seen": 26259888,
      "step": 30315
    },
    {
      "epoch": 14.295143800094294,
      "grad_norm": 0.008812078274786472,
      "learning_rate": 0.04131014923797875,
      "loss": 0.1804,
      "num_input_tokens_seen": 26263712,
      "step": 30320
    },
    {
      "epoch": 14.297501178689297,
      "grad_norm": 0.007465405855327845,
      "learning_rate": 0.04126956211333819,
      "loss": 0.1957,
      "num_input_tokens_seen": 26267760,
      "step": 30325
    },
    {
      "epoch": 14.299858557284299,
      "grad_norm": 0.011151977814733982,
      "learning_rate": 0.041228991756297545,
      "loss": 0.2282,
      "num_input_tokens_seen": 26272304,
      "step": 30330
    },
    {
      "epoch": 14.302215935879302,
      "grad_norm": 0.0072767664678394794,
      "learning_rate": 0.04118843817311332,
      "loss": 0.192,
      "num_input_tokens_seen": 26276784,
      "step": 30335
    },
    {
      "epoch": 14.304573314474304,
      "grad_norm": 0.006412986665964127,
      "learning_rate": 0.0411479013700393,
      "loss": 0.1613,
      "num_input_tokens_seen": 26281104,
      "step": 30340
    },
    {
      "epoch": 14.306930693069306,
      "grad_norm": 0.0033187298104166985,
      "learning_rate": 0.0411073813533268,
      "loss": 0.1639,
      "num_input_tokens_seen": 26285536,
      "step": 30345
    },
    {
      "epoch": 14.309288071664309,
      "grad_norm": 0.00820187944918871,
      "learning_rate": 0.04106687812922456,
      "loss": 0.1562,
      "num_input_tokens_seen": 26289616,
      "step": 30350
    },
    {
      "epoch": 14.311645450259311,
      "grad_norm": 0.0035016450565308332,
      "learning_rate": 0.041026391703978635,
      "loss": 0.1093,
      "num_input_tokens_seen": 26294000,
      "step": 30355
    },
    {
      "epoch": 14.314002828854314,
      "grad_norm": 0.006643793545663357,
      "learning_rate": 0.04098592208383259,
      "loss": 0.1349,
      "num_input_tokens_seen": 26298496,
      "step": 30360
    },
    {
      "epoch": 14.316360207449316,
      "grad_norm": 0.007168669253587723,
      "learning_rate": 0.040945469275027256,
      "loss": 0.1422,
      "num_input_tokens_seen": 26302880,
      "step": 30365
    },
    {
      "epoch": 14.318717586044318,
      "grad_norm": 0.006721879355609417,
      "learning_rate": 0.04090503328380104,
      "loss": 0.1911,
      "num_input_tokens_seen": 26307008,
      "step": 30370
    },
    {
      "epoch": 14.32107496463932,
      "grad_norm": 0.00941894855350256,
      "learning_rate": 0.04086461411638971,
      "loss": 0.1879,
      "num_input_tokens_seen": 26310720,
      "step": 30375
    },
    {
      "epoch": 14.323432343234323,
      "grad_norm": 0.01057712733745575,
      "learning_rate": 0.04082421177902631,
      "loss": 0.2031,
      "num_input_tokens_seen": 26314832,
      "step": 30380
    },
    {
      "epoch": 14.325789721829326,
      "grad_norm": 0.003999708686023951,
      "learning_rate": 0.04078382627794149,
      "loss": 0.1286,
      "num_input_tokens_seen": 26319152,
      "step": 30385
    },
    {
      "epoch": 14.328147100424328,
      "grad_norm": 0.006086652632802725,
      "learning_rate": 0.04074345761936316,
      "loss": 0.1931,
      "num_input_tokens_seen": 26322992,
      "step": 30390
    },
    {
      "epoch": 14.33050447901933,
      "grad_norm": 0.007569155655801296,
      "learning_rate": 0.04070310580951663,
      "loss": 0.167,
      "num_input_tokens_seen": 26326352,
      "step": 30395
    },
    {
      "epoch": 14.332861857614333,
      "grad_norm": 0.007616538088768721,
      "learning_rate": 0.040662770854624726,
      "loss": 0.1962,
      "num_input_tokens_seen": 26330592,
      "step": 30400
    },
    {
      "epoch": 14.332861857614333,
      "eval_loss": 0.22741538286209106,
      "eval_runtime": 21.9146,
      "eval_samples_per_second": 43.031,
      "eval_steps_per_second": 21.538,
      "num_input_tokens_seen": 26330592,
      "step": 30400
    },
    {
      "epoch": 14.335219236209335,
      "grad_norm": 0.008938301354646683,
      "learning_rate": 0.040622452760907535,
      "loss": 0.1621,
      "num_input_tokens_seen": 26334704,
      "step": 30405
    },
    {
      "epoch": 14.337576614804338,
      "grad_norm": 0.0030964931938797235,
      "learning_rate": 0.04058215153458265,
      "loss": 0.1396,
      "num_input_tokens_seen": 26340400,
      "step": 30410
    },
    {
      "epoch": 14.33993399339934,
      "grad_norm": 0.010010600090026855,
      "learning_rate": 0.04054186718186507,
      "loss": 0.1972,
      "num_input_tokens_seen": 26344192,
      "step": 30415
    },
    {
      "epoch": 14.342291371994342,
      "grad_norm": 0.003785910317674279,
      "learning_rate": 0.04050159970896708,
      "loss": 0.1787,
      "num_input_tokens_seen": 26348128,
      "step": 30420
    },
    {
      "epoch": 14.344648750589345,
      "grad_norm": 0.00928839948028326,
      "learning_rate": 0.04046134912209843,
      "loss": 0.2316,
      "num_input_tokens_seen": 26352320,
      "step": 30425
    },
    {
      "epoch": 14.347006129184347,
      "grad_norm": 0.020709043368697166,
      "learning_rate": 0.040421115427466354,
      "loss": 0.1709,
      "num_input_tokens_seen": 26356384,
      "step": 30430
    },
    {
      "epoch": 14.34936350777935,
      "grad_norm": 0.005872691050171852,
      "learning_rate": 0.04038089863127529,
      "loss": 0.1923,
      "num_input_tokens_seen": 26361152,
      "step": 30435
    },
    {
      "epoch": 14.351720886374352,
      "grad_norm": 0.0058170282281935215,
      "learning_rate": 0.04034069873972727,
      "loss": 0.1519,
      "num_input_tokens_seen": 26366256,
      "step": 30440
    },
    {
      "epoch": 14.354078264969354,
      "grad_norm": 0.006014326121658087,
      "learning_rate": 0.040300515759021514,
      "loss": 0.2227,
      "num_input_tokens_seen": 26370368,
      "step": 30445
    },
    {
      "epoch": 14.356435643564357,
      "grad_norm": 0.006300934124737978,
      "learning_rate": 0.04026034969535478,
      "loss": 0.2189,
      "num_input_tokens_seen": 26374864,
      "step": 30450
    },
    {
      "epoch": 14.35879302215936,
      "grad_norm": 0.009482618421316147,
      "learning_rate": 0.040220200554921266,
      "loss": 0.1866,
      "num_input_tokens_seen": 26378960,
      "step": 30455
    },
    {
      "epoch": 14.361150400754362,
      "grad_norm": 0.0073554557748138905,
      "learning_rate": 0.0401800683439124,
      "loss": 0.1783,
      "num_input_tokens_seen": 26383616,
      "step": 30460
    },
    {
      "epoch": 14.363507779349364,
      "grad_norm": 0.01084878109395504,
      "learning_rate": 0.04013995306851704,
      "loss": 0.2351,
      "num_input_tokens_seen": 26387696,
      "step": 30465
    },
    {
      "epoch": 14.365865157944366,
      "grad_norm": 0.006201889365911484,
      "learning_rate": 0.040099854734921545,
      "loss": 0.1994,
      "num_input_tokens_seen": 26391312,
      "step": 30470
    },
    {
      "epoch": 14.368222536539369,
      "grad_norm": 0.005605104845017195,
      "learning_rate": 0.0400597733493095,
      "loss": 0.1883,
      "num_input_tokens_seen": 26395440,
      "step": 30475
    },
    {
      "epoch": 14.370579915134371,
      "grad_norm": 0.005972946994006634,
      "learning_rate": 0.04001970891786203,
      "loss": 0.2143,
      "num_input_tokens_seen": 26399616,
      "step": 30480
    },
    {
      "epoch": 14.372937293729374,
      "grad_norm": 0.008432709611952305,
      "learning_rate": 0.03997966144675752,
      "loss": 0.2105,
      "num_input_tokens_seen": 26404800,
      "step": 30485
    },
    {
      "epoch": 14.375294672324376,
      "grad_norm": 0.009495237842202187,
      "learning_rate": 0.039939630942171796,
      "loss": 0.2216,
      "num_input_tokens_seen": 26408848,
      "step": 30490
    },
    {
      "epoch": 14.377652050919378,
      "grad_norm": 0.004275960847735405,
      "learning_rate": 0.03989961741027815,
      "loss": 0.1534,
      "num_input_tokens_seen": 26413328,
      "step": 30495
    },
    {
      "epoch": 14.38000942951438,
      "grad_norm": 0.004292964935302734,
      "learning_rate": 0.03985962085724704,
      "loss": 0.1506,
      "num_input_tokens_seen": 26417072,
      "step": 30500
    },
    {
      "epoch": 14.382366808109383,
      "grad_norm": 0.0053281886503100395,
      "learning_rate": 0.03981964128924656,
      "loss": 0.1375,
      "num_input_tokens_seen": 26420544,
      "step": 30505
    },
    {
      "epoch": 14.384724186704386,
      "grad_norm": 0.004367724061012268,
      "learning_rate": 0.03977967871244197,
      "loss": 0.2169,
      "num_input_tokens_seen": 26424464,
      "step": 30510
    },
    {
      "epoch": 14.387081565299386,
      "grad_norm": 0.007210994604974985,
      "learning_rate": 0.03973973313299602,
      "loss": 0.183,
      "num_input_tokens_seen": 26428464,
      "step": 30515
    },
    {
      "epoch": 14.389438943894389,
      "grad_norm": 0.0048298304900527,
      "learning_rate": 0.0396998045570689,
      "loss": 0.2452,
      "num_input_tokens_seen": 26432704,
      "step": 30520
    },
    {
      "epoch": 14.391796322489391,
      "grad_norm": 0.008039802312850952,
      "learning_rate": 0.03965989299081798,
      "loss": 0.1431,
      "num_input_tokens_seen": 26437472,
      "step": 30525
    },
    {
      "epoch": 14.394153701084393,
      "grad_norm": 0.007098461501300335,
      "learning_rate": 0.039619998440398235,
      "loss": 0.1644,
      "num_input_tokens_seen": 26441680,
      "step": 30530
    },
    {
      "epoch": 14.396511079679396,
      "grad_norm": 0.009242547675967216,
      "learning_rate": 0.03958012091196184,
      "loss": 0.1678,
      "num_input_tokens_seen": 26446608,
      "step": 30535
    },
    {
      "epoch": 14.398868458274398,
      "grad_norm": 0.005284856539219618,
      "learning_rate": 0.039540260411658396,
      "loss": 0.1304,
      "num_input_tokens_seen": 26451152,
      "step": 30540
    },
    {
      "epoch": 14.4012258368694,
      "grad_norm": 0.008283290080726147,
      "learning_rate": 0.03950041694563496,
      "loss": 0.21,
      "num_input_tokens_seen": 26455328,
      "step": 30545
    },
    {
      "epoch": 14.403583215464403,
      "grad_norm": 0.004776647314429283,
      "learning_rate": 0.0394605905200358,
      "loss": 0.1726,
      "num_input_tokens_seen": 26459616,
      "step": 30550
    },
    {
      "epoch": 14.405940594059405,
      "grad_norm": 0.005495809018611908,
      "learning_rate": 0.03942078114100272,
      "loss": 0.1369,
      "num_input_tokens_seen": 26464336,
      "step": 30555
    },
    {
      "epoch": 14.408297972654408,
      "grad_norm": 0.008281443268060684,
      "learning_rate": 0.03938098881467485,
      "loss": 0.2248,
      "num_input_tokens_seen": 26468160,
      "step": 30560
    },
    {
      "epoch": 14.41065535124941,
      "grad_norm": 0.007201177999377251,
      "learning_rate": 0.039341213547188586,
      "loss": 0.2077,
      "num_input_tokens_seen": 26471936,
      "step": 30565
    },
    {
      "epoch": 14.413012729844413,
      "grad_norm": 0.009223357774317265,
      "learning_rate": 0.03930145534467782,
      "loss": 0.1638,
      "num_input_tokens_seen": 26476480,
      "step": 30570
    },
    {
      "epoch": 14.415370108439415,
      "grad_norm": 0.003936063963919878,
      "learning_rate": 0.0392617142132738,
      "loss": 0.1289,
      "num_input_tokens_seen": 26481536,
      "step": 30575
    },
    {
      "epoch": 14.417727487034417,
      "grad_norm": 0.007489695213735104,
      "learning_rate": 0.03922199015910504,
      "loss": 0.1849,
      "num_input_tokens_seen": 26485952,
      "step": 30580
    },
    {
      "epoch": 14.42008486562942,
      "grad_norm": 0.008202260360121727,
      "learning_rate": 0.039182283188297556,
      "loss": 0.2161,
      "num_input_tokens_seen": 26490160,
      "step": 30585
    },
    {
      "epoch": 14.422442244224422,
      "grad_norm": 0.006327849812805653,
      "learning_rate": 0.039142593306974595,
      "loss": 0.1878,
      "num_input_tokens_seen": 26494624,
      "step": 30590
    },
    {
      "epoch": 14.424799622819425,
      "grad_norm": 0.004990175366401672,
      "learning_rate": 0.039102920521256856,
      "loss": 0.179,
      "num_input_tokens_seen": 26498400,
      "step": 30595
    },
    {
      "epoch": 14.427157001414427,
      "grad_norm": 0.006220838986337185,
      "learning_rate": 0.03906326483726243,
      "loss": 0.1747,
      "num_input_tokens_seen": 26502800,
      "step": 30600
    },
    {
      "epoch": 14.427157001414427,
      "eval_loss": 0.21772323548793793,
      "eval_runtime": 21.9051,
      "eval_samples_per_second": 43.049,
      "eval_steps_per_second": 21.548,
      "num_input_tokens_seen": 26502800,
      "step": 30600
    },
    {
      "epoch": 14.42951438000943,
      "grad_norm": 0.008114648051559925,
      "learning_rate": 0.039023626261106704,
      "loss": 0.1938,
      "num_input_tokens_seen": 26507264,
      "step": 30605
    },
    {
      "epoch": 14.431871758604432,
      "grad_norm": 0.005522445775568485,
      "learning_rate": 0.03898400479890237,
      "loss": 0.2231,
      "num_input_tokens_seen": 26511008,
      "step": 30610
    },
    {
      "epoch": 14.434229137199434,
      "grad_norm": 0.0071020303294062614,
      "learning_rate": 0.038944400456759655,
      "loss": 0.158,
      "num_input_tokens_seen": 26515712,
      "step": 30615
    },
    {
      "epoch": 14.436586515794437,
      "grad_norm": 0.007140698377043009,
      "learning_rate": 0.038904813240785964,
      "loss": 0.2201,
      "num_input_tokens_seen": 26519616,
      "step": 30620
    },
    {
      "epoch": 14.438943894389439,
      "grad_norm": 0.007632260676473379,
      "learning_rate": 0.03886524315708621,
      "loss": 0.1753,
      "num_input_tokens_seen": 26523728,
      "step": 30625
    },
    {
      "epoch": 14.441301272984441,
      "grad_norm": 0.005710085388273001,
      "learning_rate": 0.03882569021176255,
      "loss": 0.1729,
      "num_input_tokens_seen": 26527664,
      "step": 30630
    },
    {
      "epoch": 14.443658651579444,
      "grad_norm": 0.004584199283272028,
      "learning_rate": 0.038786154410914535,
      "loss": 0.1316,
      "num_input_tokens_seen": 26532528,
      "step": 30635
    },
    {
      "epoch": 14.446016030174446,
      "grad_norm": 0.003939956892281771,
      "learning_rate": 0.03874663576063917,
      "loss": 0.1594,
      "num_input_tokens_seen": 26536304,
      "step": 30640
    },
    {
      "epoch": 14.448373408769449,
      "grad_norm": 0.00636113528162241,
      "learning_rate": 0.038707134267030624,
      "loss": 0.1187,
      "num_input_tokens_seen": 26540192,
      "step": 30645
    },
    {
      "epoch": 14.450730787364451,
      "grad_norm": 0.004779467824846506,
      "learning_rate": 0.038667649936180555,
      "loss": 0.1771,
      "num_input_tokens_seen": 26544720,
      "step": 30650
    },
    {
      "epoch": 14.453088165959453,
      "grad_norm": 0.004768576472997665,
      "learning_rate": 0.038628182774178,
      "loss": 0.1833,
      "num_input_tokens_seen": 26549248,
      "step": 30655
    },
    {
      "epoch": 14.455445544554456,
      "grad_norm": 0.005710358265787363,
      "learning_rate": 0.038588732787109226,
      "loss": 0.1693,
      "num_input_tokens_seen": 26553152,
      "step": 30660
    },
    {
      "epoch": 14.457802923149458,
      "grad_norm": 0.004045804496854544,
      "learning_rate": 0.03854929998105795,
      "loss": 0.2033,
      "num_input_tokens_seen": 26557520,
      "step": 30665
    },
    {
      "epoch": 14.46016030174446,
      "grad_norm": 0.005419635213911533,
      "learning_rate": 0.03850988436210518,
      "loss": 0.238,
      "num_input_tokens_seen": 26561616,
      "step": 30670
    },
    {
      "epoch": 14.462517680339463,
      "grad_norm": 0.004515789449214935,
      "learning_rate": 0.03847048593632933,
      "loss": 0.1574,
      "num_input_tokens_seen": 26566496,
      "step": 30675
    },
    {
      "epoch": 14.464875058934465,
      "grad_norm": 0.00880457554012537,
      "learning_rate": 0.038431104709806096,
      "loss": 0.1744,
      "num_input_tokens_seen": 26570784,
      "step": 30680
    },
    {
      "epoch": 14.467232437529468,
      "grad_norm": 0.0048517887480556965,
      "learning_rate": 0.0383917406886086,
      "loss": 0.2031,
      "num_input_tokens_seen": 26574800,
      "step": 30685
    },
    {
      "epoch": 14.46958981612447,
      "grad_norm": 0.005601128097623587,
      "learning_rate": 0.03835239387880722,
      "loss": 0.1905,
      "num_input_tokens_seen": 26579152,
      "step": 30690
    },
    {
      "epoch": 14.471947194719473,
      "grad_norm": 0.006484190467745066,
      "learning_rate": 0.03831306428646979,
      "loss": 0.1749,
      "num_input_tokens_seen": 26583664,
      "step": 30695
    },
    {
      "epoch": 14.474304573314475,
      "grad_norm": 0.004080051556229591,
      "learning_rate": 0.03827375191766135,
      "loss": 0.1428,
      "num_input_tokens_seen": 26587744,
      "step": 30700
    },
    {
      "epoch": 14.476661951909477,
      "grad_norm": 0.0036432817578315735,
      "learning_rate": 0.03823445677844446,
      "loss": 0.1077,
      "num_input_tokens_seen": 26592720,
      "step": 30705
    },
    {
      "epoch": 14.47901933050448,
      "grad_norm": 0.0063684056513011456,
      "learning_rate": 0.03819517887487881,
      "loss": 0.1862,
      "num_input_tokens_seen": 26596736,
      "step": 30710
    },
    {
      "epoch": 14.481376709099482,
      "grad_norm": 0.00417819619178772,
      "learning_rate": 0.03815591821302161,
      "loss": 0.1615,
      "num_input_tokens_seen": 26600592,
      "step": 30715
    },
    {
      "epoch": 14.483734087694483,
      "grad_norm": 0.006939438171684742,
      "learning_rate": 0.03811667479892739,
      "loss": 0.2114,
      "num_input_tokens_seen": 26604896,
      "step": 30720
    },
    {
      "epoch": 14.486091466289485,
      "grad_norm": 0.0059495954774320126,
      "learning_rate": 0.03807744863864788,
      "loss": 0.155,
      "num_input_tokens_seen": 26608512,
      "step": 30725
    },
    {
      "epoch": 14.488448844884488,
      "grad_norm": 0.009088069200515747,
      "learning_rate": 0.03803823973823229,
      "loss": 0.182,
      "num_input_tokens_seen": 26612624,
      "step": 30730
    },
    {
      "epoch": 14.49080622347949,
      "grad_norm": 0.005793694406747818,
      "learning_rate": 0.03799904810372719,
      "loss": 0.2084,
      "num_input_tokens_seen": 26617072,
      "step": 30735
    },
    {
      "epoch": 14.493163602074493,
      "grad_norm": 0.00692905206233263,
      "learning_rate": 0.03795987374117632,
      "loss": 0.1498,
      "num_input_tokens_seen": 26621312,
      "step": 30740
    },
    {
      "epoch": 14.495520980669495,
      "grad_norm": 0.0051119206473231316,
      "learning_rate": 0.03792071665662093,
      "loss": 0.1586,
      "num_input_tokens_seen": 26625456,
      "step": 30745
    },
    {
      "epoch": 14.497878359264497,
      "grad_norm": 0.009916132315993309,
      "learning_rate": 0.03788157685609952,
      "loss": 0.1904,
      "num_input_tokens_seen": 26629456,
      "step": 30750
    },
    {
      "epoch": 14.5002357378595,
      "grad_norm": 0.003021150827407837,
      "learning_rate": 0.037842454345647876,
      "loss": 0.1779,
      "num_input_tokens_seen": 26634224,
      "step": 30755
    },
    {
      "epoch": 14.502593116454502,
      "grad_norm": 0.006743459030985832,
      "learning_rate": 0.03780334913129929,
      "loss": 0.1963,
      "num_input_tokens_seen": 26638624,
      "step": 30760
    },
    {
      "epoch": 14.504950495049505,
      "grad_norm": 0.001900395262055099,
      "learning_rate": 0.037764261219084175,
      "loss": 0.1511,
      "num_input_tokens_seen": 26642368,
      "step": 30765
    },
    {
      "epoch": 14.507307873644507,
      "grad_norm": 0.010378874838352203,
      "learning_rate": 0.037725190615030414,
      "loss": 0.2182,
      "num_input_tokens_seen": 26646336,
      "step": 30770
    },
    {
      "epoch": 14.50966525223951,
      "grad_norm": 0.005705143325030804,
      "learning_rate": 0.037686137325163224,
      "loss": 0.156,
      "num_input_tokens_seen": 26650016,
      "step": 30775
    },
    {
      "epoch": 14.512022630834512,
      "grad_norm": 0.005963718052953482,
      "learning_rate": 0.037647101355505065,
      "loss": 0.2239,
      "num_input_tokens_seen": 26653664,
      "step": 30780
    },
    {
      "epoch": 14.514380009429514,
      "grad_norm": 0.009307513013482094,
      "learning_rate": 0.03760808271207581,
      "loss": 0.2198,
      "num_input_tokens_seen": 26658112,
      "step": 30785
    },
    {
      "epoch": 14.516737388024517,
      "grad_norm": 0.005175680387765169,
      "learning_rate": 0.03756908140089258,
      "loss": 0.1737,
      "num_input_tokens_seen": 26663104,
      "step": 30790
    },
    {
      "epoch": 14.519094766619519,
      "grad_norm": 0.0032260785810649395,
      "learning_rate": 0.03753009742796989,
      "loss": 0.1403,
      "num_input_tokens_seen": 26667280,
      "step": 30795
    },
    {
      "epoch": 14.521452145214521,
      "grad_norm": 0.004864353686571121,
      "learning_rate": 0.037491130799319615,
      "loss": 0.2064,
      "num_input_tokens_seen": 26671584,
      "step": 30800
    },
    {
      "epoch": 14.521452145214521,
      "eval_loss": 0.21666614711284637,
      "eval_runtime": 21.928,
      "eval_samples_per_second": 43.004,
      "eval_steps_per_second": 21.525,
      "num_input_tokens_seen": 26671584,
      "step": 30800
    },
    {
      "epoch": 14.523809523809524,
      "grad_norm": 0.004547072574496269,
      "learning_rate": 0.03745218152095079,
      "loss": 0.1435,
      "num_input_tokens_seen": 26675920,
      "step": 30805
    },
    {
      "epoch": 14.526166902404526,
      "grad_norm": 0.0031764600425958633,
      "learning_rate": 0.037413249598869935,
      "loss": 0.1119,
      "num_input_tokens_seen": 26680016,
      "step": 30810
    },
    {
      "epoch": 14.528524280999529,
      "grad_norm": 0.0048849573358893394,
      "learning_rate": 0.037374335039080886,
      "loss": 0.1895,
      "num_input_tokens_seen": 26684416,
      "step": 30815
    },
    {
      "epoch": 14.530881659594531,
      "grad_norm": 0.00830648560076952,
      "learning_rate": 0.037335437847584724,
      "loss": 0.1839,
      "num_input_tokens_seen": 26688736,
      "step": 30820
    },
    {
      "epoch": 14.533239038189533,
      "grad_norm": 0.006046365015208721,
      "learning_rate": 0.03729655803037983,
      "loss": 0.1501,
      "num_input_tokens_seen": 26692848,
      "step": 30825
    },
    {
      "epoch": 14.535596416784536,
      "grad_norm": 0.006857164669781923,
      "learning_rate": 0.03725769559346207,
      "loss": 0.1747,
      "num_input_tokens_seen": 26697888,
      "step": 30830
    },
    {
      "epoch": 14.537953795379538,
      "grad_norm": 0.005469634663313627,
      "learning_rate": 0.03721885054282439,
      "loss": 0.1818,
      "num_input_tokens_seen": 26701648,
      "step": 30835
    },
    {
      "epoch": 14.54031117397454,
      "grad_norm": 0.01010989025235176,
      "learning_rate": 0.03718002288445731,
      "loss": 0.1389,
      "num_input_tokens_seen": 26706240,
      "step": 30840
    },
    {
      "epoch": 14.542668552569543,
      "grad_norm": 0.003932561259716749,
      "learning_rate": 0.03714121262434844,
      "loss": 0.1906,
      "num_input_tokens_seen": 26710656,
      "step": 30845
    },
    {
      "epoch": 14.545025931164545,
      "grad_norm": 0.0035110118333250284,
      "learning_rate": 0.037102419768482844,
      "loss": 0.148,
      "num_input_tokens_seen": 26715264,
      "step": 30850
    },
    {
      "epoch": 14.547383309759548,
      "grad_norm": 0.0077673569321632385,
      "learning_rate": 0.03706364432284293,
      "loss": 0.2135,
      "num_input_tokens_seen": 26720176,
      "step": 30855
    },
    {
      "epoch": 14.54974068835455,
      "grad_norm": 0.003392778569832444,
      "learning_rate": 0.03702488629340828,
      "loss": 0.1964,
      "num_input_tokens_seen": 26724448,
      "step": 30860
    },
    {
      "epoch": 14.552098066949553,
      "grad_norm": 0.007430478930473328,
      "learning_rate": 0.036986145686155915,
      "loss": 0.1767,
      "num_input_tokens_seen": 26728944,
      "step": 30865
    },
    {
      "epoch": 14.554455445544555,
      "grad_norm": 0.006181610282510519,
      "learning_rate": 0.036947422507060075,
      "loss": 0.1622,
      "num_input_tokens_seen": 26732432,
      "step": 30870
    },
    {
      "epoch": 14.556812824139557,
      "grad_norm": 0.007989470846951008,
      "learning_rate": 0.0369087167620924,
      "loss": 0.1908,
      "num_input_tokens_seen": 26736368,
      "step": 30875
    },
    {
      "epoch": 14.55917020273456,
      "grad_norm": 0.0057749636471271515,
      "learning_rate": 0.03687002845722183,
      "loss": 0.1814,
      "num_input_tokens_seen": 26740752,
      "step": 30880
    },
    {
      "epoch": 14.561527581329562,
      "grad_norm": 0.005783790722489357,
      "learning_rate": 0.03683135759841451,
      "loss": 0.1758,
      "num_input_tokens_seen": 26744560,
      "step": 30885
    },
    {
      "epoch": 14.563884959924565,
      "grad_norm": 0.007686834316700697,
      "learning_rate": 0.03679270419163406,
      "loss": 0.2042,
      "num_input_tokens_seen": 26748480,
      "step": 30890
    },
    {
      "epoch": 14.566242338519567,
      "grad_norm": 0.008002380840480328,
      "learning_rate": 0.03675406824284127,
      "loss": 0.1778,
      "num_input_tokens_seen": 26752448,
      "step": 30895
    },
    {
      "epoch": 14.56859971711457,
      "grad_norm": 0.0067858281545341015,
      "learning_rate": 0.03671544975799425,
      "loss": 0.1815,
      "num_input_tokens_seen": 26756816,
      "step": 30900
    },
    {
      "epoch": 14.570957095709572,
      "grad_norm": 0.005919996183365583,
      "learning_rate": 0.03667684874304854,
      "loss": 0.1594,
      "num_input_tokens_seen": 26760992,
      "step": 30905
    },
    {
      "epoch": 14.573314474304574,
      "grad_norm": 0.007145118899643421,
      "learning_rate": 0.03663826520395683,
      "loss": 0.1607,
      "num_input_tokens_seen": 26765184,
      "step": 30910
    },
    {
      "epoch": 14.575671852899575,
      "grad_norm": 0.006249037571251392,
      "learning_rate": 0.03659969914666922,
      "loss": 0.1857,
      "num_input_tokens_seen": 26769472,
      "step": 30915
    },
    {
      "epoch": 14.578029231494579,
      "grad_norm": 0.00717057753354311,
      "learning_rate": 0.036561150577133106,
      "loss": 0.1554,
      "num_input_tokens_seen": 26774112,
      "step": 30920
    },
    {
      "epoch": 14.58038661008958,
      "grad_norm": 0.005977169144898653,
      "learning_rate": 0.036522619501293103,
      "loss": 0.1653,
      "num_input_tokens_seen": 26778480,
      "step": 30925
    },
    {
      "epoch": 14.582743988684582,
      "grad_norm": 0.010449199937283993,
      "learning_rate": 0.03648410592509122,
      "loss": 0.1706,
      "num_input_tokens_seen": 26782720,
      "step": 30930
    },
    {
      "epoch": 14.585101367279584,
      "grad_norm": 0.007821097038686275,
      "learning_rate": 0.03644560985446676,
      "loss": 0.1978,
      "num_input_tokens_seen": 26786624,
      "step": 30935
    },
    {
      "epoch": 14.587458745874587,
      "grad_norm": 0.006361290346831083,
      "learning_rate": 0.036407131295356256,
      "loss": 0.1564,
      "num_input_tokens_seen": 26790512,
      "step": 30940
    },
    {
      "epoch": 14.58981612446959,
      "grad_norm": 0.007763554807752371,
      "learning_rate": 0.03636867025369362,
      "loss": 0.1472,
      "num_input_tokens_seen": 26795648,
      "step": 30945
    },
    {
      "epoch": 14.592173503064592,
      "grad_norm": 0.006469822023063898,
      "learning_rate": 0.03633022673540999,
      "loss": 0.1516,
      "num_input_tokens_seen": 26800896,
      "step": 30950
    },
    {
      "epoch": 14.594530881659594,
      "grad_norm": 0.005037639755755663,
      "learning_rate": 0.03629180074643385,
      "loss": 0.1573,
      "num_input_tokens_seen": 26805776,
      "step": 30955
    },
    {
      "epoch": 14.596888260254596,
      "grad_norm": 0.006903060246258974,
      "learning_rate": 0.03625339229269102,
      "loss": 0.1698,
      "num_input_tokens_seen": 26810432,
      "step": 30960
    },
    {
      "epoch": 14.599245638849599,
      "grad_norm": 0.009457863867282867,
      "learning_rate": 0.036215001380104535,
      "loss": 0.2112,
      "num_input_tokens_seen": 26814848,
      "step": 30965
    },
    {
      "epoch": 14.601603017444601,
      "grad_norm": 0.013125145807862282,
      "learning_rate": 0.03617662801459471,
      "loss": 0.2344,
      "num_input_tokens_seen": 26820000,
      "step": 30970
    },
    {
      "epoch": 14.603960396039604,
      "grad_norm": 0.0061284014955163,
      "learning_rate": 0.036138272202079276,
      "loss": 0.1606,
      "num_input_tokens_seen": 26824288,
      "step": 30975
    },
    {
      "epoch": 14.606317774634606,
      "grad_norm": 0.005944352131336927,
      "learning_rate": 0.036099933948473106,
      "loss": 0.1725,
      "num_input_tokens_seen": 26829280,
      "step": 30980
    },
    {
      "epoch": 14.608675153229608,
      "grad_norm": 0.0058021689765155315,
      "learning_rate": 0.03606161325968851,
      "loss": 0.1529,
      "num_input_tokens_seen": 26833456,
      "step": 30985
    },
    {
      "epoch": 14.61103253182461,
      "grad_norm": 0.004104289226233959,
      "learning_rate": 0.03602331014163496,
      "loss": 0.1988,
      "num_input_tokens_seen": 26837472,
      "step": 30990
    },
    {
      "epoch": 14.613389910419613,
      "grad_norm": 0.008772805333137512,
      "learning_rate": 0.035985024600219295,
      "loss": 0.1689,
      "num_input_tokens_seen": 26841904,
      "step": 30995
    },
    {
      "epoch": 14.615747289014616,
      "grad_norm": 0.006049638614058495,
      "learning_rate": 0.03594675664134569,
      "loss": 0.1868,
      "num_input_tokens_seen": 26845568,
      "step": 31000
    },
    {
      "epoch": 14.615747289014616,
      "eval_loss": 0.22147604823112488,
      "eval_runtime": 21.8964,
      "eval_samples_per_second": 43.067,
      "eval_steps_per_second": 21.556,
      "num_input_tokens_seen": 26845568,
      "step": 31000
    },
    {
      "epoch": 14.618104667609618,
      "grad_norm": 0.005350709427148104,
      "learning_rate": 0.03590850627091545,
      "loss": 0.1604,
      "num_input_tokens_seen": 26849472,
      "step": 31005
    },
    {
      "epoch": 14.62046204620462,
      "grad_norm": 0.006496004294604063,
      "learning_rate": 0.03587027349482731,
      "loss": 0.1983,
      "num_input_tokens_seen": 26854160,
      "step": 31010
    },
    {
      "epoch": 14.622819424799623,
      "grad_norm": 0.010208907537162304,
      "learning_rate": 0.035832058318977275,
      "loss": 0.2148,
      "num_input_tokens_seen": 26859184,
      "step": 31015
    },
    {
      "epoch": 14.625176803394625,
      "grad_norm": 0.009728824719786644,
      "learning_rate": 0.03579386074925853,
      "loss": 0.1955,
      "num_input_tokens_seen": 26862912,
      "step": 31020
    },
    {
      "epoch": 14.627534181989628,
      "grad_norm": 0.006865161005407572,
      "learning_rate": 0.035755680791561696,
      "loss": 0.2106,
      "num_input_tokens_seen": 26868416,
      "step": 31025
    },
    {
      "epoch": 14.62989156058463,
      "grad_norm": 0.006134893745183945,
      "learning_rate": 0.03571751845177454,
      "loss": 0.1957,
      "num_input_tokens_seen": 26871872,
      "step": 31030
    },
    {
      "epoch": 14.632248939179632,
      "grad_norm": 0.007742851972579956,
      "learning_rate": 0.03567937373578225,
      "loss": 0.2745,
      "num_input_tokens_seen": 26875488,
      "step": 31035
    },
    {
      "epoch": 14.634606317774635,
      "grad_norm": 0.0035222978331148624,
      "learning_rate": 0.03564124664946711,
      "loss": 0.2,
      "num_input_tokens_seen": 26879584,
      "step": 31040
    },
    {
      "epoch": 14.636963696369637,
      "grad_norm": 0.0058174519799649715,
      "learning_rate": 0.035603137198708924,
      "loss": 0.1808,
      "num_input_tokens_seen": 26883344,
      "step": 31045
    },
    {
      "epoch": 14.63932107496464,
      "grad_norm": 0.004803921561688185,
      "learning_rate": 0.035565045389384514,
      "loss": 0.2108,
      "num_input_tokens_seen": 26887840,
      "step": 31050
    },
    {
      "epoch": 14.641678453559642,
      "grad_norm": 0.005257738754153252,
      "learning_rate": 0.03552697122736823,
      "loss": 0.1544,
      "num_input_tokens_seen": 26891648,
      "step": 31055
    },
    {
      "epoch": 14.644035832154644,
      "grad_norm": 0.0031367656774818897,
      "learning_rate": 0.03548891471853153,
      "loss": 0.1669,
      "num_input_tokens_seen": 26896416,
      "step": 31060
    },
    {
      "epoch": 14.646393210749647,
      "grad_norm": 0.005687291268259287,
      "learning_rate": 0.03545087586874322,
      "loss": 0.1147,
      "num_input_tokens_seen": 26901696,
      "step": 31065
    },
    {
      "epoch": 14.64875058934465,
      "grad_norm": 0.005886594764888287,
      "learning_rate": 0.03541285468386935,
      "loss": 0.1493,
      "num_input_tokens_seen": 26905696,
      "step": 31070
    },
    {
      "epoch": 14.651107967939652,
      "grad_norm": 0.005129913333803415,
      "learning_rate": 0.03537485116977327,
      "loss": 0.178,
      "num_input_tokens_seen": 26909600,
      "step": 31075
    },
    {
      "epoch": 14.653465346534654,
      "grad_norm": 0.009497813880443573,
      "learning_rate": 0.03533686533231565,
      "loss": 0.1959,
      "num_input_tokens_seen": 26914304,
      "step": 31080
    },
    {
      "epoch": 14.655822725129656,
      "grad_norm": 0.005658426322042942,
      "learning_rate": 0.0352988971773543,
      "loss": 0.1635,
      "num_input_tokens_seen": 26918400,
      "step": 31085
    },
    {
      "epoch": 14.658180103724659,
      "grad_norm": 0.005134064704179764,
      "learning_rate": 0.03526094671074443,
      "loss": 0.1503,
      "num_input_tokens_seen": 26922720,
      "step": 31090
    },
    {
      "epoch": 14.660537482319661,
      "grad_norm": 0.005945050157606602,
      "learning_rate": 0.03522301393833852,
      "loss": 0.1871,
      "num_input_tokens_seen": 26927504,
      "step": 31095
    },
    {
      "epoch": 14.662894860914664,
      "grad_norm": 0.0078046247363090515,
      "learning_rate": 0.035185098865986204,
      "loss": 0.2361,
      "num_input_tokens_seen": 26932000,
      "step": 31100
    },
    {
      "epoch": 14.665252239509666,
      "grad_norm": 0.008537321351468563,
      "learning_rate": 0.03514720149953453,
      "loss": 0.2006,
      "num_input_tokens_seen": 26936704,
      "step": 31105
    },
    {
      "epoch": 14.667609618104667,
      "grad_norm": 0.006612342782318592,
      "learning_rate": 0.03510932184482773,
      "loss": 0.2019,
      "num_input_tokens_seen": 26941296,
      "step": 31110
    },
    {
      "epoch": 14.66996699669967,
      "grad_norm": 0.0032411497086286545,
      "learning_rate": 0.03507145990770724,
      "loss": 0.1692,
      "num_input_tokens_seen": 26946336,
      "step": 31115
    },
    {
      "epoch": 14.672324375294671,
      "grad_norm": 0.0038390459958463907,
      "learning_rate": 0.035033615694011984,
      "loss": 0.1127,
      "num_input_tokens_seen": 26950544,
      "step": 31120
    },
    {
      "epoch": 14.674681753889674,
      "grad_norm": 0.006515556015074253,
      "learning_rate": 0.03499578920957788,
      "loss": 0.1673,
      "num_input_tokens_seen": 26955088,
      "step": 31125
    },
    {
      "epoch": 14.677039132484676,
      "grad_norm": 0.004634056705981493,
      "learning_rate": 0.034957980460238375,
      "loss": 0.1804,
      "num_input_tokens_seen": 26959920,
      "step": 31130
    },
    {
      "epoch": 14.679396511079679,
      "grad_norm": 0.009707312099635601,
      "learning_rate": 0.03492018945182393,
      "loss": 0.2016,
      "num_input_tokens_seen": 26964160,
      "step": 31135
    },
    {
      "epoch": 14.681753889674681,
      "grad_norm": 0.005718636326491833,
      "learning_rate": 0.03488241619016247,
      "loss": 0.1753,
      "num_input_tokens_seen": 26968160,
      "step": 31140
    },
    {
      "epoch": 14.684111268269683,
      "grad_norm": 0.00897898431867361,
      "learning_rate": 0.03484466068107913,
      "loss": 0.169,
      "num_input_tokens_seen": 26972560,
      "step": 31145
    },
    {
      "epoch": 14.686468646864686,
      "grad_norm": 0.010096997022628784,
      "learning_rate": 0.034806922930396195,
      "loss": 0.2276,
      "num_input_tokens_seen": 26976464,
      "step": 31150
    },
    {
      "epoch": 14.688826025459688,
      "grad_norm": 0.003829028457403183,
      "learning_rate": 0.03476920294393337,
      "loss": 0.169,
      "num_input_tokens_seen": 26981040,
      "step": 31155
    },
    {
      "epoch": 14.69118340405469,
      "grad_norm": 0.012253743596374989,
      "learning_rate": 0.03473150072750755,
      "loss": 0.2044,
      "num_input_tokens_seen": 26985152,
      "step": 31160
    },
    {
      "epoch": 14.693540782649693,
      "grad_norm": 0.00750062195584178,
      "learning_rate": 0.03469381628693284,
      "loss": 0.2548,
      "num_input_tokens_seen": 26989168,
      "step": 31165
    },
    {
      "epoch": 14.695898161244696,
      "grad_norm": 0.0058988165110349655,
      "learning_rate": 0.03465614962802072,
      "loss": 0.2135,
      "num_input_tokens_seen": 26994112,
      "step": 31170
    },
    {
      "epoch": 14.698255539839698,
      "grad_norm": 0.005652610212564468,
      "learning_rate": 0.0346185007565798,
      "loss": 0.1797,
      "num_input_tokens_seen": 26998272,
      "step": 31175
    },
    {
      "epoch": 14.7006129184347,
      "grad_norm": 0.006031505297869444,
      "learning_rate": 0.03458086967841609,
      "loss": 0.1574,
      "num_input_tokens_seen": 27002240,
      "step": 31180
    },
    {
      "epoch": 14.702970297029703,
      "grad_norm": 0.007745372597128153,
      "learning_rate": 0.03454325639933266,
      "loss": 0.2028,
      "num_input_tokens_seen": 27006240,
      "step": 31185
    },
    {
      "epoch": 14.705327675624705,
      "grad_norm": 0.005938663613051176,
      "learning_rate": 0.03450566092513007,
      "loss": 0.1504,
      "num_input_tokens_seen": 27010288,
      "step": 31190
    },
    {
      "epoch": 14.707685054219708,
      "grad_norm": 0.004308302886784077,
      "learning_rate": 0.034468083261605914,
      "loss": 0.1725,
      "num_input_tokens_seen": 27014144,
      "step": 31195
    },
    {
      "epoch": 14.71004243281471,
      "grad_norm": 0.006093740463256836,
      "learning_rate": 0.03443052341455522,
      "loss": 0.179,
      "num_input_tokens_seen": 27017952,
      "step": 31200
    },
    {
      "epoch": 14.71004243281471,
      "eval_loss": 0.21963630616664886,
      "eval_runtime": 21.939,
      "eval_samples_per_second": 42.983,
      "eval_steps_per_second": 21.514,
      "num_input_tokens_seen": 27017952,
      "step": 31200
    },
    {
      "epoch": 14.712399811409712,
      "grad_norm": 0.006688254419714212,
      "learning_rate": 0.0343929813897701,
      "loss": 0.1483,
      "num_input_tokens_seen": 27021600,
      "step": 31205
    },
    {
      "epoch": 14.714757190004715,
      "grad_norm": 0.009602392092347145,
      "learning_rate": 0.034355457193040125,
      "loss": 0.1753,
      "num_input_tokens_seen": 27026400,
      "step": 31210
    },
    {
      "epoch": 14.717114568599717,
      "grad_norm": 0.003857169533148408,
      "learning_rate": 0.03431795083015186,
      "loss": 0.1761,
      "num_input_tokens_seen": 27030352,
      "step": 31215
    },
    {
      "epoch": 14.71947194719472,
      "grad_norm": 0.004795355722308159,
      "learning_rate": 0.03428046230688936,
      "loss": 0.1524,
      "num_input_tokens_seen": 27034528,
      "step": 31220
    },
    {
      "epoch": 14.721829325789722,
      "grad_norm": 0.007884330116212368,
      "learning_rate": 0.034242991629033805,
      "loss": 0.2026,
      "num_input_tokens_seen": 27038640,
      "step": 31225
    },
    {
      "epoch": 14.724186704384724,
      "grad_norm": 0.0042383079417049885,
      "learning_rate": 0.03420553880236362,
      "loss": 0.1221,
      "num_input_tokens_seen": 27042960,
      "step": 31230
    },
    {
      "epoch": 14.726544082979727,
      "grad_norm": 0.007879868149757385,
      "learning_rate": 0.03416810383265449,
      "loss": 0.2238,
      "num_input_tokens_seen": 27048592,
      "step": 31235
    },
    {
      "epoch": 14.72890146157473,
      "grad_norm": 0.005207326728850603,
      "learning_rate": 0.03413068672567944,
      "loss": 0.174,
      "num_input_tokens_seen": 27053072,
      "step": 31240
    },
    {
      "epoch": 14.731258840169732,
      "grad_norm": 0.006050557363778353,
      "learning_rate": 0.034093287487208565,
      "loss": 0.1566,
      "num_input_tokens_seen": 27056784,
      "step": 31245
    },
    {
      "epoch": 14.733616218764734,
      "grad_norm": 0.008078157901763916,
      "learning_rate": 0.03405590612300937,
      "loss": 0.2179,
      "num_input_tokens_seen": 27060288,
      "step": 31250
    },
    {
      "epoch": 14.735973597359736,
      "grad_norm": 0.007427724078297615,
      "learning_rate": 0.03401854263884646,
      "loss": 0.1677,
      "num_input_tokens_seen": 27065344,
      "step": 31255
    },
    {
      "epoch": 14.738330975954739,
      "grad_norm": 0.010830432176589966,
      "learning_rate": 0.033981197040481824,
      "loss": 0.2085,
      "num_input_tokens_seen": 27069184,
      "step": 31260
    },
    {
      "epoch": 14.740688354549741,
      "grad_norm": 0.010542825795710087,
      "learning_rate": 0.03394386933367459,
      "loss": 0.1572,
      "num_input_tokens_seen": 27074496,
      "step": 31265
    },
    {
      "epoch": 14.743045733144744,
      "grad_norm": 0.0029538762755692005,
      "learning_rate": 0.033906559524181104,
      "loss": 0.1462,
      "num_input_tokens_seen": 27078304,
      "step": 31270
    },
    {
      "epoch": 14.745403111739746,
      "grad_norm": 0.003832175163552165,
      "learning_rate": 0.033869267617755085,
      "loss": 0.2213,
      "num_input_tokens_seen": 27082848,
      "step": 31275
    },
    {
      "epoch": 14.747760490334748,
      "grad_norm": 0.008046318776905537,
      "learning_rate": 0.0338319936201474,
      "loss": 0.1584,
      "num_input_tokens_seen": 27087184,
      "step": 31280
    },
    {
      "epoch": 14.75011786892975,
      "grad_norm": 0.009863331913948059,
      "learning_rate": 0.033794737537106136,
      "loss": 0.2069,
      "num_input_tokens_seen": 27091072,
      "step": 31285
    },
    {
      "epoch": 14.752475247524753,
      "grad_norm": 0.0031112448778003454,
      "learning_rate": 0.03375749937437671,
      "loss": 0.1178,
      "num_input_tokens_seen": 27095680,
      "step": 31290
    },
    {
      "epoch": 14.754832626119756,
      "grad_norm": 0.005577183328568935,
      "learning_rate": 0.033720279137701634,
      "loss": 0.1804,
      "num_input_tokens_seen": 27100160,
      "step": 31295
    },
    {
      "epoch": 14.757190004714758,
      "grad_norm": 0.008729926310479641,
      "learning_rate": 0.03368307683282078,
      "loss": 0.205,
      "num_input_tokens_seen": 27104384,
      "step": 31300
    },
    {
      "epoch": 14.75954738330976,
      "grad_norm": 0.003878445364534855,
      "learning_rate": 0.033645892465471235,
      "loss": 0.1249,
      "num_input_tokens_seen": 27108208,
      "step": 31305
    },
    {
      "epoch": 14.761904761904763,
      "grad_norm": 0.0052953679114580154,
      "learning_rate": 0.03360872604138724,
      "loss": 0.1733,
      "num_input_tokens_seen": 27112384,
      "step": 31310
    },
    {
      "epoch": 14.764262140499763,
      "grad_norm": 0.00709918886423111,
      "learning_rate": 0.03357157756630034,
      "loss": 0.2248,
      "num_input_tokens_seen": 27116704,
      "step": 31315
    },
    {
      "epoch": 14.766619519094768,
      "grad_norm": 0.008181417360901833,
      "learning_rate": 0.033534447045939365,
      "loss": 0.2452,
      "num_input_tokens_seen": 27121776,
      "step": 31320
    },
    {
      "epoch": 14.768976897689768,
      "grad_norm": 0.007131804712116718,
      "learning_rate": 0.03349733448603026,
      "loss": 0.1893,
      "num_input_tokens_seen": 27127392,
      "step": 31325
    },
    {
      "epoch": 14.77133427628477,
      "grad_norm": 0.008365755900740623,
      "learning_rate": 0.03346023989229619,
      "loss": 0.173,
      "num_input_tokens_seen": 27131360,
      "step": 31330
    },
    {
      "epoch": 14.773691654879773,
      "grad_norm": 0.007154990918934345,
      "learning_rate": 0.03342316327045769,
      "loss": 0.1618,
      "num_input_tokens_seen": 27135536,
      "step": 31335
    },
    {
      "epoch": 14.776049033474775,
      "grad_norm": 0.005592873319983482,
      "learning_rate": 0.033386104626232385,
      "loss": 0.1738,
      "num_input_tokens_seen": 27140528,
      "step": 31340
    },
    {
      "epoch": 14.778406412069778,
      "grad_norm": 0.0059942808002233505,
      "learning_rate": 0.03334906396533525,
      "loss": 0.2613,
      "num_input_tokens_seen": 27145216,
      "step": 31345
    },
    {
      "epoch": 14.78076379066478,
      "grad_norm": 0.004958879202604294,
      "learning_rate": 0.033312041293478326,
      "loss": 0.1538,
      "num_input_tokens_seen": 27149792,
      "step": 31350
    },
    {
      "epoch": 14.783121169259783,
      "grad_norm": 0.004528856370598078,
      "learning_rate": 0.03327503661637103,
      "loss": 0.2081,
      "num_input_tokens_seen": 27154832,
      "step": 31355
    },
    {
      "epoch": 14.785478547854785,
      "grad_norm": 0.003867924911901355,
      "learning_rate": 0.03323804993971998,
      "loss": 0.1411,
      "num_input_tokens_seen": 27158352,
      "step": 31360
    },
    {
      "epoch": 14.787835926449787,
      "grad_norm": 0.00661414535716176,
      "learning_rate": 0.033201081269228924,
      "loss": 0.174,
      "num_input_tokens_seen": 27162208,
      "step": 31365
    },
    {
      "epoch": 14.79019330504479,
      "grad_norm": 0.007284439634531736,
      "learning_rate": 0.03316413061059895,
      "loss": 0.1541,
      "num_input_tokens_seen": 27166272,
      "step": 31370
    },
    {
      "epoch": 14.792550683639792,
      "grad_norm": 0.004178010392934084,
      "learning_rate": 0.03312719796952827,
      "loss": 0.1765,
      "num_input_tokens_seen": 27170208,
      "step": 31375
    },
    {
      "epoch": 14.794908062234795,
      "grad_norm": 0.010041406378149986,
      "learning_rate": 0.03309028335171236,
      "loss": 0.1686,
      "num_input_tokens_seen": 27174336,
      "step": 31380
    },
    {
      "epoch": 14.797265440829797,
      "grad_norm": 0.005167939700186253,
      "learning_rate": 0.03305338676284398,
      "loss": 0.1528,
      "num_input_tokens_seen": 27179008,
      "step": 31385
    },
    {
      "epoch": 14.7996228194248,
      "grad_norm": 0.004662918392568827,
      "learning_rate": 0.03301650820861296,
      "loss": 0.1553,
      "num_input_tokens_seen": 27183536,
      "step": 31390
    },
    {
      "epoch": 14.801980198019802,
      "grad_norm": 0.00854891911149025,
      "learning_rate": 0.03297964769470652,
      "loss": 0.2318,
      "num_input_tokens_seen": 27187728,
      "step": 31395
    },
    {
      "epoch": 14.804337576614804,
      "grad_norm": 0.007363725453615189,
      "learning_rate": 0.032942805226808945,
      "loss": 0.2227,
      "num_input_tokens_seen": 27191600,
      "step": 31400
    },
    {
      "epoch": 14.804337576614804,
      "eval_loss": 0.22277867794036865,
      "eval_runtime": 21.9277,
      "eval_samples_per_second": 43.005,
      "eval_steps_per_second": 21.525,
      "num_input_tokens_seen": 27191600,
      "step": 31400
    },
    {
      "epoch": 14.806694955209807,
      "grad_norm": 0.007005356252193451,
      "learning_rate": 0.03290598081060187,
      "loss": 0.2265,
      "num_input_tokens_seen": 27195712,
      "step": 31405
    },
    {
      "epoch": 14.809052333804809,
      "grad_norm": 0.008061530068516731,
      "learning_rate": 0.03286917445176407,
      "loss": 0.1856,
      "num_input_tokens_seen": 27199472,
      "step": 31410
    },
    {
      "epoch": 14.811409712399811,
      "grad_norm": 0.007801678031682968,
      "learning_rate": 0.032832386155971456,
      "loss": 0.1855,
      "num_input_tokens_seen": 27204000,
      "step": 31415
    },
    {
      "epoch": 14.813767090994814,
      "grad_norm": 0.008730773814022541,
      "learning_rate": 0.032795615928897334,
      "loss": 0.2689,
      "num_input_tokens_seen": 27208208,
      "step": 31420
    },
    {
      "epoch": 14.816124469589816,
      "grad_norm": 0.0058345007710158825,
      "learning_rate": 0.03275886377621215,
      "loss": 0.1538,
      "num_input_tokens_seen": 27212256,
      "step": 31425
    },
    {
      "epoch": 14.818481848184819,
      "grad_norm": 0.01002301275730133,
      "learning_rate": 0.03272212970358348,
      "loss": 0.2258,
      "num_input_tokens_seen": 27216736,
      "step": 31430
    },
    {
      "epoch": 14.820839226779821,
      "grad_norm": 0.004367421381175518,
      "learning_rate": 0.032685413716676215,
      "loss": 0.1584,
      "num_input_tokens_seen": 27221072,
      "step": 31435
    },
    {
      "epoch": 14.823196605374823,
      "grad_norm": 0.004373235162347555,
      "learning_rate": 0.032648715821152474,
      "loss": 0.192,
      "num_input_tokens_seen": 27224688,
      "step": 31440
    },
    {
      "epoch": 14.825553983969826,
      "grad_norm": 0.005812883842736483,
      "learning_rate": 0.03261203602267143,
      "loss": 0.1724,
      "num_input_tokens_seen": 27228656,
      "step": 31445
    },
    {
      "epoch": 14.827911362564828,
      "grad_norm": 0.005029674619436264,
      "learning_rate": 0.03257537432688966,
      "loss": 0.2359,
      "num_input_tokens_seen": 27232704,
      "step": 31450
    },
    {
      "epoch": 14.83026874115983,
      "grad_norm": 0.005447214934974909,
      "learning_rate": 0.03253873073946077,
      "loss": 0.2082,
      "num_input_tokens_seen": 27236640,
      "step": 31455
    },
    {
      "epoch": 14.832626119754833,
      "grad_norm": 0.0024080912116914988,
      "learning_rate": 0.03250210526603572,
      "loss": 0.1413,
      "num_input_tokens_seen": 27240704,
      "step": 31460
    },
    {
      "epoch": 14.834983498349835,
      "grad_norm": 0.008191054686903954,
      "learning_rate": 0.03246549791226266,
      "loss": 0.1989,
      "num_input_tokens_seen": 27244944,
      "step": 31465
    },
    {
      "epoch": 14.837340876944838,
      "grad_norm": 0.00554154347628355,
      "learning_rate": 0.03242890868378679,
      "loss": 0.1632,
      "num_input_tokens_seen": 27249680,
      "step": 31470
    },
    {
      "epoch": 14.83969825553984,
      "grad_norm": 0.004086107946932316,
      "learning_rate": 0.03239233758625074,
      "loss": 0.1764,
      "num_input_tokens_seen": 27254688,
      "step": 31475
    },
    {
      "epoch": 14.842055634134843,
      "grad_norm": 0.0054254597052931786,
      "learning_rate": 0.032355784625294204,
      "loss": 0.1736,
      "num_input_tokens_seen": 27258448,
      "step": 31480
    },
    {
      "epoch": 14.844413012729845,
      "grad_norm": 0.009827401489019394,
      "learning_rate": 0.03231924980655402,
      "loss": 0.2116,
      "num_input_tokens_seen": 27262848,
      "step": 31485
    },
    {
      "epoch": 14.846770391324847,
      "grad_norm": 0.005223838612437248,
      "learning_rate": 0.032282733135664446,
      "loss": 0.1797,
      "num_input_tokens_seen": 27267520,
      "step": 31490
    },
    {
      "epoch": 14.84912776991985,
      "grad_norm": 0.006793577689677477,
      "learning_rate": 0.03224623461825669,
      "loss": 0.1354,
      "num_input_tokens_seen": 27271456,
      "step": 31495
    },
    {
      "epoch": 14.851485148514852,
      "grad_norm": 0.004696226213127375,
      "learning_rate": 0.03220975425995937,
      "loss": 0.1191,
      "num_input_tokens_seen": 27275360,
      "step": 31500
    },
    {
      "epoch": 14.853842527109855,
      "grad_norm": 0.005338048562407494,
      "learning_rate": 0.032173292066398206,
      "loss": 0.2282,
      "num_input_tokens_seen": 27279312,
      "step": 31505
    },
    {
      "epoch": 14.856199905704855,
      "grad_norm": 0.0020173415541648865,
      "learning_rate": 0.03213684804319606,
      "loss": 0.1578,
      "num_input_tokens_seen": 27284432,
      "step": 31510
    },
    {
      "epoch": 14.85855728429986,
      "grad_norm": 0.007944039069116116,
      "learning_rate": 0.03210042219597312,
      "loss": 0.144,
      "num_input_tokens_seen": 27288432,
      "step": 31515
    },
    {
      "epoch": 14.86091466289486,
      "grad_norm": 0.004040560685098171,
      "learning_rate": 0.03206401453034675,
      "loss": 0.1777,
      "num_input_tokens_seen": 27293024,
      "step": 31520
    },
    {
      "epoch": 14.863272041489862,
      "grad_norm": 0.00574663607403636,
      "learning_rate": 0.03202762505193136,
      "loss": 0.1726,
      "num_input_tokens_seen": 27297424,
      "step": 31525
    },
    {
      "epoch": 14.865629420084865,
      "grad_norm": 0.005537412595003843,
      "learning_rate": 0.031991253766338754,
      "loss": 0.1859,
      "num_input_tokens_seen": 27301696,
      "step": 31530
    },
    {
      "epoch": 14.867986798679867,
      "grad_norm": 0.004909951239824295,
      "learning_rate": 0.03195490067917778,
      "loss": 0.1735,
      "num_input_tokens_seen": 27306608,
      "step": 31535
    },
    {
      "epoch": 14.87034417727487,
      "grad_norm": 0.005349540151655674,
      "learning_rate": 0.03191856579605461,
      "loss": 0.1299,
      "num_input_tokens_seen": 27310464,
      "step": 31540
    },
    {
      "epoch": 14.872701555869872,
      "grad_norm": 0.00814803596585989,
      "learning_rate": 0.031882249122572454,
      "loss": 0.1749,
      "num_input_tokens_seen": 27314992,
      "step": 31545
    },
    {
      "epoch": 14.875058934464874,
      "grad_norm": 0.004097393248230219,
      "learning_rate": 0.03184595066433188,
      "loss": 0.1773,
      "num_input_tokens_seen": 27319104,
      "step": 31550
    },
    {
      "epoch": 14.877416313059877,
      "grad_norm": 0.01075207069516182,
      "learning_rate": 0.03180967042693049,
      "loss": 0.2199,
      "num_input_tokens_seen": 27323168,
      "step": 31555
    },
    {
      "epoch": 14.87977369165488,
      "grad_norm": 0.0035620720591396093,
      "learning_rate": 0.03177340841596323,
      "loss": 0.2038,
      "num_input_tokens_seen": 27327408,
      "step": 31560
    },
    {
      "epoch": 14.882131070249882,
      "grad_norm": 0.005206974223256111,
      "learning_rate": 0.03173716463702209,
      "loss": 0.1774,
      "num_input_tokens_seen": 27331456,
      "step": 31565
    },
    {
      "epoch": 14.884488448844884,
      "grad_norm": 0.008503629826009274,
      "learning_rate": 0.03170093909569638,
      "loss": 0.201,
      "num_input_tokens_seen": 27335168,
      "step": 31570
    },
    {
      "epoch": 14.886845827439886,
      "grad_norm": 0.005828293971717358,
      "learning_rate": 0.03166473179757246,
      "loss": 0.1851,
      "num_input_tokens_seen": 27339392,
      "step": 31575
    },
    {
      "epoch": 14.889203206034889,
      "grad_norm": 0.004316146485507488,
      "learning_rate": 0.031628542748234005,
      "loss": 0.1419,
      "num_input_tokens_seen": 27343776,
      "step": 31580
    },
    {
      "epoch": 14.891560584629891,
      "grad_norm": 0.005087959114462137,
      "learning_rate": 0.03159237195326184,
      "loss": 0.1946,
      "num_input_tokens_seen": 27348016,
      "step": 31585
    },
    {
      "epoch": 14.893917963224894,
      "grad_norm": 0.007090835366398096,
      "learning_rate": 0.031556219418233875,
      "loss": 0.215,
      "num_input_tokens_seen": 27352528,
      "step": 31590
    },
    {
      "epoch": 14.896275341819896,
      "grad_norm": 0.006120446603745222,
      "learning_rate": 0.03152008514872533,
      "loss": 0.172,
      "num_input_tokens_seen": 27357856,
      "step": 31595
    },
    {
      "epoch": 14.898632720414899,
      "grad_norm": 0.00725463917478919,
      "learning_rate": 0.03148396915030862,
      "loss": 0.1798,
      "num_input_tokens_seen": 27362144,
      "step": 31600
    },
    {
      "epoch": 14.898632720414899,
      "eval_loss": 0.22433394193649292,
      "eval_runtime": 21.916,
      "eval_samples_per_second": 43.028,
      "eval_steps_per_second": 21.537,
      "num_input_tokens_seen": 27362144,
      "step": 31600
    },
    {
      "epoch": 14.900990099009901,
      "grad_norm": 0.010752711445093155,
      "learning_rate": 0.03144787142855318,
      "loss": 0.1813,
      "num_input_tokens_seen": 27366016,
      "step": 31605
    },
    {
      "epoch": 14.903347477604903,
      "grad_norm": 0.005783167667686939,
      "learning_rate": 0.031411791989025835,
      "loss": 0.2004,
      "num_input_tokens_seen": 27370320,
      "step": 31610
    },
    {
      "epoch": 14.905704856199906,
      "grad_norm": 0.008216471411287785,
      "learning_rate": 0.031375730837290394,
      "loss": 0.1683,
      "num_input_tokens_seen": 27374096,
      "step": 31615
    },
    {
      "epoch": 14.908062234794908,
      "grad_norm": 0.006392409093677998,
      "learning_rate": 0.031339687978908015,
      "loss": 0.1438,
      "num_input_tokens_seen": 27377984,
      "step": 31620
    },
    {
      "epoch": 14.91041961338991,
      "grad_norm": 0.00805665273219347,
      "learning_rate": 0.03130366341943694,
      "loss": 0.2728,
      "num_input_tokens_seen": 27382368,
      "step": 31625
    },
    {
      "epoch": 14.912776991984913,
      "grad_norm": 0.006121763028204441,
      "learning_rate": 0.031267657164432555,
      "loss": 0.1722,
      "num_input_tokens_seen": 27386832,
      "step": 31630
    },
    {
      "epoch": 14.915134370579915,
      "grad_norm": 0.010816319845616817,
      "learning_rate": 0.03123166921944752,
      "loss": 0.2638,
      "num_input_tokens_seen": 27390736,
      "step": 31635
    },
    {
      "epoch": 14.917491749174918,
      "grad_norm": 0.005303744226694107,
      "learning_rate": 0.031195699590031666,
      "loss": 0.1103,
      "num_input_tokens_seen": 27394976,
      "step": 31640
    },
    {
      "epoch": 14.91984912776992,
      "grad_norm": 0.007325414568185806,
      "learning_rate": 0.031159748281731885,
      "loss": 0.1846,
      "num_input_tokens_seen": 27399840,
      "step": 31645
    },
    {
      "epoch": 14.922206506364923,
      "grad_norm": 0.0028772037476301193,
      "learning_rate": 0.031123815300092394,
      "loss": 0.1757,
      "num_input_tokens_seen": 27404400,
      "step": 31650
    },
    {
      "epoch": 14.924563884959925,
      "grad_norm": 0.005404707510024309,
      "learning_rate": 0.031087900650654424,
      "loss": 0.1953,
      "num_input_tokens_seen": 27408176,
      "step": 31655
    },
    {
      "epoch": 14.926921263554927,
      "grad_norm": 0.007426152937114239,
      "learning_rate": 0.031052004338956534,
      "loss": 0.1566,
      "num_input_tokens_seen": 27412720,
      "step": 31660
    },
    {
      "epoch": 14.92927864214993,
      "grad_norm": 0.007629053201526403,
      "learning_rate": 0.031016126370534407,
      "loss": 0.1839,
      "num_input_tokens_seen": 27416688,
      "step": 31665
    },
    {
      "epoch": 14.931636020744932,
      "grad_norm": 0.0050683533772826195,
      "learning_rate": 0.030980266750920804,
      "loss": 0.1493,
      "num_input_tokens_seen": 27420976,
      "step": 31670
    },
    {
      "epoch": 14.933993399339935,
      "grad_norm": 0.009494072757661343,
      "learning_rate": 0.030944425485645747,
      "loss": 0.1675,
      "num_input_tokens_seen": 27425088,
      "step": 31675
    },
    {
      "epoch": 14.936350777934937,
      "grad_norm": 0.007550706155598164,
      "learning_rate": 0.03090860258023647,
      "loss": 0.1449,
      "num_input_tokens_seen": 27429952,
      "step": 31680
    },
    {
      "epoch": 14.93870815652994,
      "grad_norm": 0.003941238857805729,
      "learning_rate": 0.030872798040217236,
      "loss": 0.1754,
      "num_input_tokens_seen": 27434144,
      "step": 31685
    },
    {
      "epoch": 14.941065535124942,
      "grad_norm": 0.006010878831148148,
      "learning_rate": 0.03083701187110964,
      "loss": 0.1578,
      "num_input_tokens_seen": 27438640,
      "step": 31690
    },
    {
      "epoch": 14.943422913719944,
      "grad_norm": 0.0045889937318861485,
      "learning_rate": 0.030801244078432294,
      "loss": 0.1354,
      "num_input_tokens_seen": 27444000,
      "step": 31695
    },
    {
      "epoch": 14.945780292314947,
      "grad_norm": 0.004919214174151421,
      "learning_rate": 0.030765494667701024,
      "loss": 0.1437,
      "num_input_tokens_seen": 27448160,
      "step": 31700
    },
    {
      "epoch": 14.948137670909949,
      "grad_norm": 0.008427339605987072,
      "learning_rate": 0.030729763644428913,
      "loss": 0.1554,
      "num_input_tokens_seen": 27453040,
      "step": 31705
    },
    {
      "epoch": 14.950495049504951,
      "grad_norm": 0.010007768869400024,
      "learning_rate": 0.030694051014126048,
      "loss": 0.171,
      "num_input_tokens_seen": 27457280,
      "step": 31710
    },
    {
      "epoch": 14.952852428099952,
      "grad_norm": 0.00824948213994503,
      "learning_rate": 0.030658356782299792,
      "loss": 0.2098,
      "num_input_tokens_seen": 27461136,
      "step": 31715
    },
    {
      "epoch": 14.955209806694956,
      "grad_norm": 0.010132517665624619,
      "learning_rate": 0.030622680954454726,
      "loss": 0.143,
      "num_input_tokens_seen": 27465216,
      "step": 31720
    },
    {
      "epoch": 14.957567185289957,
      "grad_norm": 0.009320967830717564,
      "learning_rate": 0.030587023536092398,
      "loss": 0.1813,
      "num_input_tokens_seen": 27470448,
      "step": 31725
    },
    {
      "epoch": 14.95992456388496,
      "grad_norm": 0.006410930305719376,
      "learning_rate": 0.03055138453271171,
      "loss": 0.1967,
      "num_input_tokens_seen": 27474912,
      "step": 31730
    },
    {
      "epoch": 14.962281942479962,
      "grad_norm": 0.0030152248218655586,
      "learning_rate": 0.03051576394980858,
      "loss": 0.1735,
      "num_input_tokens_seen": 27479776,
      "step": 31735
    },
    {
      "epoch": 14.964639321074964,
      "grad_norm": 0.004947801120579243,
      "learning_rate": 0.030480161792876187,
      "loss": 0.114,
      "num_input_tokens_seen": 27484400,
      "step": 31740
    },
    {
      "epoch": 14.966996699669966,
      "grad_norm": 0.008130255155265331,
      "learning_rate": 0.030444578067404846,
      "loss": 0.1892,
      "num_input_tokens_seen": 27488912,
      "step": 31745
    },
    {
      "epoch": 14.969354078264969,
      "grad_norm": 0.006962026469409466,
      "learning_rate": 0.030409012778881975,
      "loss": 0.1437,
      "num_input_tokens_seen": 27493024,
      "step": 31750
    },
    {
      "epoch": 14.971711456859971,
      "grad_norm": 0.00539055373519659,
      "learning_rate": 0.030373465932792235,
      "loss": 0.18,
      "num_input_tokens_seen": 27497456,
      "step": 31755
    },
    {
      "epoch": 14.974068835454974,
      "grad_norm": 0.00877026654779911,
      "learning_rate": 0.030337937534617342,
      "loss": 0.2359,
      "num_input_tokens_seen": 27501968,
      "step": 31760
    },
    {
      "epoch": 14.976426214049976,
      "grad_norm": 0.0077596926130354404,
      "learning_rate": 0.030302427589836277,
      "loss": 0.2131,
      "num_input_tokens_seen": 27505936,
      "step": 31765
    },
    {
      "epoch": 14.978783592644978,
      "grad_norm": 0.004632691852748394,
      "learning_rate": 0.030266936103925095,
      "loss": 0.1598,
      "num_input_tokens_seen": 27510208,
      "step": 31770
    },
    {
      "epoch": 14.98114097123998,
      "grad_norm": 0.009307975880801678,
      "learning_rate": 0.030231463082356982,
      "loss": 0.2037,
      "num_input_tokens_seen": 27514640,
      "step": 31775
    },
    {
      "epoch": 14.983498349834983,
      "grad_norm": 0.003248799592256546,
      "learning_rate": 0.030196008530602367,
      "loss": 0.0909,
      "num_input_tokens_seen": 27519344,
      "step": 31780
    },
    {
      "epoch": 14.985855728429986,
      "grad_norm": 0.006391605362296104,
      "learning_rate": 0.030160572454128842,
      "loss": 0.2045,
      "num_input_tokens_seen": 27523408,
      "step": 31785
    },
    {
      "epoch": 14.988213107024988,
      "grad_norm": 0.006999519187957048,
      "learning_rate": 0.03012515485840098,
      "loss": 0.2363,
      "num_input_tokens_seen": 27528176,
      "step": 31790
    },
    {
      "epoch": 14.99057048561999,
      "grad_norm": 0.005662770010530949,
      "learning_rate": 0.030089755748880734,
      "loss": 0.1688,
      "num_input_tokens_seen": 27532640,
      "step": 31795
    },
    {
      "epoch": 14.992927864214993,
      "grad_norm": 0.004766967613250017,
      "learning_rate": 0.030054375131027003,
      "loss": 0.157,
      "num_input_tokens_seen": 27536992,
      "step": 31800
    },
    {
      "epoch": 14.992927864214993,
      "eval_loss": 0.21607425808906555,
      "eval_runtime": 21.9233,
      "eval_samples_per_second": 43.014,
      "eval_steps_per_second": 21.53,
      "num_input_tokens_seen": 27536992,
      "step": 31800
    },
    {
      "epoch": 14.995285242809995,
      "grad_norm": 0.0028982185758650303,
      "learning_rate": 0.030019013010295942,
      "loss": 0.2854,
      "num_input_tokens_seen": 27541840,
      "step": 31805
    },
    {
      "epoch": 14.997642621404998,
      "grad_norm": 0.0073151970282197,
      "learning_rate": 0.029983669392140897,
      "loss": 0.209,
      "num_input_tokens_seen": 27546128,
      "step": 31810
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.007877274416387081,
      "learning_rate": 0.029948344282012217,
      "loss": 0.2426,
      "num_input_tokens_seen": 27550432,
      "step": 31815
    },
    {
      "epoch": 15.002357378595002,
      "grad_norm": 0.003405193565413356,
      "learning_rate": 0.029913037685357507,
      "loss": 0.1081,
      "num_input_tokens_seen": 27554720,
      "step": 31820
    },
    {
      "epoch": 15.004714757190005,
      "grad_norm": 0.002649395726621151,
      "learning_rate": 0.029877749607621528,
      "loss": 0.1458,
      "num_input_tokens_seen": 27559920,
      "step": 31825
    },
    {
      "epoch": 15.007072135785007,
      "grad_norm": 0.00851258635520935,
      "learning_rate": 0.029842480054246077,
      "loss": 0.1222,
      "num_input_tokens_seen": 27563696,
      "step": 31830
    },
    {
      "epoch": 15.00942951438001,
      "grad_norm": 0.005291992798447609,
      "learning_rate": 0.02980722903067022,
      "loss": 0.1369,
      "num_input_tokens_seen": 27567888,
      "step": 31835
    },
    {
      "epoch": 15.011786892975012,
      "grad_norm": 0.005410695448517799,
      "learning_rate": 0.029771996542330113,
      "loss": 0.1323,
      "num_input_tokens_seen": 27572704,
      "step": 31840
    },
    {
      "epoch": 15.014144271570014,
      "grad_norm": 0.0050464291125535965,
      "learning_rate": 0.029736782594658954,
      "loss": 0.1253,
      "num_input_tokens_seen": 27576832,
      "step": 31845
    },
    {
      "epoch": 15.016501650165017,
      "grad_norm": 0.008322654291987419,
      "learning_rate": 0.029701587193087284,
      "loss": 0.2082,
      "num_input_tokens_seen": 27581008,
      "step": 31850
    },
    {
      "epoch": 15.01885902876002,
      "grad_norm": 0.0066838338971138,
      "learning_rate": 0.0296664103430426,
      "loss": 0.1698,
      "num_input_tokens_seen": 27585152,
      "step": 31855
    },
    {
      "epoch": 15.021216407355022,
      "grad_norm": 0.005041365046054125,
      "learning_rate": 0.029631252049949652,
      "loss": 0.1442,
      "num_input_tokens_seen": 27590336,
      "step": 31860
    },
    {
      "epoch": 15.023573785950024,
      "grad_norm": 0.006057805847376585,
      "learning_rate": 0.02959611231923031,
      "loss": 0.1657,
      "num_input_tokens_seen": 27594208,
      "step": 31865
    },
    {
      "epoch": 15.025931164545026,
      "grad_norm": 0.007288984954357147,
      "learning_rate": 0.029560991156303507,
      "loss": 0.156,
      "num_input_tokens_seen": 27598272,
      "step": 31870
    },
    {
      "epoch": 15.028288543140029,
      "grad_norm": 0.008319100365042686,
      "learning_rate": 0.02952588856658544,
      "loss": 0.1271,
      "num_input_tokens_seen": 27602064,
      "step": 31875
    },
    {
      "epoch": 15.030645921735031,
      "grad_norm": 0.00371488812379539,
      "learning_rate": 0.029490804555489296,
      "loss": 0.0898,
      "num_input_tokens_seen": 27605952,
      "step": 31880
    },
    {
      "epoch": 15.033003300330034,
      "grad_norm": 0.011071549728512764,
      "learning_rate": 0.029455739128425484,
      "loss": 0.1455,
      "num_input_tokens_seen": 27611088,
      "step": 31885
    },
    {
      "epoch": 15.035360678925036,
      "grad_norm": 0.0043593840673565865,
      "learning_rate": 0.029420692290801607,
      "loss": 0.1016,
      "num_input_tokens_seen": 27615952,
      "step": 31890
    },
    {
      "epoch": 15.037718057520038,
      "grad_norm": 0.0076629482209682465,
      "learning_rate": 0.02938566404802223,
      "loss": 0.1078,
      "num_input_tokens_seen": 27620432,
      "step": 31895
    },
    {
      "epoch": 15.04007543611504,
      "grad_norm": 0.006147939246147871,
      "learning_rate": 0.029350654405489195,
      "loss": 0.1657,
      "num_input_tokens_seen": 27624064,
      "step": 31900
    },
    {
      "epoch": 15.042432814710043,
      "grad_norm": 0.0037718666717410088,
      "learning_rate": 0.02931566336860145,
      "loss": 0.1231,
      "num_input_tokens_seen": 27628128,
      "step": 31905
    },
    {
      "epoch": 15.044790193305046,
      "grad_norm": 0.005786620546132326,
      "learning_rate": 0.02928069094275505,
      "loss": 0.1073,
      "num_input_tokens_seen": 27631920,
      "step": 31910
    },
    {
      "epoch": 15.047147571900048,
      "grad_norm": 0.0016655907966196537,
      "learning_rate": 0.02924573713334314,
      "loss": 0.1429,
      "num_input_tokens_seen": 27636032,
      "step": 31915
    },
    {
      "epoch": 15.049504950495049,
      "grad_norm": 0.008867889642715454,
      "learning_rate": 0.02921080194575603,
      "loss": 0.2735,
      "num_input_tokens_seen": 27639744,
      "step": 31920
    },
    {
      "epoch": 15.051862329090051,
      "grad_norm": 0.007612468209117651,
      "learning_rate": 0.029175885385381177,
      "loss": 0.1245,
      "num_input_tokens_seen": 27644064,
      "step": 31925
    },
    {
      "epoch": 15.054219707685053,
      "grad_norm": 0.005120874848216772,
      "learning_rate": 0.029140987457603223,
      "loss": 0.2033,
      "num_input_tokens_seen": 27648528,
      "step": 31930
    },
    {
      "epoch": 15.056577086280056,
      "grad_norm": 0.002650020644068718,
      "learning_rate": 0.029106108167803763,
      "loss": 0.1372,
      "num_input_tokens_seen": 27653072,
      "step": 31935
    },
    {
      "epoch": 15.058934464875058,
      "grad_norm": 0.005455195438116789,
      "learning_rate": 0.029071247521361674,
      "loss": 0.1963,
      "num_input_tokens_seen": 27657168,
      "step": 31940
    },
    {
      "epoch": 15.06129184347006,
      "grad_norm": 0.00237372238188982,
      "learning_rate": 0.029036405523652945,
      "loss": 0.0802,
      "num_input_tokens_seen": 27661248,
      "step": 31945
    },
    {
      "epoch": 15.063649222065063,
      "grad_norm": 0.009384174831211567,
      "learning_rate": 0.029001582180050577,
      "loss": 0.1473,
      "num_input_tokens_seen": 27665536,
      "step": 31950
    },
    {
      "epoch": 15.066006600660065,
      "grad_norm": 0.008022339083254337,
      "learning_rate": 0.02896677749592482,
      "loss": 0.2257,
      "num_input_tokens_seen": 27669392,
      "step": 31955
    },
    {
      "epoch": 15.068363979255068,
      "grad_norm": 0.009980757720768452,
      "learning_rate": 0.028931991476642938,
      "loss": 0.1437,
      "num_input_tokens_seen": 27673216,
      "step": 31960
    },
    {
      "epoch": 15.07072135785007,
      "grad_norm": 0.007686107885092497,
      "learning_rate": 0.028897224127569412,
      "loss": 0.1695,
      "num_input_tokens_seen": 27677744,
      "step": 31965
    },
    {
      "epoch": 15.073078736445073,
      "grad_norm": 0.006691863760352135,
      "learning_rate": 0.028862475454065832,
      "loss": 0.158,
      "num_input_tokens_seen": 27681504,
      "step": 31970
    },
    {
      "epoch": 15.075436115040075,
      "grad_norm": 0.008031461387872696,
      "learning_rate": 0.028827745461490806,
      "loss": 0.1719,
      "num_input_tokens_seen": 27685712,
      "step": 31975
    },
    {
      "epoch": 15.077793493635077,
      "grad_norm": 0.012139732018113136,
      "learning_rate": 0.028793034155200212,
      "loss": 0.1486,
      "num_input_tokens_seen": 27690000,
      "step": 31980
    },
    {
      "epoch": 15.08015087223008,
      "grad_norm": 0.010516410693526268,
      "learning_rate": 0.028758341540546944,
      "loss": 0.245,
      "num_input_tokens_seen": 27693968,
      "step": 31985
    },
    {
      "epoch": 15.082508250825082,
      "grad_norm": 0.005717732012271881,
      "learning_rate": 0.02872366762288098,
      "loss": 0.1468,
      "num_input_tokens_seen": 27698544,
      "step": 31990
    },
    {
      "epoch": 15.084865629420085,
      "grad_norm": 0.00811387225985527,
      "learning_rate": 0.028689012407549567,
      "loss": 0.1458,
      "num_input_tokens_seen": 27703424,
      "step": 31995
    },
    {
      "epoch": 15.087223008015087,
      "grad_norm": 0.006718137301504612,
      "learning_rate": 0.028654375899896892,
      "loss": 0.1691,
      "num_input_tokens_seen": 27707728,
      "step": 32000
    },
    {
      "epoch": 15.087223008015087,
      "eval_loss": 0.2260258048772812,
      "eval_runtime": 21.9299,
      "eval_samples_per_second": 43.001,
      "eval_steps_per_second": 21.523,
      "num_input_tokens_seen": 27707728,
      "step": 32000
    },
    {
      "epoch": 15.08958038661009,
      "grad_norm": 0.007443042006343603,
      "learning_rate": 0.02861975810526437,
      "loss": 0.1584,
      "num_input_tokens_seen": 27712640,
      "step": 32005
    },
    {
      "epoch": 15.091937765205092,
      "grad_norm": 0.009210008196532726,
      "learning_rate": 0.02858515902899056,
      "loss": 0.1056,
      "num_input_tokens_seen": 27717984,
      "step": 32010
    },
    {
      "epoch": 15.094295143800094,
      "grad_norm": 0.011758270673453808,
      "learning_rate": 0.028550578676410976,
      "loss": 0.146,
      "num_input_tokens_seen": 27722592,
      "step": 32015
    },
    {
      "epoch": 15.096652522395097,
      "grad_norm": 0.005912365857511759,
      "learning_rate": 0.02851601705285837,
      "loss": 0.1376,
      "num_input_tokens_seen": 27727152,
      "step": 32020
    },
    {
      "epoch": 15.099009900990099,
      "grad_norm": 0.006046711932867765,
      "learning_rate": 0.028481474163662666,
      "loss": 0.1638,
      "num_input_tokens_seen": 27730688,
      "step": 32025
    },
    {
      "epoch": 15.101367279585101,
      "grad_norm": 0.007585310842841864,
      "learning_rate": 0.028446950014150683,
      "loss": 0.1222,
      "num_input_tokens_seen": 27736304,
      "step": 32030
    },
    {
      "epoch": 15.103724658180104,
      "grad_norm": 0.011047706007957458,
      "learning_rate": 0.028412444609646596,
      "loss": 0.208,
      "num_input_tokens_seen": 27740560,
      "step": 32035
    },
    {
      "epoch": 15.106082036775106,
      "grad_norm": 0.0077481744810938835,
      "learning_rate": 0.028377957955471465,
      "loss": 0.1224,
      "num_input_tokens_seen": 27745744,
      "step": 32040
    },
    {
      "epoch": 15.108439415370109,
      "grad_norm": 0.007697937544435263,
      "learning_rate": 0.0283434900569436,
      "loss": 0.1679,
      "num_input_tokens_seen": 27749568,
      "step": 32045
    },
    {
      "epoch": 15.110796793965111,
      "grad_norm": 0.00943811982870102,
      "learning_rate": 0.028309040919378456,
      "loss": 0.1606,
      "num_input_tokens_seen": 27754656,
      "step": 32050
    },
    {
      "epoch": 15.113154172560114,
      "grad_norm": 0.009678568691015244,
      "learning_rate": 0.02827461054808848,
      "loss": 0.1234,
      "num_input_tokens_seen": 27759248,
      "step": 32055
    },
    {
      "epoch": 15.115511551155116,
      "grad_norm": 0.013338626362383366,
      "learning_rate": 0.028240198948383186,
      "loss": 0.1475,
      "num_input_tokens_seen": 27763824,
      "step": 32060
    },
    {
      "epoch": 15.117868929750118,
      "grad_norm": 0.007759434171020985,
      "learning_rate": 0.028205806125569402,
      "loss": 0.1533,
      "num_input_tokens_seen": 27768528,
      "step": 32065
    },
    {
      "epoch": 15.12022630834512,
      "grad_norm": 0.005333269480615854,
      "learning_rate": 0.028171432084950834,
      "loss": 0.1024,
      "num_input_tokens_seen": 27772736,
      "step": 32070
    },
    {
      "epoch": 15.122583686940123,
      "grad_norm": 0.013363608159124851,
      "learning_rate": 0.028137076831828478,
      "loss": 0.2082,
      "num_input_tokens_seen": 27776800,
      "step": 32075
    },
    {
      "epoch": 15.124941065535126,
      "grad_norm": 0.014401925727725029,
      "learning_rate": 0.028102740371500238,
      "loss": 0.1827,
      "num_input_tokens_seen": 27781232,
      "step": 32080
    },
    {
      "epoch": 15.127298444130128,
      "grad_norm": 0.0119165712967515,
      "learning_rate": 0.0280684227092613,
      "loss": 0.1402,
      "num_input_tokens_seen": 27785968,
      "step": 32085
    },
    {
      "epoch": 15.12965582272513,
      "grad_norm": 0.00829023215919733,
      "learning_rate": 0.02803412385040392,
      "loss": 0.1435,
      "num_input_tokens_seen": 27791472,
      "step": 32090
    },
    {
      "epoch": 15.132013201320133,
      "grad_norm": 0.004531940910965204,
      "learning_rate": 0.027999843800217306,
      "loss": 0.13,
      "num_input_tokens_seen": 27796064,
      "step": 32095
    },
    {
      "epoch": 15.134370579915135,
      "grad_norm": 0.007127948570996523,
      "learning_rate": 0.027965582563987932,
      "loss": 0.129,
      "num_input_tokens_seen": 27801072,
      "step": 32100
    },
    {
      "epoch": 15.136727958510138,
      "grad_norm": 0.012917138636112213,
      "learning_rate": 0.027931340146999346,
      "loss": 0.1769,
      "num_input_tokens_seen": 27805344,
      "step": 32105
    },
    {
      "epoch": 15.13908533710514,
      "grad_norm": 0.01270768791437149,
      "learning_rate": 0.02789711655453208,
      "loss": 0.1435,
      "num_input_tokens_seen": 27808880,
      "step": 32110
    },
    {
      "epoch": 15.14144271570014,
      "grad_norm": 0.007484283298254013,
      "learning_rate": 0.02786291179186392,
      "loss": 0.1309,
      "num_input_tokens_seen": 27813120,
      "step": 32115
    },
    {
      "epoch": 15.143800094295143,
      "grad_norm": 0.0063737197779119015,
      "learning_rate": 0.02782872586426961,
      "loss": 0.1326,
      "num_input_tokens_seen": 27816880,
      "step": 32120
    },
    {
      "epoch": 15.146157472890145,
      "grad_norm": 0.011078761890530586,
      "learning_rate": 0.027794558777021083,
      "loss": 0.176,
      "num_input_tokens_seen": 27820976,
      "step": 32125
    },
    {
      "epoch": 15.148514851485148,
      "grad_norm": 0.010743359103798866,
      "learning_rate": 0.02776041053538734,
      "loss": 0.1915,
      "num_input_tokens_seen": 27824944,
      "step": 32130
    },
    {
      "epoch": 15.15087223008015,
      "grad_norm": 0.006372596602886915,
      "learning_rate": 0.027726281144634407,
      "loss": 0.1796,
      "num_input_tokens_seen": 27830144,
      "step": 32135
    },
    {
      "epoch": 15.153229608675153,
      "grad_norm": 0.0029856194742023945,
      "learning_rate": 0.02769217061002552,
      "loss": 0.1138,
      "num_input_tokens_seen": 27834496,
      "step": 32140
    },
    {
      "epoch": 15.155586987270155,
      "grad_norm": 0.009944400750100613,
      "learning_rate": 0.027658078936820967,
      "loss": 0.1804,
      "num_input_tokens_seen": 27838944,
      "step": 32145
    },
    {
      "epoch": 15.157944365865157,
      "grad_norm": 0.008034724742174149,
      "learning_rate": 0.02762400613027805,
      "loss": 0.143,
      "num_input_tokens_seen": 27842752,
      "step": 32150
    },
    {
      "epoch": 15.16030174446016,
      "grad_norm": 0.00780272064730525,
      "learning_rate": 0.027589952195651295,
      "loss": 0.1631,
      "num_input_tokens_seen": 27846848,
      "step": 32155
    },
    {
      "epoch": 15.162659123055162,
      "grad_norm": 0.00837457925081253,
      "learning_rate": 0.027555917138192186,
      "loss": 0.1466,
      "num_input_tokens_seen": 27850992,
      "step": 32160
    },
    {
      "epoch": 15.165016501650165,
      "grad_norm": 0.0036268997937440872,
      "learning_rate": 0.027521900963149375,
      "loss": 0.117,
      "num_input_tokens_seen": 27854992,
      "step": 32165
    },
    {
      "epoch": 15.167373880245167,
      "grad_norm": 0.008317976258695126,
      "learning_rate": 0.027487903675768633,
      "loss": 0.1558,
      "num_input_tokens_seen": 27859584,
      "step": 32170
    },
    {
      "epoch": 15.16973125884017,
      "grad_norm": 0.005387894809246063,
      "learning_rate": 0.027453925281292677,
      "loss": 0.1251,
      "num_input_tokens_seen": 27864304,
      "step": 32175
    },
    {
      "epoch": 15.172088637435172,
      "grad_norm": 0.010136988945305347,
      "learning_rate": 0.027419965784961475,
      "loss": 0.1449,
      "num_input_tokens_seen": 27868416,
      "step": 32180
    },
    {
      "epoch": 15.174446016030174,
      "grad_norm": 0.006001410540193319,
      "learning_rate": 0.027386025192012015,
      "loss": 0.075,
      "num_input_tokens_seen": 27873520,
      "step": 32185
    },
    {
      "epoch": 15.176803394625177,
      "grad_norm": 0.010405183769762516,
      "learning_rate": 0.027352103507678277,
      "loss": 0.1941,
      "num_input_tokens_seen": 27878128,
      "step": 32190
    },
    {
      "epoch": 15.179160773220179,
      "grad_norm": 0.005759806837886572,
      "learning_rate": 0.027318200737191527,
      "loss": 0.1769,
      "num_input_tokens_seen": 27881968,
      "step": 32195
    },
    {
      "epoch": 15.181518151815181,
      "grad_norm": 0.007023729383945465,
      "learning_rate": 0.027284316885779935,
      "loss": 0.119,
      "num_input_tokens_seen": 27886368,
      "step": 32200
    },
    {
      "epoch": 15.181518151815181,
      "eval_loss": 0.23457427322864532,
      "eval_runtime": 21.9257,
      "eval_samples_per_second": 43.009,
      "eval_steps_per_second": 21.527,
      "num_input_tokens_seen": 27886368,
      "step": 32200
    },
    {
      "epoch": 15.183875530410184,
      "grad_norm": 0.011687003076076508,
      "learning_rate": 0.027250451958668785,
      "loss": 0.1364,
      "num_input_tokens_seen": 27891920,
      "step": 32205
    },
    {
      "epoch": 15.186232909005186,
      "grad_norm": 0.012156837619841099,
      "learning_rate": 0.027216605961080536,
      "loss": 0.1366,
      "num_input_tokens_seen": 27896736,
      "step": 32210
    },
    {
      "epoch": 15.188590287600189,
      "grad_norm": 0.01599881239235401,
      "learning_rate": 0.02718277889823461,
      "loss": 0.1653,
      "num_input_tokens_seen": 27901440,
      "step": 32215
    },
    {
      "epoch": 15.190947666195191,
      "grad_norm": 0.00851558055728674,
      "learning_rate": 0.027148970775347604,
      "loss": 0.1242,
      "num_input_tokens_seen": 27906208,
      "step": 32220
    },
    {
      "epoch": 15.193305044790193,
      "grad_norm": 0.010315995663404465,
      "learning_rate": 0.027115181597633174,
      "loss": 0.1324,
      "num_input_tokens_seen": 27911120,
      "step": 32225
    },
    {
      "epoch": 15.195662423385196,
      "grad_norm": 0.007938310503959656,
      "learning_rate": 0.027081411370301976,
      "loss": 0.1201,
      "num_input_tokens_seen": 27915488,
      "step": 32230
    },
    {
      "epoch": 15.198019801980198,
      "grad_norm": 0.009246890433132648,
      "learning_rate": 0.027047660098561875,
      "loss": 0.1125,
      "num_input_tokens_seen": 27920528,
      "step": 32235
    },
    {
      "epoch": 15.2003771805752,
      "grad_norm": 0.012387375347316265,
      "learning_rate": 0.02701392778761766,
      "loss": 0.1834,
      "num_input_tokens_seen": 27925120,
      "step": 32240
    },
    {
      "epoch": 15.202734559170203,
      "grad_norm": 0.010826407000422478,
      "learning_rate": 0.02698021444267133,
      "loss": 0.1961,
      "num_input_tokens_seen": 27929072,
      "step": 32245
    },
    {
      "epoch": 15.205091937765205,
      "grad_norm": 0.009084058925509453,
      "learning_rate": 0.026946520068921915,
      "loss": 0.1845,
      "num_input_tokens_seen": 27932976,
      "step": 32250
    },
    {
      "epoch": 15.207449316360208,
      "grad_norm": 0.004440312273800373,
      "learning_rate": 0.02691284467156547,
      "loss": 0.0927,
      "num_input_tokens_seen": 27936912,
      "step": 32255
    },
    {
      "epoch": 15.20980669495521,
      "grad_norm": 0.005042764358222485,
      "learning_rate": 0.026879188255795182,
      "loss": 0.1043,
      "num_input_tokens_seen": 27941568,
      "step": 32260
    },
    {
      "epoch": 15.212164073550213,
      "grad_norm": 0.014170131646096706,
      "learning_rate": 0.026845550826801328,
      "loss": 0.1204,
      "num_input_tokens_seen": 27946512,
      "step": 32265
    },
    {
      "epoch": 15.214521452145215,
      "grad_norm": 0.0058250706642866135,
      "learning_rate": 0.02681193238977121,
      "loss": 0.0894,
      "num_input_tokens_seen": 27950832,
      "step": 32270
    },
    {
      "epoch": 15.216878830740217,
      "grad_norm": 0.01427895575761795,
      "learning_rate": 0.026778332949889145,
      "loss": 0.1988,
      "num_input_tokens_seen": 27955056,
      "step": 32275
    },
    {
      "epoch": 15.21923620933522,
      "grad_norm": 0.010294276289641857,
      "learning_rate": 0.026744752512336673,
      "loss": 0.1178,
      "num_input_tokens_seen": 27959280,
      "step": 32280
    },
    {
      "epoch": 15.221593587930222,
      "grad_norm": 0.008662639185786247,
      "learning_rate": 0.02671119108229225,
      "loss": 0.1252,
      "num_input_tokens_seen": 27963424,
      "step": 32285
    },
    {
      "epoch": 15.223950966525225,
      "grad_norm": 0.013079563155770302,
      "learning_rate": 0.026677648664931556,
      "loss": 0.1099,
      "num_input_tokens_seen": 27966928,
      "step": 32290
    },
    {
      "epoch": 15.226308345120227,
      "grad_norm": 0.02364530600607395,
      "learning_rate": 0.026644125265427154,
      "loss": 0.2131,
      "num_input_tokens_seen": 27971360,
      "step": 32295
    },
    {
      "epoch": 15.22866572371523,
      "grad_norm": 0.0031454747077077627,
      "learning_rate": 0.026610620888948822,
      "loss": 0.1227,
      "num_input_tokens_seen": 27975328,
      "step": 32300
    },
    {
      "epoch": 15.231023102310232,
      "grad_norm": 0.009328093379735947,
      "learning_rate": 0.026577135540663408,
      "loss": 0.1263,
      "num_input_tokens_seen": 27979840,
      "step": 32305
    },
    {
      "epoch": 15.233380480905234,
      "grad_norm": 0.008871783502399921,
      "learning_rate": 0.026543669225734673,
      "loss": 0.1742,
      "num_input_tokens_seen": 27984320,
      "step": 32310
    },
    {
      "epoch": 15.235737859500237,
      "grad_norm": 0.016636818647384644,
      "learning_rate": 0.02651022194932363,
      "loss": 0.1467,
      "num_input_tokens_seen": 27988320,
      "step": 32315
    },
    {
      "epoch": 15.238095238095237,
      "grad_norm": 0.01185997948050499,
      "learning_rate": 0.026476793716588194,
      "loss": 0.1516,
      "num_input_tokens_seen": 27992304,
      "step": 32320
    },
    {
      "epoch": 15.24045261669024,
      "grad_norm": 0.015266108326613903,
      "learning_rate": 0.026443384532683467,
      "loss": 0.255,
      "num_input_tokens_seen": 27996496,
      "step": 32325
    },
    {
      "epoch": 15.242809995285242,
      "grad_norm": 0.007138614542782307,
      "learning_rate": 0.026409994402761584,
      "loss": 0.1003,
      "num_input_tokens_seen": 28000576,
      "step": 32330
    },
    {
      "epoch": 15.245167373880244,
      "grad_norm": 0.007402620743960142,
      "learning_rate": 0.026376623331971653,
      "loss": 0.1026,
      "num_input_tokens_seen": 28005040,
      "step": 32335
    },
    {
      "epoch": 15.247524752475247,
      "grad_norm": 0.004511822015047073,
      "learning_rate": 0.026343271325459997,
      "loss": 0.1355,
      "num_input_tokens_seen": 28008960,
      "step": 32340
    },
    {
      "epoch": 15.24988213107025,
      "grad_norm": 0.007138450164347887,
      "learning_rate": 0.02630993838836987,
      "loss": 0.1322,
      "num_input_tokens_seen": 28013392,
      "step": 32345
    },
    {
      "epoch": 15.252239509665252,
      "grad_norm": 0.014267225749790668,
      "learning_rate": 0.026276624525841584,
      "loss": 0.1655,
      "num_input_tokens_seen": 28016720,
      "step": 32350
    },
    {
      "epoch": 15.254596888260254,
      "grad_norm": 0.01218926440924406,
      "learning_rate": 0.026243329743012637,
      "loss": 0.1167,
      "num_input_tokens_seen": 28021424,
      "step": 32355
    },
    {
      "epoch": 15.256954266855256,
      "grad_norm": 0.015085089951753616,
      "learning_rate": 0.026210054045017438,
      "loss": 0.1404,
      "num_input_tokens_seen": 28026336,
      "step": 32360
    },
    {
      "epoch": 15.259311645450259,
      "grad_norm": 0.005296699237078428,
      "learning_rate": 0.02617679743698755,
      "loss": 0.1618,
      "num_input_tokens_seen": 28030656,
      "step": 32365
    },
    {
      "epoch": 15.261669024045261,
      "grad_norm": 0.0031926780939102173,
      "learning_rate": 0.02614355992405158,
      "loss": 0.0835,
      "num_input_tokens_seen": 28035152,
      "step": 32370
    },
    {
      "epoch": 15.264026402640264,
      "grad_norm": 0.004572013393044472,
      "learning_rate": 0.026110341511335115,
      "loss": 0.1672,
      "num_input_tokens_seen": 28039632,
      "step": 32375
    },
    {
      "epoch": 15.266383781235266,
      "grad_norm": 0.006843676790595055,
      "learning_rate": 0.02607714220396093,
      "loss": 0.1061,
      "num_input_tokens_seen": 28043680,
      "step": 32380
    },
    {
      "epoch": 15.268741159830268,
      "grad_norm": 0.018468564376235008,
      "learning_rate": 0.02604396200704869,
      "loss": 0.2295,
      "num_input_tokens_seen": 28048432,
      "step": 32385
    },
    {
      "epoch": 15.27109853842527,
      "grad_norm": 0.011436278000473976,
      "learning_rate": 0.02601080092571523,
      "loss": 0.303,
      "num_input_tokens_seen": 28052128,
      "step": 32390
    },
    {
      "epoch": 15.273455917020273,
      "grad_norm": 0.012343303300440311,
      "learning_rate": 0.025977658965074455,
      "loss": 0.1626,
      "num_input_tokens_seen": 28056176,
      "step": 32395
    },
    {
      "epoch": 15.275813295615276,
      "grad_norm": 0.004792632535099983,
      "learning_rate": 0.02594453613023719,
      "loss": 0.1075,
      "num_input_tokens_seen": 28061984,
      "step": 32400
    },
    {
      "epoch": 15.275813295615276,
      "eval_loss": 0.23268653452396393,
      "eval_runtime": 21.8986,
      "eval_samples_per_second": 43.062,
      "eval_steps_per_second": 21.554,
      "num_input_tokens_seen": 28061984,
      "step": 32400
    },
    {
      "epoch": 15.278170674210278,
      "grad_norm": 0.010497204028069973,
      "learning_rate": 0.025911432426311443,
      "loss": 0.1392,
      "num_input_tokens_seen": 28066528,
      "step": 32405
    },
    {
      "epoch": 15.28052805280528,
      "grad_norm": 0.0029115735087543726,
      "learning_rate": 0.025878347858402234,
      "loss": 0.1236,
      "num_input_tokens_seen": 28070688,
      "step": 32410
    },
    {
      "epoch": 15.282885431400283,
      "grad_norm": 0.013748629949986935,
      "learning_rate": 0.025845282431611598,
      "loss": 0.2091,
      "num_input_tokens_seen": 28074704,
      "step": 32415
    },
    {
      "epoch": 15.285242809995285,
      "grad_norm": 0.018689455464482307,
      "learning_rate": 0.025812236151038608,
      "loss": 0.1809,
      "num_input_tokens_seen": 28078816,
      "step": 32420
    },
    {
      "epoch": 15.287600188590288,
      "grad_norm": 0.009783794172108173,
      "learning_rate": 0.025779209021779468,
      "loss": 0.0955,
      "num_input_tokens_seen": 28082736,
      "step": 32425
    },
    {
      "epoch": 15.28995756718529,
      "grad_norm": 0.008072003722190857,
      "learning_rate": 0.025746201048927324,
      "loss": 0.1087,
      "num_input_tokens_seen": 28087248,
      "step": 32430
    },
    {
      "epoch": 15.292314945780292,
      "grad_norm": 0.005828424356877804,
      "learning_rate": 0.025713212237572485,
      "loss": 0.0935,
      "num_input_tokens_seen": 28091744,
      "step": 32435
    },
    {
      "epoch": 15.294672324375295,
      "grad_norm": 0.013339896686375141,
      "learning_rate": 0.025680242592802164,
      "loss": 0.2137,
      "num_input_tokens_seen": 28095360,
      "step": 32440
    },
    {
      "epoch": 15.297029702970297,
      "grad_norm": 0.011880171485245228,
      "learning_rate": 0.02564729211970073,
      "loss": 0.1427,
      "num_input_tokens_seen": 28099888,
      "step": 32445
    },
    {
      "epoch": 15.2993870815653,
      "grad_norm": 0.005814662203192711,
      "learning_rate": 0.025614360823349617,
      "loss": 0.1569,
      "num_input_tokens_seen": 28103472,
      "step": 32450
    },
    {
      "epoch": 15.301744460160302,
      "grad_norm": 0.016350964084267616,
      "learning_rate": 0.025581448708827146,
      "loss": 0.1274,
      "num_input_tokens_seen": 28107680,
      "step": 32455
    },
    {
      "epoch": 15.304101838755304,
      "grad_norm": 0.01193738728761673,
      "learning_rate": 0.025548555781208876,
      "loss": 0.1875,
      "num_input_tokens_seen": 28111872,
      "step": 32460
    },
    {
      "epoch": 15.306459217350307,
      "grad_norm": 0.008230690844357014,
      "learning_rate": 0.02551568204556721,
      "loss": 0.1123,
      "num_input_tokens_seen": 28117280,
      "step": 32465
    },
    {
      "epoch": 15.30881659594531,
      "grad_norm": 0.007615272421389818,
      "learning_rate": 0.02548282750697173,
      "loss": 0.1059,
      "num_input_tokens_seen": 28121136,
      "step": 32470
    },
    {
      "epoch": 15.311173974540312,
      "grad_norm": 0.00945815909653902,
      "learning_rate": 0.02544999217048909,
      "loss": 0.2026,
      "num_input_tokens_seen": 28127248,
      "step": 32475
    },
    {
      "epoch": 15.313531353135314,
      "grad_norm": 0.013622204773128033,
      "learning_rate": 0.025417176041182793,
      "loss": 0.2165,
      "num_input_tokens_seen": 28131184,
      "step": 32480
    },
    {
      "epoch": 15.315888731730317,
      "grad_norm": 0.010845323093235493,
      "learning_rate": 0.025384379124113596,
      "loss": 0.1234,
      "num_input_tokens_seen": 28135248,
      "step": 32485
    },
    {
      "epoch": 15.318246110325319,
      "grad_norm": 0.011358315125107765,
      "learning_rate": 0.025351601424339124,
      "loss": 0.1676,
      "num_input_tokens_seen": 28139424,
      "step": 32490
    },
    {
      "epoch": 15.320603488920321,
      "grad_norm": 0.009925664402544498,
      "learning_rate": 0.025318842946914184,
      "loss": 0.1459,
      "num_input_tokens_seen": 28143776,
      "step": 32495
    },
    {
      "epoch": 15.322960867515324,
      "grad_norm": 0.005954282823950052,
      "learning_rate": 0.025286103696890494,
      "loss": 0.1412,
      "num_input_tokens_seen": 28148032,
      "step": 32500
    },
    {
      "epoch": 15.325318246110326,
      "grad_norm": 0.004367055371403694,
      "learning_rate": 0.025253383679316836,
      "loss": 0.1272,
      "num_input_tokens_seen": 28151888,
      "step": 32505
    },
    {
      "epoch": 15.327675624705329,
      "grad_norm": 0.008908172138035297,
      "learning_rate": 0.025220682899239077,
      "loss": 0.1631,
      "num_input_tokens_seen": 28155760,
      "step": 32510
    },
    {
      "epoch": 15.33003300330033,
      "grad_norm": 0.011793509125709534,
      "learning_rate": 0.02518800136170013,
      "loss": 0.1322,
      "num_input_tokens_seen": 28160560,
      "step": 32515
    },
    {
      "epoch": 15.332390381895332,
      "grad_norm": 0.00810137391090393,
      "learning_rate": 0.02515533907173981,
      "loss": 0.1744,
      "num_input_tokens_seen": 28164672,
      "step": 32520
    },
    {
      "epoch": 15.334747760490334,
      "grad_norm": 0.003926781006157398,
      "learning_rate": 0.025122696034395115,
      "loss": 0.1067,
      "num_input_tokens_seen": 28169056,
      "step": 32525
    },
    {
      "epoch": 15.337105139085336,
      "grad_norm": 0.008796206675469875,
      "learning_rate": 0.025090072254700023,
      "loss": 0.143,
      "num_input_tokens_seen": 28173040,
      "step": 32530
    },
    {
      "epoch": 15.339462517680339,
      "grad_norm": 0.009944225661456585,
      "learning_rate": 0.025057467737685468,
      "loss": 0.1069,
      "num_input_tokens_seen": 28176736,
      "step": 32535
    },
    {
      "epoch": 15.341819896275341,
      "grad_norm": 0.01030266098678112,
      "learning_rate": 0.025024882488379557,
      "loss": 0.1501,
      "num_input_tokens_seen": 28182240,
      "step": 32540
    },
    {
      "epoch": 15.344177274870344,
      "grad_norm": 0.013475790619850159,
      "learning_rate": 0.02499231651180727,
      "loss": 0.12,
      "num_input_tokens_seen": 28186784,
      "step": 32545
    },
    {
      "epoch": 15.346534653465346,
      "grad_norm": 0.012259495444595814,
      "learning_rate": 0.024959769812990713,
      "loss": 0.2102,
      "num_input_tokens_seen": 28191776,
      "step": 32550
    },
    {
      "epoch": 15.348892032060348,
      "grad_norm": 0.0028135532047599554,
      "learning_rate": 0.024927242396949045,
      "loss": 0.0531,
      "num_input_tokens_seen": 28195568,
      "step": 32555
    },
    {
      "epoch": 15.35124941065535,
      "grad_norm": 0.015003572218120098,
      "learning_rate": 0.02489473426869836,
      "loss": 0.1547,
      "num_input_tokens_seen": 28199488,
      "step": 32560
    },
    {
      "epoch": 15.353606789250353,
      "grad_norm": 0.011886035092175007,
      "learning_rate": 0.024862245433251776,
      "loss": 0.1522,
      "num_input_tokens_seen": 28204384,
      "step": 32565
    },
    {
      "epoch": 15.355964167845356,
      "grad_norm": 0.005336808506399393,
      "learning_rate": 0.024829775895619577,
      "loss": 0.176,
      "num_input_tokens_seen": 28208464,
      "step": 32570
    },
    {
      "epoch": 15.358321546440358,
      "grad_norm": 0.008342361077666283,
      "learning_rate": 0.024797325660808882,
      "loss": 0.1445,
      "num_input_tokens_seen": 28212000,
      "step": 32575
    },
    {
      "epoch": 15.36067892503536,
      "grad_norm": 0.00827929936349392,
      "learning_rate": 0.02476489473382401,
      "loss": 0.1003,
      "num_input_tokens_seen": 28216064,
      "step": 32580
    },
    {
      "epoch": 15.363036303630363,
      "grad_norm": 0.015575846657156944,
      "learning_rate": 0.024732483119666127,
      "loss": 0.1585,
      "num_input_tokens_seen": 28220320,
      "step": 32585
    },
    {
      "epoch": 15.365393682225365,
      "grad_norm": 0.006159981247037649,
      "learning_rate": 0.024700090823333548,
      "loss": 0.1183,
      "num_input_tokens_seen": 28224240,
      "step": 32590
    },
    {
      "epoch": 15.367751060820368,
      "grad_norm": 0.006392660550773144,
      "learning_rate": 0.02466771784982163,
      "loss": 0.1169,
      "num_input_tokens_seen": 28229184,
      "step": 32595
    },
    {
      "epoch": 15.37010843941537,
      "grad_norm": 0.015322172082960606,
      "learning_rate": 0.024635364204122594,
      "loss": 0.1668,
      "num_input_tokens_seen": 28233360,
      "step": 32600
    },
    {
      "epoch": 15.37010843941537,
      "eval_loss": 0.23528781533241272,
      "eval_runtime": 21.9257,
      "eval_samples_per_second": 43.009,
      "eval_steps_per_second": 21.527,
      "num_input_tokens_seen": 28233360,
      "step": 32600
    },
    {
      "epoch": 15.372465818010372,
      "grad_norm": 0.008008717559278011,
      "learning_rate": 0.024603029891225852,
      "loss": 0.1526,
      "num_input_tokens_seen": 28237472,
      "step": 32605
    },
    {
      "epoch": 15.374823196605375,
      "grad_norm": 0.018177973106503487,
      "learning_rate": 0.024570714916117748,
      "loss": 0.1215,
      "num_input_tokens_seen": 28242368,
      "step": 32610
    },
    {
      "epoch": 15.377180575200377,
      "grad_norm": 0.01232017669826746,
      "learning_rate": 0.024538419283781625,
      "loss": 0.1661,
      "num_input_tokens_seen": 28247520,
      "step": 32615
    },
    {
      "epoch": 15.37953795379538,
      "grad_norm": 0.012947475537657738,
      "learning_rate": 0.024506142999197938,
      "loss": 0.2171,
      "num_input_tokens_seen": 28251488,
      "step": 32620
    },
    {
      "epoch": 15.381895332390382,
      "grad_norm": 0.017344379797577858,
      "learning_rate": 0.024473886067344002,
      "loss": 0.1159,
      "num_input_tokens_seen": 28256240,
      "step": 32625
    },
    {
      "epoch": 15.384252710985384,
      "grad_norm": 0.010178913362324238,
      "learning_rate": 0.02444164849319434,
      "loss": 0.1642,
      "num_input_tokens_seen": 28260000,
      "step": 32630
    },
    {
      "epoch": 15.386610089580387,
      "grad_norm": 0.010834285989403725,
      "learning_rate": 0.024409430281720306,
      "loss": 0.1919,
      "num_input_tokens_seen": 28264800,
      "step": 32635
    },
    {
      "epoch": 15.38896746817539,
      "grad_norm": 0.008635144680738449,
      "learning_rate": 0.024377231437890428,
      "loss": 0.151,
      "num_input_tokens_seen": 28269808,
      "step": 32640
    },
    {
      "epoch": 15.391324846770392,
      "grad_norm": 0.011760484427213669,
      "learning_rate": 0.024345051966670115,
      "loss": 0.1578,
      "num_input_tokens_seen": 28273936,
      "step": 32645
    },
    {
      "epoch": 15.393682225365394,
      "grad_norm": 0.011929628439247608,
      "learning_rate": 0.024312891873021884,
      "loss": 0.2178,
      "num_input_tokens_seen": 28278048,
      "step": 32650
    },
    {
      "epoch": 15.396039603960396,
      "grad_norm": 0.008102137595415115,
      "learning_rate": 0.024280751161905183,
      "loss": 0.1768,
      "num_input_tokens_seen": 28282160,
      "step": 32655
    },
    {
      "epoch": 15.398396982555399,
      "grad_norm": 0.014904254116117954,
      "learning_rate": 0.02424862983827658,
      "loss": 0.1655,
      "num_input_tokens_seen": 28286256,
      "step": 32660
    },
    {
      "epoch": 15.400754361150401,
      "grad_norm": 0.010600886307656765,
      "learning_rate": 0.024216527907089495,
      "loss": 0.1873,
      "num_input_tokens_seen": 28290256,
      "step": 32665
    },
    {
      "epoch": 15.403111739745404,
      "grad_norm": 0.008769355714321136,
      "learning_rate": 0.024184445373294505,
      "loss": 0.1756,
      "num_input_tokens_seen": 28295280,
      "step": 32670
    },
    {
      "epoch": 15.405469118340406,
      "grad_norm": 0.009978369809687138,
      "learning_rate": 0.02415238224183918,
      "loss": 0.1877,
      "num_input_tokens_seen": 28299328,
      "step": 32675
    },
    {
      "epoch": 15.407826496935408,
      "grad_norm": 0.01037834957242012,
      "learning_rate": 0.024120338517667973,
      "loss": 0.1389,
      "num_input_tokens_seen": 28303296,
      "step": 32680
    },
    {
      "epoch": 15.41018387553041,
      "grad_norm": 0.004951676353812218,
      "learning_rate": 0.02408831420572247,
      "loss": 0.1269,
      "num_input_tokens_seen": 28307712,
      "step": 32685
    },
    {
      "epoch": 15.412541254125413,
      "grad_norm": 0.0045358226634562016,
      "learning_rate": 0.024056309310941264,
      "loss": 0.0969,
      "num_input_tokens_seen": 28311696,
      "step": 32690
    },
    {
      "epoch": 15.414898632720416,
      "grad_norm": 0.0031718246173113585,
      "learning_rate": 0.02402432383825982,
      "loss": 0.1114,
      "num_input_tokens_seen": 28316256,
      "step": 32695
    },
    {
      "epoch": 15.417256011315418,
      "grad_norm": 0.020901277661323547,
      "learning_rate": 0.023992357792610792,
      "loss": 0.1903,
      "num_input_tokens_seen": 28320960,
      "step": 32700
    },
    {
      "epoch": 15.41961338991042,
      "grad_norm": 0.012570049613714218,
      "learning_rate": 0.0239604111789237,
      "loss": 0.1551,
      "num_input_tokens_seen": 28325232,
      "step": 32705
    },
    {
      "epoch": 15.421970768505423,
      "grad_norm": 0.0062584588304162025,
      "learning_rate": 0.023928484002125095,
      "loss": 0.1091,
      "num_input_tokens_seen": 28329472,
      "step": 32710
    },
    {
      "epoch": 15.424328147100425,
      "grad_norm": 0.007957147434353828,
      "learning_rate": 0.023896576267138595,
      "loss": 0.1437,
      "num_input_tokens_seen": 28333232,
      "step": 32715
    },
    {
      "epoch": 15.426685525695426,
      "grad_norm": 0.014776704832911491,
      "learning_rate": 0.02386468797888471,
      "loss": 0.1503,
      "num_input_tokens_seen": 28337312,
      "step": 32720
    },
    {
      "epoch": 15.429042904290428,
      "grad_norm": 0.012422189116477966,
      "learning_rate": 0.023832819142281057,
      "loss": 0.2676,
      "num_input_tokens_seen": 28342336,
      "step": 32725
    },
    {
      "epoch": 15.43140028288543,
      "grad_norm": 0.008457721211016178,
      "learning_rate": 0.02380096976224225,
      "loss": 0.165,
      "num_input_tokens_seen": 28347264,
      "step": 32730
    },
    {
      "epoch": 15.433757661480433,
      "grad_norm": 0.014382749795913696,
      "learning_rate": 0.023769139843679777,
      "loss": 0.1552,
      "num_input_tokens_seen": 28350816,
      "step": 32735
    },
    {
      "epoch": 15.436115040075435,
      "grad_norm": 0.006935228127986193,
      "learning_rate": 0.023737329391502287,
      "loss": 0.1641,
      "num_input_tokens_seen": 28355712,
      "step": 32740
    },
    {
      "epoch": 15.438472418670438,
      "grad_norm": 0.011144048534333706,
      "learning_rate": 0.023705538410615293,
      "loss": 0.2076,
      "num_input_tokens_seen": 28359856,
      "step": 32745
    },
    {
      "epoch": 15.44082979726544,
      "grad_norm": 0.012775902636349201,
      "learning_rate": 0.023673766905921396,
      "loss": 0.1565,
      "num_input_tokens_seen": 28364464,
      "step": 32750
    },
    {
      "epoch": 15.443187175860443,
      "grad_norm": 0.0087564866989851,
      "learning_rate": 0.0236420148823202,
      "loss": 0.1157,
      "num_input_tokens_seen": 28368704,
      "step": 32755
    },
    {
      "epoch": 15.445544554455445,
      "grad_norm": 0.007694755215197802,
      "learning_rate": 0.02361028234470816,
      "loss": 0.1123,
      "num_input_tokens_seen": 28374096,
      "step": 32760
    },
    {
      "epoch": 15.447901933050447,
      "grad_norm": 0.016539813950657845,
      "learning_rate": 0.023578569297978913,
      "loss": 0.1311,
      "num_input_tokens_seen": 28380000,
      "step": 32765
    },
    {
      "epoch": 15.45025931164545,
      "grad_norm": 0.0044408938847482204,
      "learning_rate": 0.023546875747023025,
      "loss": 0.1436,
      "num_input_tokens_seen": 28383792,
      "step": 32770
    },
    {
      "epoch": 15.452616690240452,
      "grad_norm": 0.015383691526949406,
      "learning_rate": 0.02351520169672801,
      "loss": 0.1408,
      "num_input_tokens_seen": 28388000,
      "step": 32775
    },
    {
      "epoch": 15.454974068835455,
      "grad_norm": 0.013608130626380444,
      "learning_rate": 0.023483547151978357,
      "loss": 0.2037,
      "num_input_tokens_seen": 28392400,
      "step": 32780
    },
    {
      "epoch": 15.457331447430457,
      "grad_norm": 0.008210189640522003,
      "learning_rate": 0.023451912117655675,
      "loss": 0.1,
      "num_input_tokens_seen": 28396736,
      "step": 32785
    },
    {
      "epoch": 15.45968882602546,
      "grad_norm": 0.014253182336688042,
      "learning_rate": 0.023420296598638417,
      "loss": 0.1575,
      "num_input_tokens_seen": 28401616,
      "step": 32790
    },
    {
      "epoch": 15.462046204620462,
      "grad_norm": 0.013996430672705173,
      "learning_rate": 0.023388700599802165,
      "loss": 0.199,
      "num_input_tokens_seen": 28405552,
      "step": 32795
    },
    {
      "epoch": 15.464403583215464,
      "grad_norm": 0.0024401608388870955,
      "learning_rate": 0.023357124126019334,
      "loss": 0.1905,
      "num_input_tokens_seen": 28411200,
      "step": 32800
    },
    {
      "epoch": 15.464403583215464,
      "eval_loss": 0.22916005551815033,
      "eval_runtime": 21.9385,
      "eval_samples_per_second": 42.984,
      "eval_steps_per_second": 21.515,
      "num_input_tokens_seen": 28411200,
      "step": 32800
    },
    {
      "epoch": 15.466760961810467,
      "grad_norm": 0.01107438188046217,
      "learning_rate": 0.02332556718215945,
      "loss": 0.1225,
      "num_input_tokens_seen": 28414832,
      "step": 32805
    },
    {
      "epoch": 15.469118340405469,
      "grad_norm": 0.012400981970131397,
      "learning_rate": 0.023294029773089035,
      "loss": 0.1532,
      "num_input_tokens_seen": 28419440,
      "step": 32810
    },
    {
      "epoch": 15.471475719000471,
      "grad_norm": 0.015646561980247498,
      "learning_rate": 0.023262511903671484,
      "loss": 0.1513,
      "num_input_tokens_seen": 28423136,
      "step": 32815
    },
    {
      "epoch": 15.473833097595474,
      "grad_norm": 0.008530253544449806,
      "learning_rate": 0.023231013578767324,
      "loss": 0.1741,
      "num_input_tokens_seen": 28427616,
      "step": 32820
    },
    {
      "epoch": 15.476190476190476,
      "grad_norm": 0.00967363826930523,
      "learning_rate": 0.0231995348032339,
      "loss": 0.1868,
      "num_input_tokens_seen": 28432768,
      "step": 32825
    },
    {
      "epoch": 15.478547854785479,
      "grad_norm": 0.011692674830555916,
      "learning_rate": 0.023168075581925685,
      "loss": 0.2336,
      "num_input_tokens_seen": 28437488,
      "step": 32830
    },
    {
      "epoch": 15.480905233380481,
      "grad_norm": 0.01673063263297081,
      "learning_rate": 0.023136635919694126,
      "loss": 0.204,
      "num_input_tokens_seen": 28442176,
      "step": 32835
    },
    {
      "epoch": 15.483262611975483,
      "grad_norm": 0.012262137606739998,
      "learning_rate": 0.02310521582138753,
      "loss": 0.2024,
      "num_input_tokens_seen": 28446512,
      "step": 32840
    },
    {
      "epoch": 15.485619990570486,
      "grad_norm": 0.006570841651409864,
      "learning_rate": 0.023073815291851357,
      "loss": 0.1525,
      "num_input_tokens_seen": 28450288,
      "step": 32845
    },
    {
      "epoch": 15.487977369165488,
      "grad_norm": 0.002762504620477557,
      "learning_rate": 0.02304243433592788,
      "loss": 0.0825,
      "num_input_tokens_seen": 28454208,
      "step": 32850
    },
    {
      "epoch": 15.49033474776049,
      "grad_norm": 0.007456216495484114,
      "learning_rate": 0.023011072958456513,
      "loss": 0.1826,
      "num_input_tokens_seen": 28458736,
      "step": 32855
    },
    {
      "epoch": 15.492692126355493,
      "grad_norm": 0.0064378040842711926,
      "learning_rate": 0.022979731164273536,
      "loss": 0.1267,
      "num_input_tokens_seen": 28462528,
      "step": 32860
    },
    {
      "epoch": 15.495049504950495,
      "grad_norm": 0.008060024119913578,
      "learning_rate": 0.022948408958212218,
      "loss": 0.1064,
      "num_input_tokens_seen": 28467744,
      "step": 32865
    },
    {
      "epoch": 15.497406883545498,
      "grad_norm": 0.006059776991605759,
      "learning_rate": 0.022917106345102876,
      "loss": 0.1255,
      "num_input_tokens_seen": 28471824,
      "step": 32870
    },
    {
      "epoch": 15.4997642621405,
      "grad_norm": 0.006881760898977518,
      "learning_rate": 0.022885823329772785,
      "loss": 0.137,
      "num_input_tokens_seen": 28476288,
      "step": 32875
    },
    {
      "epoch": 15.502121640735503,
      "grad_norm": 0.022588232532143593,
      "learning_rate": 0.02285455991704612,
      "loss": 0.2013,
      "num_input_tokens_seen": 28481152,
      "step": 32880
    },
    {
      "epoch": 15.504479019330505,
      "grad_norm": 0.009498757310211658,
      "learning_rate": 0.022823316111744117,
      "loss": 0.1672,
      "num_input_tokens_seen": 28485920,
      "step": 32885
    },
    {
      "epoch": 15.506836397925507,
      "grad_norm": 0.008905145339667797,
      "learning_rate": 0.022792091918685014,
      "loss": 0.1441,
      "num_input_tokens_seen": 28490096,
      "step": 32890
    },
    {
      "epoch": 15.50919377652051,
      "grad_norm": 0.01259632222354412,
      "learning_rate": 0.022760887342683906,
      "loss": 0.1796,
      "num_input_tokens_seen": 28493312,
      "step": 32895
    },
    {
      "epoch": 15.511551155115512,
      "grad_norm": 0.006613368634134531,
      "learning_rate": 0.022729702388552975,
      "loss": 0.1888,
      "num_input_tokens_seen": 28497328,
      "step": 32900
    },
    {
      "epoch": 15.513908533710515,
      "grad_norm": 0.009593715891242027,
      "learning_rate": 0.022698537061101292,
      "loss": 0.1592,
      "num_input_tokens_seen": 28501648,
      "step": 32905
    },
    {
      "epoch": 15.516265912305517,
      "grad_norm": 0.01375671848654747,
      "learning_rate": 0.022667391365134962,
      "loss": 0.1351,
      "num_input_tokens_seen": 28505808,
      "step": 32910
    },
    {
      "epoch": 15.518623290900518,
      "grad_norm": 0.011229746043682098,
      "learning_rate": 0.022636265305457065,
      "loss": 0.1989,
      "num_input_tokens_seen": 28510592,
      "step": 32915
    },
    {
      "epoch": 15.520980669495522,
      "grad_norm": 0.00799809955060482,
      "learning_rate": 0.02260515888686764,
      "loss": 0.1701,
      "num_input_tokens_seen": 28514336,
      "step": 32920
    },
    {
      "epoch": 15.523338048090523,
      "grad_norm": 0.010452601127326488,
      "learning_rate": 0.022574072114163596,
      "loss": 0.1602,
      "num_input_tokens_seen": 28518512,
      "step": 32925
    },
    {
      "epoch": 15.525695426685525,
      "grad_norm": 0.0068704853765666485,
      "learning_rate": 0.022543004992139005,
      "loss": 0.1985,
      "num_input_tokens_seen": 28522544,
      "step": 32930
    },
    {
      "epoch": 15.528052805280527,
      "grad_norm": 0.0053809937089681625,
      "learning_rate": 0.022511957525584745,
      "loss": 0.1423,
      "num_input_tokens_seen": 28526352,
      "step": 32935
    },
    {
      "epoch": 15.53041018387553,
      "grad_norm": 0.009612223133444786,
      "learning_rate": 0.022480929719288778,
      "loss": 0.1964,
      "num_input_tokens_seen": 28531200,
      "step": 32940
    },
    {
      "epoch": 15.532767562470532,
      "grad_norm": 0.0064047579653561115,
      "learning_rate": 0.02244992157803592,
      "loss": 0.1792,
      "num_input_tokens_seen": 28535616,
      "step": 32945
    },
    {
      "epoch": 15.535124941065535,
      "grad_norm": 0.007627466693520546,
      "learning_rate": 0.022418933106608047,
      "loss": 0.1245,
      "num_input_tokens_seen": 28540560,
      "step": 32950
    },
    {
      "epoch": 15.537482319660537,
      "grad_norm": 0.005930561572313309,
      "learning_rate": 0.022387964309784018,
      "loss": 0.1853,
      "num_input_tokens_seen": 28544432,
      "step": 32955
    },
    {
      "epoch": 15.53983969825554,
      "grad_norm": 0.010000316426157951,
      "learning_rate": 0.022357015192339517,
      "loss": 0.2115,
      "num_input_tokens_seen": 28548528,
      "step": 32960
    },
    {
      "epoch": 15.542197076850542,
      "grad_norm": 0.0052168527618050575,
      "learning_rate": 0.02232608575904734,
      "loss": 0.1203,
      "num_input_tokens_seen": 28552448,
      "step": 32965
    },
    {
      "epoch": 15.544554455445544,
      "grad_norm": 0.010701922699809074,
      "learning_rate": 0.022295176014677225,
      "loss": 0.1806,
      "num_input_tokens_seen": 28557136,
      "step": 32970
    },
    {
      "epoch": 15.546911834040547,
      "grad_norm": 0.007220966275781393,
      "learning_rate": 0.02226428596399577,
      "loss": 0.1931,
      "num_input_tokens_seen": 28561424,
      "step": 32975
    },
    {
      "epoch": 15.549269212635549,
      "grad_norm": 0.006489204242825508,
      "learning_rate": 0.02223341561176669,
      "loss": 0.0887,
      "num_input_tokens_seen": 28565792,
      "step": 32980
    },
    {
      "epoch": 15.551626591230551,
      "grad_norm": 0.006054729223251343,
      "learning_rate": 0.0222025649627505,
      "loss": 0.1257,
      "num_input_tokens_seen": 28569376,
      "step": 32985
    },
    {
      "epoch": 15.553983969825554,
      "grad_norm": 0.004876998718827963,
      "learning_rate": 0.022171734021704814,
      "loss": 0.0981,
      "num_input_tokens_seen": 28573824,
      "step": 32990
    },
    {
      "epoch": 15.556341348420556,
      "grad_norm": 0.012059914879500866,
      "learning_rate": 0.022140922793384116,
      "loss": 0.1754,
      "num_input_tokens_seen": 28577872,
      "step": 32995
    },
    {
      "epoch": 15.558698727015559,
      "grad_norm": 0.003697044914588332,
      "learning_rate": 0.022110131282539934,
      "loss": 0.1303,
      "num_input_tokens_seen": 28582944,
      "step": 33000
    },
    {
      "epoch": 15.558698727015559,
      "eval_loss": 0.22632470726966858,
      "eval_runtime": 21.9125,
      "eval_samples_per_second": 43.035,
      "eval_steps_per_second": 21.54,
      "num_input_tokens_seen": 28582944,
      "step": 33000
    },
    {
      "epoch": 15.561056105610561,
      "grad_norm": 0.005530308000743389,
      "learning_rate": 0.022079359493920675,
      "loss": 0.1171,
      "num_input_tokens_seen": 28587136,
      "step": 33005
    },
    {
      "epoch": 15.563413484205563,
      "grad_norm": 0.014832360669970512,
      "learning_rate": 0.02204860743227169,
      "loss": 0.1795,
      "num_input_tokens_seen": 28591248,
      "step": 33010
    },
    {
      "epoch": 15.565770862800566,
      "grad_norm": 0.013271516188979149,
      "learning_rate": 0.022017875102335365,
      "loss": 0.1751,
      "num_input_tokens_seen": 28595344,
      "step": 33015
    },
    {
      "epoch": 15.568128241395568,
      "grad_norm": 0.009843386709690094,
      "learning_rate": 0.02198716250885108,
      "loss": 0.1233,
      "num_input_tokens_seen": 28600016,
      "step": 33020
    },
    {
      "epoch": 15.57048561999057,
      "grad_norm": 0.009697039611637592,
      "learning_rate": 0.021956469656555,
      "loss": 0.1388,
      "num_input_tokens_seen": 28603984,
      "step": 33025
    },
    {
      "epoch": 15.572842998585573,
      "grad_norm": 0.014772730879485607,
      "learning_rate": 0.0219257965501804,
      "loss": 0.201,
      "num_input_tokens_seen": 28608672,
      "step": 33030
    },
    {
      "epoch": 15.575200377180575,
      "grad_norm": 0.007973120547831059,
      "learning_rate": 0.021895143194457494,
      "loss": 0.134,
      "num_input_tokens_seen": 28613008,
      "step": 33035
    },
    {
      "epoch": 15.577557755775578,
      "grad_norm": 0.009184773080050945,
      "learning_rate": 0.021864509594113322,
      "loss": 0.1354,
      "num_input_tokens_seen": 28617552,
      "step": 33040
    },
    {
      "epoch": 15.57991513437058,
      "grad_norm": 0.010211185552179813,
      "learning_rate": 0.02183389575387207,
      "loss": 0.1181,
      "num_input_tokens_seen": 28621200,
      "step": 33045
    },
    {
      "epoch": 15.582272512965583,
      "grad_norm": 0.0062277293764054775,
      "learning_rate": 0.021803301678454682,
      "loss": 0.1405,
      "num_input_tokens_seen": 28625152,
      "step": 33050
    },
    {
      "epoch": 15.584629891560585,
      "grad_norm": 0.011773655191063881,
      "learning_rate": 0.021772727372579213,
      "loss": 0.2368,
      "num_input_tokens_seen": 28629168,
      "step": 33055
    },
    {
      "epoch": 15.586987270155587,
      "grad_norm": 0.008062131702899933,
      "learning_rate": 0.02174217284096061,
      "loss": 0.1005,
      "num_input_tokens_seen": 28633808,
      "step": 33060
    },
    {
      "epoch": 15.58934464875059,
      "grad_norm": 0.004558291286230087,
      "learning_rate": 0.0217116380883107,
      "loss": 0.1445,
      "num_input_tokens_seen": 28637760,
      "step": 33065
    },
    {
      "epoch": 15.591702027345592,
      "grad_norm": 0.008162260986864567,
      "learning_rate": 0.021681123119338425,
      "loss": 0.191,
      "num_input_tokens_seen": 28642960,
      "step": 33070
    },
    {
      "epoch": 15.594059405940595,
      "grad_norm": 0.007699870970100164,
      "learning_rate": 0.02165062793874951,
      "loss": 0.096,
      "num_input_tokens_seen": 28646848,
      "step": 33075
    },
    {
      "epoch": 15.596416784535597,
      "grad_norm": 0.004813583567738533,
      "learning_rate": 0.021620152551246666,
      "loss": 0.1478,
      "num_input_tokens_seen": 28651952,
      "step": 33080
    },
    {
      "epoch": 15.5987741631306,
      "grad_norm": 0.00886498298496008,
      "learning_rate": 0.02158969696152967,
      "loss": 0.1688,
      "num_input_tokens_seen": 28656064,
      "step": 33085
    },
    {
      "epoch": 15.601131541725602,
      "grad_norm": 0.011855809018015862,
      "learning_rate": 0.021559261174295057,
      "loss": 0.1867,
      "num_input_tokens_seen": 28660832,
      "step": 33090
    },
    {
      "epoch": 15.603488920320604,
      "grad_norm": 0.014365799725055695,
      "learning_rate": 0.02152884519423646,
      "loss": 0.1611,
      "num_input_tokens_seen": 28665184,
      "step": 33095
    },
    {
      "epoch": 15.605846298915607,
      "grad_norm": 0.007142855320125818,
      "learning_rate": 0.021498449026044447,
      "loss": 0.1594,
      "num_input_tokens_seen": 28668816,
      "step": 33100
    },
    {
      "epoch": 15.608203677510609,
      "grad_norm": 0.010502105578780174,
      "learning_rate": 0.021468072674406414,
      "loss": 0.1483,
      "num_input_tokens_seen": 28673296,
      "step": 33105
    },
    {
      "epoch": 15.61056105610561,
      "grad_norm": 0.0037844388280063868,
      "learning_rate": 0.021437716144006795,
      "loss": 0.1121,
      "num_input_tokens_seen": 28677376,
      "step": 33110
    },
    {
      "epoch": 15.612918434700614,
      "grad_norm": 0.006872489582747221,
      "learning_rate": 0.021407379439527002,
      "loss": 0.1363,
      "num_input_tokens_seen": 28681648,
      "step": 33115
    },
    {
      "epoch": 15.615275813295614,
      "grad_norm": 0.005345071200281382,
      "learning_rate": 0.021377062565645255,
      "loss": 0.1524,
      "num_input_tokens_seen": 28686144,
      "step": 33120
    },
    {
      "epoch": 15.617633191890617,
      "grad_norm": 0.0056538814678788185,
      "learning_rate": 0.02134676552703688,
      "loss": 0.1361,
      "num_input_tokens_seen": 28690688,
      "step": 33125
    },
    {
      "epoch": 15.61999057048562,
      "grad_norm": 0.003952366299927235,
      "learning_rate": 0.02131648832837398,
      "loss": 0.1043,
      "num_input_tokens_seen": 28695232,
      "step": 33130
    },
    {
      "epoch": 15.622347949080622,
      "grad_norm": 0.006569413933902979,
      "learning_rate": 0.02128623097432574,
      "loss": 0.1617,
      "num_input_tokens_seen": 28699584,
      "step": 33135
    },
    {
      "epoch": 15.624705327675624,
      "grad_norm": 0.0025312684010714293,
      "learning_rate": 0.021255993469558192,
      "loss": 0.1215,
      "num_input_tokens_seen": 28704720,
      "step": 33140
    },
    {
      "epoch": 15.627062706270626,
      "grad_norm": 0.013719763606786728,
      "learning_rate": 0.021225775818734364,
      "loss": 0.1913,
      "num_input_tokens_seen": 28708528,
      "step": 33145
    },
    {
      "epoch": 15.629420084865629,
      "grad_norm": 0.009359954856336117,
      "learning_rate": 0.021195578026514166,
      "loss": 0.1774,
      "num_input_tokens_seen": 28713104,
      "step": 33150
    },
    {
      "epoch": 15.631777463460631,
      "grad_norm": 0.01313064992427826,
      "learning_rate": 0.02116540009755452,
      "loss": 0.1637,
      "num_input_tokens_seen": 28718208,
      "step": 33155
    },
    {
      "epoch": 15.634134842055634,
      "grad_norm": 0.006834178231656551,
      "learning_rate": 0.021135242036509173,
      "loss": 0.1665,
      "num_input_tokens_seen": 28722096,
      "step": 33160
    },
    {
      "epoch": 15.636492220650636,
      "grad_norm": 0.015169400721788406,
      "learning_rate": 0.021105103848028967,
      "loss": 0.1416,
      "num_input_tokens_seen": 28726368,
      "step": 33165
    },
    {
      "epoch": 15.638849599245638,
      "grad_norm": 0.006066963076591492,
      "learning_rate": 0.021074985536761504,
      "loss": 0.1379,
      "num_input_tokens_seen": 28731856,
      "step": 33170
    },
    {
      "epoch": 15.64120697784064,
      "grad_norm": 0.010219521820545197,
      "learning_rate": 0.021044887107351435,
      "loss": 0.1324,
      "num_input_tokens_seen": 28736592,
      "step": 33175
    },
    {
      "epoch": 15.643564356435643,
      "grad_norm": 0.011590374633669853,
      "learning_rate": 0.021014808564440362,
      "loss": 0.1324,
      "num_input_tokens_seen": 28740144,
      "step": 33180
    },
    {
      "epoch": 15.645921735030646,
      "grad_norm": 0.005227481480687857,
      "learning_rate": 0.02098474991266671,
      "loss": 0.1142,
      "num_input_tokens_seen": 28744080,
      "step": 33185
    },
    {
      "epoch": 15.648279113625648,
      "grad_norm": 0.010854237712919712,
      "learning_rate": 0.02095471115666592,
      "loss": 0.1438,
      "num_input_tokens_seen": 28748320,
      "step": 33190
    },
    {
      "epoch": 15.65063649222065,
      "grad_norm": 0.013706477358937263,
      "learning_rate": 0.020924692301070406,
      "loss": 0.1998,
      "num_input_tokens_seen": 28752832,
      "step": 33195
    },
    {
      "epoch": 15.652993870815653,
      "grad_norm": 0.010388951748609543,
      "learning_rate": 0.020894693350509346,
      "loss": 0.1954,
      "num_input_tokens_seen": 28756240,
      "step": 33200
    },
    {
      "epoch": 15.652993870815653,
      "eval_loss": 0.2356317937374115,
      "eval_runtime": 21.9273,
      "eval_samples_per_second": 43.006,
      "eval_steps_per_second": 21.526,
      "num_input_tokens_seen": 28756240,
      "step": 33200
    },
    {
      "epoch": 15.655351249410655,
      "grad_norm": 0.014369914308190346,
      "learning_rate": 0.020864714309609057,
      "loss": 0.1867,
      "num_input_tokens_seen": 28760752,
      "step": 33205
    },
    {
      "epoch": 15.657708628005658,
      "grad_norm": 0.009331897832453251,
      "learning_rate": 0.020834755182992604,
      "loss": 0.1384,
      "num_input_tokens_seen": 28764896,
      "step": 33210
    },
    {
      "epoch": 15.66006600660066,
      "grad_norm": 0.004013505764305592,
      "learning_rate": 0.02080481597528011,
      "loss": 0.1087,
      "num_input_tokens_seen": 28769168,
      "step": 33215
    },
    {
      "epoch": 15.662423385195662,
      "grad_norm": 0.014961844310164452,
      "learning_rate": 0.020774896691088583,
      "loss": 0.1583,
      "num_input_tokens_seen": 28772704,
      "step": 33220
    },
    {
      "epoch": 15.664780763790665,
      "grad_norm": 0.0034656417556107044,
      "learning_rate": 0.020744997335031882,
      "loss": 0.0825,
      "num_input_tokens_seen": 28776736,
      "step": 33225
    },
    {
      "epoch": 15.667138142385667,
      "grad_norm": 0.009257973171770573,
      "learning_rate": 0.02071511791172092,
      "loss": 0.1528,
      "num_input_tokens_seen": 28780704,
      "step": 33230
    },
    {
      "epoch": 15.66949552098067,
      "grad_norm": 0.015145860612392426,
      "learning_rate": 0.02068525842576351,
      "loss": 0.1493,
      "num_input_tokens_seen": 28784576,
      "step": 33235
    },
    {
      "epoch": 15.671852899575672,
      "grad_norm": 0.00983085110783577,
      "learning_rate": 0.020655418881764264,
      "loss": 0.1725,
      "num_input_tokens_seen": 28789024,
      "step": 33240
    },
    {
      "epoch": 15.674210278170674,
      "grad_norm": 0.009945601224899292,
      "learning_rate": 0.020625599284324923,
      "loss": 0.1173,
      "num_input_tokens_seen": 28793648,
      "step": 33245
    },
    {
      "epoch": 15.676567656765677,
      "grad_norm": 0.010404549539089203,
      "learning_rate": 0.02059579963804396,
      "loss": 0.1638,
      "num_input_tokens_seen": 28798048,
      "step": 33250
    },
    {
      "epoch": 15.67892503536068,
      "grad_norm": 0.01724894344806671,
      "learning_rate": 0.02056601994751688,
      "loss": 0.1953,
      "num_input_tokens_seen": 28802128,
      "step": 33255
    },
    {
      "epoch": 15.681282413955682,
      "grad_norm": 0.007362672593444586,
      "learning_rate": 0.02053626021733614,
      "loss": 0.1876,
      "num_input_tokens_seen": 28806688,
      "step": 33260
    },
    {
      "epoch": 15.683639792550684,
      "grad_norm": 0.00902239978313446,
      "learning_rate": 0.02050652045209097,
      "loss": 0.1657,
      "num_input_tokens_seen": 28811072,
      "step": 33265
    },
    {
      "epoch": 15.685997171145686,
      "grad_norm": 0.007616869639605284,
      "learning_rate": 0.020476800656367672,
      "loss": 0.1479,
      "num_input_tokens_seen": 28815520,
      "step": 33270
    },
    {
      "epoch": 15.688354549740689,
      "grad_norm": 0.006913819350302219,
      "learning_rate": 0.020447100834749425,
      "loss": 0.1482,
      "num_input_tokens_seen": 28820880,
      "step": 33275
    },
    {
      "epoch": 15.690711928335691,
      "grad_norm": 0.01078998763114214,
      "learning_rate": 0.02041742099181627,
      "loss": 0.0884,
      "num_input_tokens_seen": 28825440,
      "step": 33280
    },
    {
      "epoch": 15.693069306930694,
      "grad_norm": 0.010669144801795483,
      "learning_rate": 0.02038776113214526,
      "loss": 0.1015,
      "num_input_tokens_seen": 28829584,
      "step": 33285
    },
    {
      "epoch": 15.695426685525696,
      "grad_norm": 0.007945742458105087,
      "learning_rate": 0.0203581212603103,
      "loss": 0.1498,
      "num_input_tokens_seen": 28833712,
      "step": 33290
    },
    {
      "epoch": 15.697784064120698,
      "grad_norm": 0.01267133466899395,
      "learning_rate": 0.02032850138088219,
      "loss": 0.1393,
      "num_input_tokens_seen": 28837920,
      "step": 33295
    },
    {
      "epoch": 15.700141442715701,
      "grad_norm": 0.00546290585771203,
      "learning_rate": 0.020298901498428754,
      "loss": 0.1067,
      "num_input_tokens_seen": 28842176,
      "step": 33300
    },
    {
      "epoch": 15.702498821310703,
      "grad_norm": 0.013740873895585537,
      "learning_rate": 0.020269321617514595,
      "loss": 0.2519,
      "num_input_tokens_seen": 28846192,
      "step": 33305
    },
    {
      "epoch": 15.704856199905706,
      "grad_norm": 0.01193067617714405,
      "learning_rate": 0.020239761742701343,
      "loss": 0.1727,
      "num_input_tokens_seen": 28851712,
      "step": 33310
    },
    {
      "epoch": 15.707213578500706,
      "grad_norm": 0.013204175047576427,
      "learning_rate": 0.02021022187854754,
      "loss": 0.1607,
      "num_input_tokens_seen": 28855472,
      "step": 33315
    },
    {
      "epoch": 15.70957095709571,
      "grad_norm": 0.007259982172399759,
      "learning_rate": 0.020180702029608522,
      "loss": 0.1463,
      "num_input_tokens_seen": 28859088,
      "step": 33320
    },
    {
      "epoch": 15.711928335690711,
      "grad_norm": 0.015096140094101429,
      "learning_rate": 0.020151202200436695,
      "loss": 0.1688,
      "num_input_tokens_seen": 28863456,
      "step": 33325
    },
    {
      "epoch": 15.714285714285714,
      "grad_norm": 0.005905061028897762,
      "learning_rate": 0.020121722395581226,
      "loss": 0.1493,
      "num_input_tokens_seen": 28867360,
      "step": 33330
    },
    {
      "epoch": 15.716643092880716,
      "grad_norm": 0.011165544390678406,
      "learning_rate": 0.020092262619588342,
      "loss": 0.1188,
      "num_input_tokens_seen": 28872000,
      "step": 33335
    },
    {
      "epoch": 15.719000471475718,
      "grad_norm": 0.01145342830568552,
      "learning_rate": 0.02006282287700109,
      "loss": 0.1876,
      "num_input_tokens_seen": 28876192,
      "step": 33340
    },
    {
      "epoch": 15.72135785007072,
      "grad_norm": 0.010686317458748817,
      "learning_rate": 0.020033403172359427,
      "loss": 0.0789,
      "num_input_tokens_seen": 28879824,
      "step": 33345
    },
    {
      "epoch": 15.723715228665723,
      "grad_norm": 0.010326235555112362,
      "learning_rate": 0.020004003510200284,
      "loss": 0.1262,
      "num_input_tokens_seen": 28883776,
      "step": 33350
    },
    {
      "epoch": 15.726072607260726,
      "grad_norm": 0.005926843732595444,
      "learning_rate": 0.019974623895057407,
      "loss": 0.0664,
      "num_input_tokens_seen": 28887904,
      "step": 33355
    },
    {
      "epoch": 15.728429985855728,
      "grad_norm": 0.009972752071917057,
      "learning_rate": 0.019945264331461553,
      "loss": 0.2003,
      "num_input_tokens_seen": 28892256,
      "step": 33360
    },
    {
      "epoch": 15.73078736445073,
      "grad_norm": 0.00403845589607954,
      "learning_rate": 0.019915924823940317,
      "loss": 0.1447,
      "num_input_tokens_seen": 28897184,
      "step": 33365
    },
    {
      "epoch": 15.733144743045733,
      "grad_norm": 0.015507747419178486,
      "learning_rate": 0.01988660537701816,
      "loss": 0.3053,
      "num_input_tokens_seen": 28901040,
      "step": 33370
    },
    {
      "epoch": 15.735502121640735,
      "grad_norm": 0.009736848063766956,
      "learning_rate": 0.01985730599521659,
      "loss": 0.1295,
      "num_input_tokens_seen": 28905600,
      "step": 33375
    },
    {
      "epoch": 15.737859500235738,
      "grad_norm": 0.005619247909635305,
      "learning_rate": 0.019828026683053918,
      "loss": 0.1238,
      "num_input_tokens_seen": 28909552,
      "step": 33380
    },
    {
      "epoch": 15.74021687883074,
      "grad_norm": 0.015858426690101624,
      "learning_rate": 0.01979876744504535,
      "loss": 0.1267,
      "num_input_tokens_seen": 28913632,
      "step": 33385
    },
    {
      "epoch": 15.742574257425742,
      "grad_norm": 0.019311323761940002,
      "learning_rate": 0.019769528285703046,
      "loss": 0.1957,
      "num_input_tokens_seen": 28917888,
      "step": 33390
    },
    {
      "epoch": 15.744931636020745,
      "grad_norm": 0.009724169969558716,
      "learning_rate": 0.019740309209536098,
      "loss": 0.1797,
      "num_input_tokens_seen": 28922080,
      "step": 33395
    },
    {
      "epoch": 15.747289014615747,
      "grad_norm": 0.008935193531215191,
      "learning_rate": 0.019711110221050387,
      "loss": 0.1933,
      "num_input_tokens_seen": 28926208,
      "step": 33400
    },
    {
      "epoch": 15.747289014615747,
      "eval_loss": 0.2358691394329071,
      "eval_runtime": 21.9343,
      "eval_samples_per_second": 42.992,
      "eval_steps_per_second": 21.519,
      "num_input_tokens_seen": 28926208,
      "step": 33400
    },
    {
      "epoch": 15.74964639321075,
      "grad_norm": 0.009668223559856415,
      "learning_rate": 0.019681931324748825,
      "loss": 0.1476,
      "num_input_tokens_seen": 28930288,
      "step": 33405
    },
    {
      "epoch": 15.752003771805752,
      "grad_norm": 0.006874902173876762,
      "learning_rate": 0.019652772525131094,
      "loss": 0.1687,
      "num_input_tokens_seen": 28935376,
      "step": 33410
    },
    {
      "epoch": 15.754361150400754,
      "grad_norm": 0.010143081657588482,
      "learning_rate": 0.019623633826693885,
      "loss": 0.1433,
      "num_input_tokens_seen": 28939248,
      "step": 33415
    },
    {
      "epoch": 15.756718528995757,
      "grad_norm": 0.014642305672168732,
      "learning_rate": 0.019594515233930788,
      "loss": 0.1338,
      "num_input_tokens_seen": 28943392,
      "step": 33420
    },
    {
      "epoch": 15.75907590759076,
      "grad_norm": 0.015993531793355942,
      "learning_rate": 0.019565416751332186,
      "loss": 0.17,
      "num_input_tokens_seen": 28947808,
      "step": 33425
    },
    {
      "epoch": 15.761433286185762,
      "grad_norm": 0.006854436360299587,
      "learning_rate": 0.019536338383385497,
      "loss": 0.1535,
      "num_input_tokens_seen": 28951984,
      "step": 33430
    },
    {
      "epoch": 15.763790664780764,
      "grad_norm": 0.005140727385878563,
      "learning_rate": 0.019507280134574933,
      "loss": 0.1544,
      "num_input_tokens_seen": 28956032,
      "step": 33435
    },
    {
      "epoch": 15.766148043375766,
      "grad_norm": 0.011818690225481987,
      "learning_rate": 0.019478242009381624,
      "loss": 0.134,
      "num_input_tokens_seen": 28960672,
      "step": 33440
    },
    {
      "epoch": 15.768505421970769,
      "grad_norm": 0.011516189202666283,
      "learning_rate": 0.01944922401228367,
      "loss": 0.1608,
      "num_input_tokens_seen": 28965248,
      "step": 33445
    },
    {
      "epoch": 15.770862800565771,
      "grad_norm": 0.00805049017071724,
      "learning_rate": 0.01942022614775593,
      "loss": 0.0847,
      "num_input_tokens_seen": 28969824,
      "step": 33450
    },
    {
      "epoch": 15.773220179160774,
      "grad_norm": 0.011468291282653809,
      "learning_rate": 0.01939124842027029,
      "loss": 0.1301,
      "num_input_tokens_seen": 28974368,
      "step": 33455
    },
    {
      "epoch": 15.775577557755776,
      "grad_norm": 0.014600384049117565,
      "learning_rate": 0.01936229083429551,
      "loss": 0.2214,
      "num_input_tokens_seen": 28978128,
      "step": 33460
    },
    {
      "epoch": 15.777934936350778,
      "grad_norm": 0.010637049563229084,
      "learning_rate": 0.019333353394297148,
      "loss": 0.1204,
      "num_input_tokens_seen": 28982048,
      "step": 33465
    },
    {
      "epoch": 15.78029231494578,
      "grad_norm": 0.01019955798983574,
      "learning_rate": 0.019304436104737754,
      "loss": 0.1452,
      "num_input_tokens_seen": 28986336,
      "step": 33470
    },
    {
      "epoch": 15.782649693540783,
      "grad_norm": 0.012605104595422745,
      "learning_rate": 0.019275538970076778,
      "loss": 0.1324,
      "num_input_tokens_seen": 28990560,
      "step": 33475
    },
    {
      "epoch": 15.785007072135786,
      "grad_norm": 0.006216756068170071,
      "learning_rate": 0.019246661994770434,
      "loss": 0.1466,
      "num_input_tokens_seen": 28994416,
      "step": 33480
    },
    {
      "epoch": 15.787364450730788,
      "grad_norm": 0.011961767449975014,
      "learning_rate": 0.019217805183271985,
      "loss": 0.1477,
      "num_input_tokens_seen": 28998304,
      "step": 33485
    },
    {
      "epoch": 15.78972182932579,
      "grad_norm": 0.011355184018611908,
      "learning_rate": 0.019188968540031465,
      "loss": 0.143,
      "num_input_tokens_seen": 29002160,
      "step": 33490
    },
    {
      "epoch": 15.792079207920793,
      "grad_norm": 0.010464862920343876,
      "learning_rate": 0.019160152069495867,
      "loss": 0.1324,
      "num_input_tokens_seen": 29006768,
      "step": 33495
    },
    {
      "epoch": 15.794436586515795,
      "grad_norm": 0.008440044708549976,
      "learning_rate": 0.019131355776109103,
      "loss": 0.2708,
      "num_input_tokens_seen": 29011472,
      "step": 33500
    },
    {
      "epoch": 15.796793965110798,
      "grad_norm": 0.00617208331823349,
      "learning_rate": 0.019102579664311857,
      "loss": 0.2133,
      "num_input_tokens_seen": 29015984,
      "step": 33505
    },
    {
      "epoch": 15.799151343705798,
      "grad_norm": 0.006971802096813917,
      "learning_rate": 0.019073823738541763,
      "loss": 0.1476,
      "num_input_tokens_seen": 29020672,
      "step": 33510
    },
    {
      "epoch": 15.801508722300802,
      "grad_norm": 0.011096819303929806,
      "learning_rate": 0.0190450880032334,
      "loss": 0.1167,
      "num_input_tokens_seen": 29024608,
      "step": 33515
    },
    {
      "epoch": 15.803866100895803,
      "grad_norm": 0.008110621944069862,
      "learning_rate": 0.019016372462818114,
      "loss": 0.1247,
      "num_input_tokens_seen": 29029024,
      "step": 33520
    },
    {
      "epoch": 15.806223479490805,
      "grad_norm": 0.005207305308431387,
      "learning_rate": 0.018987677121724278,
      "loss": 0.144,
      "num_input_tokens_seen": 29033280,
      "step": 33525
    },
    {
      "epoch": 15.808580858085808,
      "grad_norm": 0.01477016881108284,
      "learning_rate": 0.018959001984377,
      "loss": 0.2034,
      "num_input_tokens_seen": 29037584,
      "step": 33530
    },
    {
      "epoch": 15.81093823668081,
      "grad_norm": 0.011677647940814495,
      "learning_rate": 0.018930347055198377,
      "loss": 0.1885,
      "num_input_tokens_seen": 29041312,
      "step": 33535
    },
    {
      "epoch": 15.813295615275813,
      "grad_norm": 0.00546837504953146,
      "learning_rate": 0.01890171233860739,
      "loss": 0.1615,
      "num_input_tokens_seen": 29045360,
      "step": 33540
    },
    {
      "epoch": 15.815652993870815,
      "grad_norm": 0.01631225273013115,
      "learning_rate": 0.018873097839019807,
      "loss": 0.1882,
      "num_input_tokens_seen": 29049792,
      "step": 33545
    },
    {
      "epoch": 15.818010372465817,
      "grad_norm": 0.01510283350944519,
      "learning_rate": 0.0188445035608484,
      "loss": 0.2107,
      "num_input_tokens_seen": 29054096,
      "step": 33550
    },
    {
      "epoch": 15.82036775106082,
      "grad_norm": 0.013078639283776283,
      "learning_rate": 0.018815929508502777,
      "loss": 0.1537,
      "num_input_tokens_seen": 29057952,
      "step": 33555
    },
    {
      "epoch": 15.822725129655822,
      "grad_norm": 0.012507554143667221,
      "learning_rate": 0.01878737568638934,
      "loss": 0.1465,
      "num_input_tokens_seen": 29062544,
      "step": 33560
    },
    {
      "epoch": 15.825082508250825,
      "grad_norm": 0.011849344708025455,
      "learning_rate": 0.01875884209891152,
      "loss": 0.1219,
      "num_input_tokens_seen": 29066992,
      "step": 33565
    },
    {
      "epoch": 15.827439886845827,
      "grad_norm": 0.004110679030418396,
      "learning_rate": 0.018730328750469514,
      "loss": 0.1161,
      "num_input_tokens_seen": 29070704,
      "step": 33570
    },
    {
      "epoch": 15.82979726544083,
      "grad_norm": 0.014476512558758259,
      "learning_rate": 0.018701835645460473,
      "loss": 0.1797,
      "num_input_tokens_seen": 29074736,
      "step": 33575
    },
    {
      "epoch": 15.832154644035832,
      "grad_norm": 0.005878007505089045,
      "learning_rate": 0.01867336278827838,
      "loss": 0.1282,
      "num_input_tokens_seen": 29079136,
      "step": 33580
    },
    {
      "epoch": 15.834512022630834,
      "grad_norm": 0.009585722349584103,
      "learning_rate": 0.018644910183314056,
      "loss": 0.1112,
      "num_input_tokens_seen": 29083904,
      "step": 33585
    },
    {
      "epoch": 15.836869401225837,
      "grad_norm": 0.011633875779807568,
      "learning_rate": 0.01861647783495531,
      "loss": 0.1665,
      "num_input_tokens_seen": 29088608,
      "step": 33590
    },
    {
      "epoch": 15.839226779820839,
      "grad_norm": 0.005122533068060875,
      "learning_rate": 0.01858806574758676,
      "loss": 0.1591,
      "num_input_tokens_seen": 29092688,
      "step": 33595
    },
    {
      "epoch": 15.841584158415841,
      "grad_norm": 0.009182100184261799,
      "learning_rate": 0.01855967392558988,
      "loss": 0.1241,
      "num_input_tokens_seen": 29096816,
      "step": 33600
    },
    {
      "epoch": 15.841584158415841,
      "eval_loss": 0.23447120189666748,
      "eval_runtime": 21.9341,
      "eval_samples_per_second": 42.992,
      "eval_steps_per_second": 21.519,
      "num_input_tokens_seen": 29096816,
      "step": 33600
    },
    {
      "epoch": 15.843941537010844,
      "grad_norm": 0.007577728945761919,
      "learning_rate": 0.018531302373343096,
      "loss": 0.1571,
      "num_input_tokens_seen": 29101056,
      "step": 33605
    },
    {
      "epoch": 15.846298915605846,
      "grad_norm": 0.01298991683870554,
      "learning_rate": 0.018502951095221588,
      "loss": 0.1744,
      "num_input_tokens_seen": 29105568,
      "step": 33610
    },
    {
      "epoch": 15.848656294200849,
      "grad_norm": 0.01238330453634262,
      "learning_rate": 0.01847462009559751,
      "loss": 0.2041,
      "num_input_tokens_seen": 29109104,
      "step": 33615
    },
    {
      "epoch": 15.851013672795851,
      "grad_norm": 0.010474913753569126,
      "learning_rate": 0.01844630937883992,
      "loss": 0.2178,
      "num_input_tokens_seen": 29113312,
      "step": 33620
    },
    {
      "epoch": 15.853371051390853,
      "grad_norm": 0.006106297019869089,
      "learning_rate": 0.018418018949314573,
      "loss": 0.0885,
      "num_input_tokens_seen": 29117600,
      "step": 33625
    },
    {
      "epoch": 15.855728429985856,
      "grad_norm": 0.010363249108195305,
      "learning_rate": 0.018389748811384315,
      "loss": 0.1538,
      "num_input_tokens_seen": 29122096,
      "step": 33630
    },
    {
      "epoch": 15.858085808580858,
      "grad_norm": 0.010607166215777397,
      "learning_rate": 0.018361498969408658,
      "loss": 0.13,
      "num_input_tokens_seen": 29125824,
      "step": 33635
    },
    {
      "epoch": 15.86044318717586,
      "grad_norm": 0.009844205342233181,
      "learning_rate": 0.01833326942774415,
      "loss": 0.1411,
      "num_input_tokens_seen": 29130240,
      "step": 33640
    },
    {
      "epoch": 15.862800565770863,
      "grad_norm": 0.003402505535632372,
      "learning_rate": 0.018305060190744155,
      "loss": 0.0859,
      "num_input_tokens_seen": 29134912,
      "step": 33645
    },
    {
      "epoch": 15.865157944365865,
      "grad_norm": 0.009187222458422184,
      "learning_rate": 0.018276871262758846,
      "loss": 0.1345,
      "num_input_tokens_seen": 29139456,
      "step": 33650
    },
    {
      "epoch": 15.867515322960868,
      "grad_norm": 0.0094832768663764,
      "learning_rate": 0.0182487026481353,
      "loss": 0.1225,
      "num_input_tokens_seen": 29143600,
      "step": 33655
    },
    {
      "epoch": 15.86987270155587,
      "grad_norm": 0.013676558621227741,
      "learning_rate": 0.018220554351217538,
      "loss": 0.1362,
      "num_input_tokens_seen": 29147712,
      "step": 33660
    },
    {
      "epoch": 15.872230080150873,
      "grad_norm": 0.009403099305927753,
      "learning_rate": 0.01819242637634629,
      "loss": 0.1227,
      "num_input_tokens_seen": 29151776,
      "step": 33665
    },
    {
      "epoch": 15.874587458745875,
      "grad_norm": 0.008690647780895233,
      "learning_rate": 0.01816431872785933,
      "loss": 0.1529,
      "num_input_tokens_seen": 29156512,
      "step": 33670
    },
    {
      "epoch": 15.876944837340877,
      "grad_norm": 0.007721145171672106,
      "learning_rate": 0.018136231410091148,
      "loss": 0.1352,
      "num_input_tokens_seen": 29161456,
      "step": 33675
    },
    {
      "epoch": 15.87930221593588,
      "grad_norm": 0.00759466178715229,
      "learning_rate": 0.018108164427373175,
      "loss": 0.1698,
      "num_input_tokens_seen": 29165488,
      "step": 33680
    },
    {
      "epoch": 15.881659594530882,
      "grad_norm": 0.004428298212587833,
      "learning_rate": 0.01808011778403375,
      "loss": 0.1291,
      "num_input_tokens_seen": 29169648,
      "step": 33685
    },
    {
      "epoch": 15.884016973125885,
      "grad_norm": 0.004801807459443808,
      "learning_rate": 0.01805209148439793,
      "loss": 0.1274,
      "num_input_tokens_seen": 29173344,
      "step": 33690
    },
    {
      "epoch": 15.886374351720887,
      "grad_norm": 0.008379009552299976,
      "learning_rate": 0.018024085532787757,
      "loss": 0.158,
      "num_input_tokens_seen": 29177312,
      "step": 33695
    },
    {
      "epoch": 15.88873173031589,
      "grad_norm": 0.008536525070667267,
      "learning_rate": 0.017996099933522164,
      "loss": 0.1474,
      "num_input_tokens_seen": 29181696,
      "step": 33700
    },
    {
      "epoch": 15.891089108910892,
      "grad_norm": 0.016969969496130943,
      "learning_rate": 0.017968134690916775,
      "loss": 0.2143,
      "num_input_tokens_seen": 29185808,
      "step": 33705
    },
    {
      "epoch": 15.893446487505894,
      "grad_norm": 0.007745060138404369,
      "learning_rate": 0.017940189809284263,
      "loss": 0.2069,
      "num_input_tokens_seen": 29190064,
      "step": 33710
    },
    {
      "epoch": 15.895803866100895,
      "grad_norm": 0.012543420307338238,
      "learning_rate": 0.017912265292934024,
      "loss": 0.2238,
      "num_input_tokens_seen": 29193632,
      "step": 33715
    },
    {
      "epoch": 15.898161244695899,
      "grad_norm": 0.005619990173727274,
      "learning_rate": 0.017884361146172423,
      "loss": 0.1212,
      "num_input_tokens_seen": 29197520,
      "step": 33720
    },
    {
      "epoch": 15.9005186232909,
      "grad_norm": 0.011378907598555088,
      "learning_rate": 0.01785647737330261,
      "loss": 0.147,
      "num_input_tokens_seen": 29202544,
      "step": 33725
    },
    {
      "epoch": 15.902876001885902,
      "grad_norm": 0.01348639465868473,
      "learning_rate": 0.017828613978624563,
      "loss": 0.1805,
      "num_input_tokens_seen": 29207168,
      "step": 33730
    },
    {
      "epoch": 15.905233380480905,
      "grad_norm": 0.004668994806706905,
      "learning_rate": 0.01780077096643523,
      "loss": 0.1572,
      "num_input_tokens_seen": 29211728,
      "step": 33735
    },
    {
      "epoch": 15.907590759075907,
      "grad_norm": 0.011346595361828804,
      "learning_rate": 0.017772948341028345,
      "loss": 0.1175,
      "num_input_tokens_seen": 29215616,
      "step": 33740
    },
    {
      "epoch": 15.90994813767091,
      "grad_norm": 0.01404068898409605,
      "learning_rate": 0.01774514610669447,
      "loss": 0.1313,
      "num_input_tokens_seen": 29220928,
      "step": 33745
    },
    {
      "epoch": 15.912305516265912,
      "grad_norm": 0.018039526417851448,
      "learning_rate": 0.017717364267721112,
      "loss": 0.1942,
      "num_input_tokens_seen": 29224608,
      "step": 33750
    },
    {
      "epoch": 15.914662894860914,
      "grad_norm": 0.014733195304870605,
      "learning_rate": 0.017689602828392513,
      "loss": 0.173,
      "num_input_tokens_seen": 29229120,
      "step": 33755
    },
    {
      "epoch": 15.917020273455917,
      "grad_norm": 0.008668680675327778,
      "learning_rate": 0.017661861792989897,
      "loss": 0.1692,
      "num_input_tokens_seen": 29233408,
      "step": 33760
    },
    {
      "epoch": 15.919377652050919,
      "grad_norm": 0.00762353278696537,
      "learning_rate": 0.017634141165791272,
      "loss": 0.1764,
      "num_input_tokens_seen": 29237008,
      "step": 33765
    },
    {
      "epoch": 15.921735030645921,
      "grad_norm": 0.019390935078263283,
      "learning_rate": 0.017606440951071455,
      "loss": 0.2788,
      "num_input_tokens_seen": 29240384,
      "step": 33770
    },
    {
      "epoch": 15.924092409240924,
      "grad_norm": 0.012553144246339798,
      "learning_rate": 0.017578761153102213,
      "loss": 0.134,
      "num_input_tokens_seen": 29244832,
      "step": 33775
    },
    {
      "epoch": 15.926449787835926,
      "grad_norm": 0.008866786025464535,
      "learning_rate": 0.017551101776152146,
      "loss": 0.1484,
      "num_input_tokens_seen": 29250896,
      "step": 33780
    },
    {
      "epoch": 15.928807166430929,
      "grad_norm": 0.008158031851053238,
      "learning_rate": 0.017523462824486608,
      "loss": 0.1289,
      "num_input_tokens_seen": 29255168,
      "step": 33785
    },
    {
      "epoch": 15.931164545025931,
      "grad_norm": 0.01971360668540001,
      "learning_rate": 0.01749584430236794,
      "loss": 0.1636,
      "num_input_tokens_seen": 29258880,
      "step": 33790
    },
    {
      "epoch": 15.933521923620933,
      "grad_norm": 0.011381003074347973,
      "learning_rate": 0.01746824621405524,
      "loss": 0.1776,
      "num_input_tokens_seen": 29263200,
      "step": 33795
    },
    {
      "epoch": 15.935879302215936,
      "grad_norm": 0.015404395759105682,
      "learning_rate": 0.017440668563804412,
      "loss": 0.1931,
      "num_input_tokens_seen": 29267072,
      "step": 33800
    },
    {
      "epoch": 15.935879302215936,
      "eval_loss": 0.23921844363212585,
      "eval_runtime": 21.9135,
      "eval_samples_per_second": 43.033,
      "eval_steps_per_second": 21.539,
      "num_input_tokens_seen": 29267072,
      "step": 33800
    },
    {
      "epoch": 15.938236680810938,
      "grad_norm": 0.015389579348266125,
      "learning_rate": 0.017413111355868392,
      "loss": 0.194,
      "num_input_tokens_seen": 29272208,
      "step": 33805
    },
    {
      "epoch": 15.94059405940594,
      "grad_norm": 0.003462106455117464,
      "learning_rate": 0.017385574594496748,
      "loss": 0.1395,
      "num_input_tokens_seen": 29276528,
      "step": 33810
    },
    {
      "epoch": 15.942951438000943,
      "grad_norm": 0.017849478870630264,
      "learning_rate": 0.01735805828393605,
      "loss": 0.155,
      "num_input_tokens_seen": 29280416,
      "step": 33815
    },
    {
      "epoch": 15.945308816595945,
      "grad_norm": 0.008412381634116173,
      "learning_rate": 0.017330562428429667,
      "loss": 0.1658,
      "num_input_tokens_seen": 29285088,
      "step": 33820
    },
    {
      "epoch": 15.947666195190948,
      "grad_norm": 0.012100562453269958,
      "learning_rate": 0.01730308703221776,
      "loss": 0.1381,
      "num_input_tokens_seen": 29289168,
      "step": 33825
    },
    {
      "epoch": 15.95002357378595,
      "grad_norm": 0.00923791155219078,
      "learning_rate": 0.01727563209953744,
      "loss": 0.1078,
      "num_input_tokens_seen": 29293296,
      "step": 33830
    },
    {
      "epoch": 15.952380952380953,
      "grad_norm": 0.012612216174602509,
      "learning_rate": 0.017248197634622535,
      "loss": 0.1394,
      "num_input_tokens_seen": 29297696,
      "step": 33835
    },
    {
      "epoch": 15.954738330975955,
      "grad_norm": 0.01085917092859745,
      "learning_rate": 0.01722078364170383,
      "loss": 0.1766,
      "num_input_tokens_seen": 29301664,
      "step": 33840
    },
    {
      "epoch": 15.957095709570957,
      "grad_norm": 0.008514147251844406,
      "learning_rate": 0.017193390125008905,
      "loss": 0.1367,
      "num_input_tokens_seen": 29305776,
      "step": 33845
    },
    {
      "epoch": 15.95945308816596,
      "grad_norm": 0.006366985384374857,
      "learning_rate": 0.017166017088762153,
      "loss": 0.105,
      "num_input_tokens_seen": 29309856,
      "step": 33850
    },
    {
      "epoch": 15.961810466760962,
      "grad_norm": 0.003963693045079708,
      "learning_rate": 0.017138664537184878,
      "loss": 0.0947,
      "num_input_tokens_seen": 29313328,
      "step": 33855
    },
    {
      "epoch": 15.964167845355965,
      "grad_norm": 0.0078471340239048,
      "learning_rate": 0.017111332474495172,
      "loss": 0.1166,
      "num_input_tokens_seen": 29317344,
      "step": 33860
    },
    {
      "epoch": 15.966525223950967,
      "grad_norm": 0.004692547954618931,
      "learning_rate": 0.017084020904907998,
      "loss": 0.1121,
      "num_input_tokens_seen": 29321344,
      "step": 33865
    },
    {
      "epoch": 15.96888260254597,
      "grad_norm": 0.007308132015168667,
      "learning_rate": 0.017056729832635103,
      "loss": 0.1648,
      "num_input_tokens_seen": 29325200,
      "step": 33870
    },
    {
      "epoch": 15.971239981140972,
      "grad_norm": 0.01259162649512291,
      "learning_rate": 0.017029459261885153,
      "loss": 0.1617,
      "num_input_tokens_seen": 29329248,
      "step": 33875
    },
    {
      "epoch": 15.973597359735974,
      "grad_norm": 0.013526651076972485,
      "learning_rate": 0.01700220919686359,
      "loss": 0.1016,
      "num_input_tokens_seen": 29333248,
      "step": 33880
    },
    {
      "epoch": 15.975954738330977,
      "grad_norm": 0.0137288523837924,
      "learning_rate": 0.016974979641772723,
      "loss": 0.2651,
      "num_input_tokens_seen": 29337312,
      "step": 33885
    },
    {
      "epoch": 15.978312116925979,
      "grad_norm": 0.009665065445005894,
      "learning_rate": 0.01694777060081169,
      "loss": 0.1286,
      "num_input_tokens_seen": 29340896,
      "step": 33890
    },
    {
      "epoch": 15.980669495520981,
      "grad_norm": 0.012758394703269005,
      "learning_rate": 0.016920582078176444,
      "loss": 0.1184,
      "num_input_tokens_seen": 29344928,
      "step": 33895
    },
    {
      "epoch": 15.983026874115984,
      "grad_norm": 0.009794783778488636,
      "learning_rate": 0.016893414078059863,
      "loss": 0.1799,
      "num_input_tokens_seen": 29349616,
      "step": 33900
    },
    {
      "epoch": 15.985384252710986,
      "grad_norm": 0.012445393018424511,
      "learning_rate": 0.016866266604651535,
      "loss": 0.1709,
      "num_input_tokens_seen": 29353760,
      "step": 33905
    },
    {
      "epoch": 15.987741631305987,
      "grad_norm": 0.013093075715005398,
      "learning_rate": 0.016839139662137976,
      "loss": 0.1578,
      "num_input_tokens_seen": 29357584,
      "step": 33910
    },
    {
      "epoch": 15.990099009900991,
      "grad_norm": 0.0046911947429180145,
      "learning_rate": 0.01681203325470245,
      "loss": 0.0715,
      "num_input_tokens_seen": 29362144,
      "step": 33915
    },
    {
      "epoch": 15.992456388495992,
      "grad_norm": 0.009837457910180092,
      "learning_rate": 0.016784947386525157,
      "loss": 0.1248,
      "num_input_tokens_seen": 29367296,
      "step": 33920
    },
    {
      "epoch": 15.994813767090994,
      "grad_norm": 0.008478086441755295,
      "learning_rate": 0.01675788206178308,
      "loss": 0.131,
      "num_input_tokens_seen": 29371600,
      "step": 33925
    },
    {
      "epoch": 15.997171145685996,
      "grad_norm": 0.011700219474732876,
      "learning_rate": 0.016730837284649986,
      "loss": 0.1461,
      "num_input_tokens_seen": 29376656,
      "step": 33930
    },
    {
      "epoch": 15.999528524280999,
      "grad_norm": 0.013135616667568684,
      "learning_rate": 0.016703813059296583,
      "loss": 0.1576,
      "num_input_tokens_seen": 29379904,
      "step": 33935
    },
    {
      "epoch": 16.001885902876,
      "grad_norm": 0.0085666598752141,
      "learning_rate": 0.016676809389890294,
      "loss": 0.1269,
      "num_input_tokens_seen": 29384608,
      "step": 33940
    },
    {
      "epoch": 16.004243281471005,
      "grad_norm": 0.006550874561071396,
      "learning_rate": 0.016649826280595435,
      "loss": 0.0718,
      "num_input_tokens_seen": 29388624,
      "step": 33945
    },
    {
      "epoch": 16.006600660066006,
      "grad_norm": 0.014352720230817795,
      "learning_rate": 0.016622863735573163,
      "loss": 0.1668,
      "num_input_tokens_seen": 29391808,
      "step": 33950
    },
    {
      "epoch": 16.00895803866101,
      "grad_norm": 0.009166819974780083,
      "learning_rate": 0.016595921758981395,
      "loss": 0.1297,
      "num_input_tokens_seen": 29395840,
      "step": 33955
    },
    {
      "epoch": 16.01131541725601,
      "grad_norm": 0.0112850870937109,
      "learning_rate": 0.01656900035497495,
      "loss": 0.0972,
      "num_input_tokens_seen": 29400608,
      "step": 33960
    },
    {
      "epoch": 16.013672795851015,
      "grad_norm": 0.007533607073128223,
      "learning_rate": 0.016542099527705485,
      "loss": 0.1238,
      "num_input_tokens_seen": 29404592,
      "step": 33965
    },
    {
      "epoch": 16.016030174446016,
      "grad_norm": 0.00611309427767992,
      "learning_rate": 0.01651521928132138,
      "loss": 0.158,
      "num_input_tokens_seen": 29409552,
      "step": 33970
    },
    {
      "epoch": 16.01838755304102,
      "grad_norm": 0.01574980467557907,
      "learning_rate": 0.01648835961996794,
      "loss": 0.2151,
      "num_input_tokens_seen": 29413168,
      "step": 33975
    },
    {
      "epoch": 16.02074493163602,
      "grad_norm": 0.010679124854505062,
      "learning_rate": 0.016461520547787285,
      "loss": 0.1013,
      "num_input_tokens_seen": 29418032,
      "step": 33980
    },
    {
      "epoch": 16.023102310231025,
      "grad_norm": 0.006675763987004757,
      "learning_rate": 0.016434702068918266,
      "loss": 0.0614,
      "num_input_tokens_seen": 29422608,
      "step": 33985
    },
    {
      "epoch": 16.025459688826025,
      "grad_norm": 0.011866366490721703,
      "learning_rate": 0.01640790418749673,
      "loss": 0.1228,
      "num_input_tokens_seen": 29426800,
      "step": 33990
    },
    {
      "epoch": 16.02781706742103,
      "grad_norm": 0.012925009243190289,
      "learning_rate": 0.016381126907655134,
      "loss": 0.1378,
      "num_input_tokens_seen": 29431328,
      "step": 33995
    },
    {
      "epoch": 16.03017444601603,
      "grad_norm": 0.00643345108255744,
      "learning_rate": 0.016354370233522948,
      "loss": 0.1113,
      "num_input_tokens_seen": 29435360,
      "step": 34000
    },
    {
      "epoch": 16.03017444601603,
      "eval_loss": 0.2413286715745926,
      "eval_runtime": 21.9184,
      "eval_samples_per_second": 43.023,
      "eval_steps_per_second": 21.534,
      "num_input_tokens_seen": 29435360,
      "step": 34000
    },
    {
      "epoch": 16.032531824611034,
      "grad_norm": 0.0017099065007641912,
      "learning_rate": 0.016327634169226394,
      "loss": 0.0813,
      "num_input_tokens_seen": 29439888,
      "step": 34005
    },
    {
      "epoch": 16.034889203206035,
      "grad_norm": 0.015047271735966206,
      "learning_rate": 0.016300918718888485,
      "loss": 0.1037,
      "num_input_tokens_seen": 29444192,
      "step": 34010
    },
    {
      "epoch": 16.03724658180104,
      "grad_norm": 0.011723118834197521,
      "learning_rate": 0.016274223886629052,
      "loss": 0.1355,
      "num_input_tokens_seen": 29448560,
      "step": 34015
    },
    {
      "epoch": 16.03960396039604,
      "grad_norm": 0.012239443138241768,
      "learning_rate": 0.01624754967656482,
      "loss": 0.1516,
      "num_input_tokens_seen": 29452960,
      "step": 34020
    },
    {
      "epoch": 16.04196133899104,
      "grad_norm": 0.008117412216961384,
      "learning_rate": 0.016220896092809235,
      "loss": 0.1357,
      "num_input_tokens_seen": 29458400,
      "step": 34025
    },
    {
      "epoch": 16.044318717586044,
      "grad_norm": 0.021050719544291496,
      "learning_rate": 0.01619426313947267,
      "loss": 0.1378,
      "num_input_tokens_seen": 29462000,
      "step": 34030
    },
    {
      "epoch": 16.046676096181045,
      "grad_norm": 0.0037561622448265553,
      "learning_rate": 0.016167650820662228,
      "loss": 0.1174,
      "num_input_tokens_seen": 29465824,
      "step": 34035
    },
    {
      "epoch": 16.04903347477605,
      "grad_norm": 0.00926631223410368,
      "learning_rate": 0.016141059140481855,
      "loss": 0.1479,
      "num_input_tokens_seen": 29471280,
      "step": 34040
    },
    {
      "epoch": 16.05139085337105,
      "grad_norm": 0.005544474348425865,
      "learning_rate": 0.016114488103032374,
      "loss": 0.1105,
      "num_input_tokens_seen": 29475728,
      "step": 34045
    },
    {
      "epoch": 16.053748231966054,
      "grad_norm": 0.009672624059021473,
      "learning_rate": 0.016087937712411293,
      "loss": 0.0985,
      "num_input_tokens_seen": 29479824,
      "step": 34050
    },
    {
      "epoch": 16.056105610561055,
      "grad_norm": 0.0070793721824884415,
      "learning_rate": 0.01606140797271308,
      "loss": 0.083,
      "num_input_tokens_seen": 29484112,
      "step": 34055
    },
    {
      "epoch": 16.05846298915606,
      "grad_norm": 0.015368602238595486,
      "learning_rate": 0.01603489888802897,
      "loss": 0.0985,
      "num_input_tokens_seen": 29488272,
      "step": 34060
    },
    {
      "epoch": 16.06082036775106,
      "grad_norm": 0.009617436677217484,
      "learning_rate": 0.016008410462446918,
      "loss": 0.1189,
      "num_input_tokens_seen": 29493088,
      "step": 34065
    },
    {
      "epoch": 16.063177746346064,
      "grad_norm": 0.014716090634465218,
      "learning_rate": 0.01598194270005185,
      "loss": 0.1195,
      "num_input_tokens_seen": 29497168,
      "step": 34070
    },
    {
      "epoch": 16.065535124941064,
      "grad_norm": 0.009490973316133022,
      "learning_rate": 0.015955495604925356,
      "loss": 0.0911,
      "num_input_tokens_seen": 29501760,
      "step": 34075
    },
    {
      "epoch": 16.06789250353607,
      "grad_norm": 0.008850552141666412,
      "learning_rate": 0.01592906918114598,
      "loss": 0.0988,
      "num_input_tokens_seen": 29506336,
      "step": 34080
    },
    {
      "epoch": 16.07024988213107,
      "grad_norm": 0.013250972144305706,
      "learning_rate": 0.015902663432788965,
      "loss": 0.1062,
      "num_input_tokens_seen": 29511888,
      "step": 34085
    },
    {
      "epoch": 16.072607260726073,
      "grad_norm": 0.011141255497932434,
      "learning_rate": 0.01587627836392643,
      "loss": 0.0778,
      "num_input_tokens_seen": 29516096,
      "step": 34090
    },
    {
      "epoch": 16.074964639321074,
      "grad_norm": 0.022495299577713013,
      "learning_rate": 0.01584991397862726,
      "loss": 0.1076,
      "num_input_tokens_seen": 29520144,
      "step": 34095
    },
    {
      "epoch": 16.077322017916078,
      "grad_norm": 0.007952597923576832,
      "learning_rate": 0.015823570280957214,
      "loss": 0.121,
      "num_input_tokens_seen": 29524640,
      "step": 34100
    },
    {
      "epoch": 16.07967939651108,
      "grad_norm": 0.009321251884102821,
      "learning_rate": 0.015797247274978766,
      "loss": 0.1009,
      "num_input_tokens_seen": 29529504,
      "step": 34105
    },
    {
      "epoch": 16.082036775106083,
      "grad_norm": 0.008311799727380276,
      "learning_rate": 0.015770944964751326,
      "loss": 0.1105,
      "num_input_tokens_seen": 29533328,
      "step": 34110
    },
    {
      "epoch": 16.084394153701083,
      "grad_norm": 0.01356179267168045,
      "learning_rate": 0.015744663354330956,
      "loss": 0.1256,
      "num_input_tokens_seen": 29538080,
      "step": 34115
    },
    {
      "epoch": 16.086751532296088,
      "grad_norm": 0.005060568451881409,
      "learning_rate": 0.015718402447770664,
      "loss": 0.1516,
      "num_input_tokens_seen": 29542032,
      "step": 34120
    },
    {
      "epoch": 16.08910891089109,
      "grad_norm": 0.012311452999711037,
      "learning_rate": 0.015692162249120224,
      "loss": 0.1631,
      "num_input_tokens_seen": 29546432,
      "step": 34125
    },
    {
      "epoch": 16.091466289486092,
      "grad_norm": 0.020479561761021614,
      "learning_rate": 0.01566594276242615,
      "loss": 0.1478,
      "num_input_tokens_seen": 29550944,
      "step": 34130
    },
    {
      "epoch": 16.093823668081093,
      "grad_norm": 0.007709856610745192,
      "learning_rate": 0.015639743991731857,
      "loss": 0.1055,
      "num_input_tokens_seen": 29555520,
      "step": 34135
    },
    {
      "epoch": 16.096181046676097,
      "grad_norm": 0.012737628072500229,
      "learning_rate": 0.01561356594107755,
      "loss": 0.1283,
      "num_input_tokens_seen": 29559856,
      "step": 34140
    },
    {
      "epoch": 16.098538425271098,
      "grad_norm": 0.005424327217042446,
      "learning_rate": 0.015587408614500147,
      "loss": 0.1503,
      "num_input_tokens_seen": 29563696,
      "step": 34145
    },
    {
      "epoch": 16.100895803866102,
      "grad_norm": 0.01442993525415659,
      "learning_rate": 0.015561272016033505,
      "loss": 0.1722,
      "num_input_tokens_seen": 29569152,
      "step": 34150
    },
    {
      "epoch": 16.103253182461103,
      "grad_norm": 0.004335211124271154,
      "learning_rate": 0.015535156149708167,
      "loss": 0.0425,
      "num_input_tokens_seen": 29573120,
      "step": 34155
    },
    {
      "epoch": 16.105610561056107,
      "grad_norm": 0.003364636329934001,
      "learning_rate": 0.015509061019551528,
      "loss": 0.0883,
      "num_input_tokens_seen": 29577456,
      "step": 34160
    },
    {
      "epoch": 16.107967939651108,
      "grad_norm": 0.019437070935964584,
      "learning_rate": 0.015482986629587818,
      "loss": 0.1475,
      "num_input_tokens_seen": 29581424,
      "step": 34165
    },
    {
      "epoch": 16.11032531824611,
      "grad_norm": 0.007611352950334549,
      "learning_rate": 0.01545693298383799,
      "loss": 0.1431,
      "num_input_tokens_seen": 29585872,
      "step": 34170
    },
    {
      "epoch": 16.112682696841112,
      "grad_norm": 0.015791453421115875,
      "learning_rate": 0.015430900086319858,
      "loss": 0.1353,
      "num_input_tokens_seen": 29589984,
      "step": 34175
    },
    {
      "epoch": 16.115040075436116,
      "grad_norm": 0.005937471054494381,
      "learning_rate": 0.015404887941048084,
      "loss": 0.0874,
      "num_input_tokens_seen": 29594672,
      "step": 34180
    },
    {
      "epoch": 16.117397454031117,
      "grad_norm": 0.005718894302845001,
      "learning_rate": 0.01537889655203397,
      "loss": 0.1511,
      "num_input_tokens_seen": 29598992,
      "step": 34185
    },
    {
      "epoch": 16.11975483262612,
      "grad_norm": 0.005655954591929913,
      "learning_rate": 0.015352925923285798,
      "loss": 0.1737,
      "num_input_tokens_seen": 29602528,
      "step": 34190
    },
    {
      "epoch": 16.122112211221122,
      "grad_norm": 0.0035175681114196777,
      "learning_rate": 0.015326976058808511,
      "loss": 0.0605,
      "num_input_tokens_seen": 29606384,
      "step": 34195
    },
    {
      "epoch": 16.124469589816126,
      "grad_norm": 0.011124420911073685,
      "learning_rate": 0.015301046962603908,
      "loss": 0.1501,
      "num_input_tokens_seen": 29610720,
      "step": 34200
    },
    {
      "epoch": 16.124469589816126,
      "eval_loss": 0.2551458179950714,
      "eval_runtime": 21.8876,
      "eval_samples_per_second": 43.084,
      "eval_steps_per_second": 21.565,
      "num_input_tokens_seen": 29610720,
      "step": 34200
    },
    {
      "epoch": 16.126826968411127,
      "grad_norm": 0.011172446422278881,
      "learning_rate": 0.015275138638670626,
      "loss": 0.1045,
      "num_input_tokens_seen": 29615744,
      "step": 34205
    },
    {
      "epoch": 16.12918434700613,
      "grad_norm": 0.003085650037974119,
      "learning_rate": 0.015249251091004001,
      "loss": 0.1339,
      "num_input_tokens_seen": 29619392,
      "step": 34210
    },
    {
      "epoch": 16.13154172560113,
      "grad_norm": 0.01371519360691309,
      "learning_rate": 0.01522338432359624,
      "loss": 0.161,
      "num_input_tokens_seen": 29623248,
      "step": 34215
    },
    {
      "epoch": 16.133899104196132,
      "grad_norm": 0.020540283992886543,
      "learning_rate": 0.01519753834043635,
      "loss": 0.1606,
      "num_input_tokens_seen": 29626816,
      "step": 34220
    },
    {
      "epoch": 16.136256482791136,
      "grad_norm": 0.020900508388876915,
      "learning_rate": 0.015171713145510095,
      "loss": 0.2167,
      "num_input_tokens_seen": 29632784,
      "step": 34225
    },
    {
      "epoch": 16.138613861386137,
      "grad_norm": 0.005213696975260973,
      "learning_rate": 0.01514590874279999,
      "loss": 0.0919,
      "num_input_tokens_seen": 29637200,
      "step": 34230
    },
    {
      "epoch": 16.14097123998114,
      "grad_norm": 0.004236621782183647,
      "learning_rate": 0.015120125136285467,
      "loss": 0.0758,
      "num_input_tokens_seen": 29641216,
      "step": 34235
    },
    {
      "epoch": 16.14332861857614,
      "grad_norm": 0.008980554528534412,
      "learning_rate": 0.015094362329942629,
      "loss": 0.1154,
      "num_input_tokens_seen": 29645856,
      "step": 34240
    },
    {
      "epoch": 16.145685997171146,
      "grad_norm": 0.014634541235864162,
      "learning_rate": 0.01506862032774448,
      "loss": 0.1558,
      "num_input_tokens_seen": 29649776,
      "step": 34245
    },
    {
      "epoch": 16.148043375766147,
      "grad_norm": 0.025339709594845772,
      "learning_rate": 0.015042899133660697,
      "loss": 0.1495,
      "num_input_tokens_seen": 29654016,
      "step": 34250
    },
    {
      "epoch": 16.15040075436115,
      "grad_norm": 0.01310226134955883,
      "learning_rate": 0.01501719875165789,
      "loss": 0.209,
      "num_input_tokens_seen": 29657952,
      "step": 34255
    },
    {
      "epoch": 16.15275813295615,
      "grad_norm": 0.018156424164772034,
      "learning_rate": 0.014991519185699286,
      "loss": 0.161,
      "num_input_tokens_seen": 29661968,
      "step": 34260
    },
    {
      "epoch": 16.155115511551156,
      "grad_norm": 0.018348902463912964,
      "learning_rate": 0.014965860439745054,
      "loss": 0.148,
      "num_input_tokens_seen": 29666480,
      "step": 34265
    },
    {
      "epoch": 16.157472890146156,
      "grad_norm": 0.019490554928779602,
      "learning_rate": 0.01494022251775211,
      "loss": 0.1927,
      "num_input_tokens_seen": 29670512,
      "step": 34270
    },
    {
      "epoch": 16.15983026874116,
      "grad_norm": 0.009946472942829132,
      "learning_rate": 0.014914605423674109,
      "loss": 0.1582,
      "num_input_tokens_seen": 29674432,
      "step": 34275
    },
    {
      "epoch": 16.16218764733616,
      "grad_norm": 0.010929797776043415,
      "learning_rate": 0.014889009161461525,
      "loss": 0.15,
      "num_input_tokens_seen": 29678720,
      "step": 34280
    },
    {
      "epoch": 16.164545025931165,
      "grad_norm": 0.008386784233152866,
      "learning_rate": 0.014863433735061665,
      "loss": 0.1338,
      "num_input_tokens_seen": 29683696,
      "step": 34285
    },
    {
      "epoch": 16.166902404526166,
      "grad_norm": 0.008403047919273376,
      "learning_rate": 0.014837879148418541,
      "loss": 0.1101,
      "num_input_tokens_seen": 29687824,
      "step": 34290
    },
    {
      "epoch": 16.16925978312117,
      "grad_norm": 0.009904601611196995,
      "learning_rate": 0.01481234540547302,
      "loss": 0.1115,
      "num_input_tokens_seen": 29692160,
      "step": 34295
    },
    {
      "epoch": 16.17161716171617,
      "grad_norm": 0.02155277132987976,
      "learning_rate": 0.014786832510162717,
      "loss": 0.1864,
      "num_input_tokens_seen": 29696304,
      "step": 34300
    },
    {
      "epoch": 16.173974540311175,
      "grad_norm": 0.008372167125344276,
      "learning_rate": 0.014761340466422017,
      "loss": 0.1728,
      "num_input_tokens_seen": 29700208,
      "step": 34305
    },
    {
      "epoch": 16.176331918906175,
      "grad_norm": 0.020202049985527992,
      "learning_rate": 0.014735869278182144,
      "loss": 0.1581,
      "num_input_tokens_seen": 29703920,
      "step": 34310
    },
    {
      "epoch": 16.17868929750118,
      "grad_norm": 0.008375285193324089,
      "learning_rate": 0.014710418949371057,
      "loss": 0.0898,
      "num_input_tokens_seen": 29708400,
      "step": 34315
    },
    {
      "epoch": 16.18104667609618,
      "grad_norm": 0.01661125011742115,
      "learning_rate": 0.014684989483913495,
      "loss": 0.1036,
      "num_input_tokens_seen": 29712656,
      "step": 34320
    },
    {
      "epoch": 16.183404054691184,
      "grad_norm": 0.024773744866251945,
      "learning_rate": 0.014659580885731077,
      "loss": 0.1323,
      "num_input_tokens_seen": 29717376,
      "step": 34325
    },
    {
      "epoch": 16.185761433286185,
      "grad_norm": 0.0049279904924333096,
      "learning_rate": 0.014634193158742047,
      "loss": 0.117,
      "num_input_tokens_seen": 29721200,
      "step": 34330
    },
    {
      "epoch": 16.18811881188119,
      "grad_norm": 0.009918139316141605,
      "learning_rate": 0.014608826306861576,
      "loss": 0.1296,
      "num_input_tokens_seen": 29725472,
      "step": 34335
    },
    {
      "epoch": 16.19047619047619,
      "grad_norm": 0.003971131052821875,
      "learning_rate": 0.014583480334001486,
      "loss": 0.139,
      "num_input_tokens_seen": 29729376,
      "step": 34340
    },
    {
      "epoch": 16.192833569071194,
      "grad_norm": 0.007473228499293327,
      "learning_rate": 0.014558155244070496,
      "loss": 0.1976,
      "num_input_tokens_seen": 29733648,
      "step": 34345
    },
    {
      "epoch": 16.195190947666195,
      "grad_norm": 0.006171034183353186,
      "learning_rate": 0.014532851040974036,
      "loss": 0.1324,
      "num_input_tokens_seen": 29737168,
      "step": 34350
    },
    {
      "epoch": 16.1975483262612,
      "grad_norm": 0.014796918258070946,
      "learning_rate": 0.014507567728614335,
      "loss": 0.0887,
      "num_input_tokens_seen": 29742768,
      "step": 34355
    },
    {
      "epoch": 16.1999057048562,
      "grad_norm": 0.004246920812875032,
      "learning_rate": 0.01448230531089037,
      "loss": 0.085,
      "num_input_tokens_seen": 29747328,
      "step": 34360
    },
    {
      "epoch": 16.202263083451204,
      "grad_norm": 0.002736500231549144,
      "learning_rate": 0.014457063791697993,
      "loss": 0.1054,
      "num_input_tokens_seen": 29751328,
      "step": 34365
    },
    {
      "epoch": 16.204620462046204,
      "grad_norm": 0.008553816936910152,
      "learning_rate": 0.01443184317492971,
      "loss": 0.1654,
      "num_input_tokens_seen": 29755504,
      "step": 34370
    },
    {
      "epoch": 16.20697784064121,
      "grad_norm": 0.011818241328001022,
      "learning_rate": 0.014406643464474822,
      "loss": 0.1231,
      "num_input_tokens_seen": 29760336,
      "step": 34375
    },
    {
      "epoch": 16.20933521923621,
      "grad_norm": 0.010860808193683624,
      "learning_rate": 0.014381464664219539,
      "loss": 0.0831,
      "num_input_tokens_seen": 29764752,
      "step": 34380
    },
    {
      "epoch": 16.211692597831213,
      "grad_norm": 0.009971903637051582,
      "learning_rate": 0.014356306778046656,
      "loss": 0.0758,
      "num_input_tokens_seen": 29768816,
      "step": 34385
    },
    {
      "epoch": 16.214049976426214,
      "grad_norm": 0.021600401028990746,
      "learning_rate": 0.014331169809835885,
      "loss": 0.1825,
      "num_input_tokens_seen": 29772992,
      "step": 34390
    },
    {
      "epoch": 16.216407355021218,
      "grad_norm": 0.014235047623515129,
      "learning_rate": 0.014306053763463644,
      "loss": 0.153,
      "num_input_tokens_seen": 29776848,
      "step": 34395
    },
    {
      "epoch": 16.21876473361622,
      "grad_norm": 0.009120653383433819,
      "learning_rate": 0.014280958642803147,
      "loss": 0.1441,
      "num_input_tokens_seen": 29781472,
      "step": 34400
    },
    {
      "epoch": 16.21876473361622,
      "eval_loss": 0.24954015016555786,
      "eval_runtime": 21.9481,
      "eval_samples_per_second": 42.965,
      "eval_steps_per_second": 21.505,
      "num_input_tokens_seen": 29781472,
      "step": 34400
    },
    {
      "epoch": 16.221122112211223,
      "grad_norm": 0.00858235638588667,
      "learning_rate": 0.014255884451724404,
      "loss": 0.0996,
      "num_input_tokens_seen": 29786416,
      "step": 34405
    },
    {
      "epoch": 16.223479490806223,
      "grad_norm": 0.010800439864397049,
      "learning_rate": 0.014230831194094101,
      "loss": 0.0756,
      "num_input_tokens_seen": 29790480,
      "step": 34410
    },
    {
      "epoch": 16.225836869401228,
      "grad_norm": 0.008445179089903831,
      "learning_rate": 0.014205798873775865,
      "loss": 0.1106,
      "num_input_tokens_seen": 29795136,
      "step": 34415
    },
    {
      "epoch": 16.22819424799623,
      "grad_norm": 0.01712847501039505,
      "learning_rate": 0.014180787494629893,
      "loss": 0.0673,
      "num_input_tokens_seen": 29800080,
      "step": 34420
    },
    {
      "epoch": 16.23055162659123,
      "grad_norm": 0.0026848330162465572,
      "learning_rate": 0.014155797060513314,
      "loss": 0.0754,
      "num_input_tokens_seen": 29804224,
      "step": 34425
    },
    {
      "epoch": 16.232909005186233,
      "grad_norm": 0.021138524636626244,
      "learning_rate": 0.014130827575279963,
      "loss": 0.1533,
      "num_input_tokens_seen": 29808896,
      "step": 34430
    },
    {
      "epoch": 16.235266383781234,
      "grad_norm": 0.00976987462490797,
      "learning_rate": 0.014105879042780427,
      "loss": 0.0743,
      "num_input_tokens_seen": 29813664,
      "step": 34435
    },
    {
      "epoch": 16.237623762376238,
      "grad_norm": 0.004852553363889456,
      "learning_rate": 0.014080951466862113,
      "loss": 0.0926,
      "num_input_tokens_seen": 29817184,
      "step": 34440
    },
    {
      "epoch": 16.23998114097124,
      "grad_norm": 0.004030173644423485,
      "learning_rate": 0.014056044851369126,
      "loss": 0.0587,
      "num_input_tokens_seen": 29821408,
      "step": 34445
    },
    {
      "epoch": 16.242338519566243,
      "grad_norm": 0.0057837339118123055,
      "learning_rate": 0.014031159200142428,
      "loss": 0.127,
      "num_input_tokens_seen": 29826368,
      "step": 34450
    },
    {
      "epoch": 16.244695898161243,
      "grad_norm": 0.014906048774719238,
      "learning_rate": 0.014006294517019667,
      "loss": 0.1672,
      "num_input_tokens_seen": 29830912,
      "step": 34455
    },
    {
      "epoch": 16.247053276756247,
      "grad_norm": 0.0037369453348219395,
      "learning_rate": 0.013981450805835276,
      "loss": 0.1381,
      "num_input_tokens_seen": 29835872,
      "step": 34460
    },
    {
      "epoch": 16.249410655351248,
      "grad_norm": 0.012458516284823418,
      "learning_rate": 0.01395662807042049,
      "loss": 0.1128,
      "num_input_tokens_seen": 29839936,
      "step": 34465
    },
    {
      "epoch": 16.251768033946252,
      "grad_norm": 0.006020260043442249,
      "learning_rate": 0.013931826314603296,
      "loss": 0.1025,
      "num_input_tokens_seen": 29844016,
      "step": 34470
    },
    {
      "epoch": 16.254125412541253,
      "grad_norm": 0.006804877892136574,
      "learning_rate": 0.013907045542208401,
      "loss": 0.1231,
      "num_input_tokens_seen": 29848512,
      "step": 34475
    },
    {
      "epoch": 16.256482791136257,
      "grad_norm": 0.017228273674845695,
      "learning_rate": 0.013882285757057333,
      "loss": 0.2225,
      "num_input_tokens_seen": 29853648,
      "step": 34480
    },
    {
      "epoch": 16.258840169731258,
      "grad_norm": 0.009123594500124454,
      "learning_rate": 0.013857546962968403,
      "loss": 0.1231,
      "num_input_tokens_seen": 29857552,
      "step": 34485
    },
    {
      "epoch": 16.261197548326262,
      "grad_norm": 0.0066408757120370865,
      "learning_rate": 0.013832829163756577,
      "loss": 0.0607,
      "num_input_tokens_seen": 29861472,
      "step": 34490
    },
    {
      "epoch": 16.263554926921262,
      "grad_norm": 0.012659543193876743,
      "learning_rate": 0.013808132363233689,
      "loss": 0.1055,
      "num_input_tokens_seen": 29865616,
      "step": 34495
    },
    {
      "epoch": 16.265912305516267,
      "grad_norm": 0.008858910761773586,
      "learning_rate": 0.013783456565208256,
      "loss": 0.0772,
      "num_input_tokens_seen": 29870112,
      "step": 34500
    },
    {
      "epoch": 16.268269684111267,
      "grad_norm": 0.016577212139964104,
      "learning_rate": 0.01375880177348564,
      "loss": 0.1375,
      "num_input_tokens_seen": 29873840,
      "step": 34505
    },
    {
      "epoch": 16.27062706270627,
      "grad_norm": 0.009850556030869484,
      "learning_rate": 0.013734167991867928,
      "loss": 0.1311,
      "num_input_tokens_seen": 29878320,
      "step": 34510
    },
    {
      "epoch": 16.272984441301272,
      "grad_norm": 0.0062917061150074005,
      "learning_rate": 0.013709555224153935,
      "loss": 0.1423,
      "num_input_tokens_seen": 29883280,
      "step": 34515
    },
    {
      "epoch": 16.275341819896276,
      "grad_norm": 0.0030026084277778864,
      "learning_rate": 0.013684963474139222,
      "loss": 0.1527,
      "num_input_tokens_seen": 29887600,
      "step": 34520
    },
    {
      "epoch": 16.277699198491277,
      "grad_norm": 0.01357159111648798,
      "learning_rate": 0.013660392745616224,
      "loss": 0.1067,
      "num_input_tokens_seen": 29891840,
      "step": 34525
    },
    {
      "epoch": 16.28005657708628,
      "grad_norm": 0.008259536698460579,
      "learning_rate": 0.013635843042373974,
      "loss": 0.1329,
      "num_input_tokens_seen": 29896608,
      "step": 34530
    },
    {
      "epoch": 16.28241395568128,
      "grad_norm": 0.004500230774283409,
      "learning_rate": 0.01361131436819843,
      "loss": 0.1115,
      "num_input_tokens_seen": 29901792,
      "step": 34535
    },
    {
      "epoch": 16.284771334276286,
      "grad_norm": 0.01196984015405178,
      "learning_rate": 0.013586806726872147,
      "loss": 0.1022,
      "num_input_tokens_seen": 29906048,
      "step": 34540
    },
    {
      "epoch": 16.287128712871286,
      "grad_norm": 0.023855049163103104,
      "learning_rate": 0.013562320122174537,
      "loss": 0.1964,
      "num_input_tokens_seen": 29910000,
      "step": 34545
    },
    {
      "epoch": 16.28948609146629,
      "grad_norm": 0.011452157981693745,
      "learning_rate": 0.013537854557881762,
      "loss": 0.1599,
      "num_input_tokens_seen": 29914608,
      "step": 34550
    },
    {
      "epoch": 16.29184347006129,
      "grad_norm": 0.004847514443099499,
      "learning_rate": 0.013513410037766687,
      "loss": 0.0825,
      "num_input_tokens_seen": 29918528,
      "step": 34555
    },
    {
      "epoch": 16.294200848656295,
      "grad_norm": 0.013607257045805454,
      "learning_rate": 0.013488986565598998,
      "loss": 0.1309,
      "num_input_tokens_seen": 29923232,
      "step": 34560
    },
    {
      "epoch": 16.296558227251296,
      "grad_norm": 0.00858873501420021,
      "learning_rate": 0.013464584145145097,
      "loss": 0.1285,
      "num_input_tokens_seen": 29927520,
      "step": 34565
    },
    {
      "epoch": 16.2989156058463,
      "grad_norm": 0.007658522110432386,
      "learning_rate": 0.013440202780168109,
      "loss": 0.1252,
      "num_input_tokens_seen": 29932176,
      "step": 34570
    },
    {
      "epoch": 16.3012729844413,
      "grad_norm": 0.004131216090172529,
      "learning_rate": 0.01341584247442799,
      "loss": 0.1646,
      "num_input_tokens_seen": 29936880,
      "step": 34575
    },
    {
      "epoch": 16.303630363036305,
      "grad_norm": 0.004858119413256645,
      "learning_rate": 0.013391503231681355,
      "loss": 0.0712,
      "num_input_tokens_seen": 29941360,
      "step": 34580
    },
    {
      "epoch": 16.305987741631306,
      "grad_norm": 0.004984191618859768,
      "learning_rate": 0.013367185055681685,
      "loss": 0.0622,
      "num_input_tokens_seen": 29945920,
      "step": 34585
    },
    {
      "epoch": 16.30834512022631,
      "grad_norm": 0.010351119562983513,
      "learning_rate": 0.013342887950179095,
      "loss": 0.1319,
      "num_input_tokens_seen": 29950912,
      "step": 34590
    },
    {
      "epoch": 16.31070249882131,
      "grad_norm": 0.008251049555838108,
      "learning_rate": 0.013318611918920554,
      "loss": 0.083,
      "num_input_tokens_seen": 29955408,
      "step": 34595
    },
    {
      "epoch": 16.313059877416315,
      "grad_norm": 0.017348725348711014,
      "learning_rate": 0.01329435696564965,
      "loss": 0.1606,
      "num_input_tokens_seen": 29959568,
      "step": 34600
    },
    {
      "epoch": 16.313059877416315,
      "eval_loss": 0.25914889574050903,
      "eval_runtime": 21.9054,
      "eval_samples_per_second": 43.049,
      "eval_steps_per_second": 21.547,
      "num_input_tokens_seen": 29959568,
      "step": 34600
    },
    {
      "epoch": 16.315417256011315,
      "grad_norm": 0.017015403136610985,
      "learning_rate": 0.013270123094106894,
      "loss": 0.226,
      "num_input_tokens_seen": 29964400,
      "step": 34605
    },
    {
      "epoch": 16.31777463460632,
      "grad_norm": 0.022247862070798874,
      "learning_rate": 0.013245910308029395,
      "loss": 0.1817,
      "num_input_tokens_seen": 29968304,
      "step": 34610
    },
    {
      "epoch": 16.32013201320132,
      "grad_norm": 0.008536193519830704,
      "learning_rate": 0.0132217186111511,
      "loss": 0.1196,
      "num_input_tokens_seen": 29973008,
      "step": 34615
    },
    {
      "epoch": 16.32248939179632,
      "grad_norm": 0.019585397094488144,
      "learning_rate": 0.013197548007202626,
      "loss": 0.1354,
      "num_input_tokens_seen": 29977168,
      "step": 34620
    },
    {
      "epoch": 16.324846770391325,
      "grad_norm": 0.014536433853209019,
      "learning_rate": 0.01317339849991142,
      "loss": 0.1086,
      "num_input_tokens_seen": 29981760,
      "step": 34625
    },
    {
      "epoch": 16.327204148986326,
      "grad_norm": 0.007709859870374203,
      "learning_rate": 0.013149270093001675,
      "loss": 0.1206,
      "num_input_tokens_seen": 29985456,
      "step": 34630
    },
    {
      "epoch": 16.32956152758133,
      "grad_norm": 0.019150685518980026,
      "learning_rate": 0.013125162790194227,
      "loss": 0.1136,
      "num_input_tokens_seen": 29989616,
      "step": 34635
    },
    {
      "epoch": 16.33191890617633,
      "grad_norm": 0.0024337507784366608,
      "learning_rate": 0.01310107659520674,
      "loss": 0.0821,
      "num_input_tokens_seen": 29994160,
      "step": 34640
    },
    {
      "epoch": 16.334276284771335,
      "grad_norm": 0.022555958479642868,
      "learning_rate": 0.013077011511753655,
      "loss": 0.1087,
      "num_input_tokens_seen": 29998432,
      "step": 34645
    },
    {
      "epoch": 16.336633663366335,
      "grad_norm": 0.020689763128757477,
      "learning_rate": 0.013052967543546056,
      "loss": 0.1549,
      "num_input_tokens_seen": 30002992,
      "step": 34650
    },
    {
      "epoch": 16.33899104196134,
      "grad_norm": 0.009866178967058659,
      "learning_rate": 0.01302894469429186,
      "loss": 0.1435,
      "num_input_tokens_seen": 30006832,
      "step": 34655
    },
    {
      "epoch": 16.34134842055634,
      "grad_norm": 0.012270376086235046,
      "learning_rate": 0.013004942967695653,
      "loss": 0.1944,
      "num_input_tokens_seen": 30010400,
      "step": 34660
    },
    {
      "epoch": 16.343705799151344,
      "grad_norm": 0.00685886712744832,
      "learning_rate": 0.012980962367458859,
      "loss": 0.1062,
      "num_input_tokens_seen": 30014880,
      "step": 34665
    },
    {
      "epoch": 16.346063177746345,
      "grad_norm": 0.005834975279867649,
      "learning_rate": 0.012957002897279567,
      "loss": 0.0517,
      "num_input_tokens_seen": 30019328,
      "step": 34670
    },
    {
      "epoch": 16.34842055634135,
      "grad_norm": 0.01022653840482235,
      "learning_rate": 0.012933064560852576,
      "loss": 0.1214,
      "num_input_tokens_seen": 30023888,
      "step": 34675
    },
    {
      "epoch": 16.35077793493635,
      "grad_norm": 0.008776349015533924,
      "learning_rate": 0.012909147361869527,
      "loss": 0.0944,
      "num_input_tokens_seen": 30030016,
      "step": 34680
    },
    {
      "epoch": 16.353135313531354,
      "grad_norm": 0.001837646821513772,
      "learning_rate": 0.012885251304018774,
      "loss": 0.0604,
      "num_input_tokens_seen": 30034256,
      "step": 34685
    },
    {
      "epoch": 16.355492692126354,
      "grad_norm": 0.018766913563013077,
      "learning_rate": 0.012861376390985335,
      "loss": 0.1473,
      "num_input_tokens_seen": 30039488,
      "step": 34690
    },
    {
      "epoch": 16.35785007072136,
      "grad_norm": 0.01562689244747162,
      "learning_rate": 0.012837522626451063,
      "loss": 0.1066,
      "num_input_tokens_seen": 30044512,
      "step": 34695
    },
    {
      "epoch": 16.36020744931636,
      "grad_norm": 0.0022444245405495167,
      "learning_rate": 0.01281369001409447,
      "loss": 0.0492,
      "num_input_tokens_seen": 30048272,
      "step": 34700
    },
    {
      "epoch": 16.362564827911363,
      "grad_norm": 0.00583273358643055,
      "learning_rate": 0.012789878557590877,
      "loss": 0.1079,
      "num_input_tokens_seen": 30052768,
      "step": 34705
    },
    {
      "epoch": 16.364922206506364,
      "grad_norm": 0.007334201596677303,
      "learning_rate": 0.012766088260612334,
      "loss": 0.1351,
      "num_input_tokens_seen": 30057568,
      "step": 34710
    },
    {
      "epoch": 16.367279585101368,
      "grad_norm": 0.008131795562803745,
      "learning_rate": 0.012742319126827523,
      "loss": 0.0645,
      "num_input_tokens_seen": 30061856,
      "step": 34715
    },
    {
      "epoch": 16.36963696369637,
      "grad_norm": 0.017093287780880928,
      "learning_rate": 0.012718571159902008,
      "loss": 0.1468,
      "num_input_tokens_seen": 30066704,
      "step": 34720
    },
    {
      "epoch": 16.371994342291373,
      "grad_norm": 0.007563186343759298,
      "learning_rate": 0.01269484436349803,
      "loss": 0.1399,
      "num_input_tokens_seen": 30071040,
      "step": 34725
    },
    {
      "epoch": 16.374351720886374,
      "grad_norm": 0.011832297779619694,
      "learning_rate": 0.012671138741274528,
      "loss": 0.1369,
      "num_input_tokens_seen": 30075776,
      "step": 34730
    },
    {
      "epoch": 16.376709099481378,
      "grad_norm": 0.01238256972283125,
      "learning_rate": 0.012647454296887194,
      "loss": 0.0931,
      "num_input_tokens_seen": 30079520,
      "step": 34735
    },
    {
      "epoch": 16.37906647807638,
      "grad_norm": 0.0019004677888005972,
      "learning_rate": 0.012623791033988507,
      "loss": 0.1415,
      "num_input_tokens_seen": 30083888,
      "step": 34740
    },
    {
      "epoch": 16.381423856671383,
      "grad_norm": 0.011642579920589924,
      "learning_rate": 0.012600148956227597,
      "loss": 0.1875,
      "num_input_tokens_seen": 30087936,
      "step": 34745
    },
    {
      "epoch": 16.383781235266383,
      "grad_norm": 0.01616702228784561,
      "learning_rate": 0.012576528067250414,
      "loss": 0.1014,
      "num_input_tokens_seen": 30092272,
      "step": 34750
    },
    {
      "epoch": 16.386138613861387,
      "grad_norm": 0.01834462396800518,
      "learning_rate": 0.012552928370699561,
      "loss": 0.1512,
      "num_input_tokens_seen": 30096144,
      "step": 34755
    },
    {
      "epoch": 16.388495992456388,
      "grad_norm": 0.0014152417425066233,
      "learning_rate": 0.012529349870214411,
      "loss": 0.1572,
      "num_input_tokens_seen": 30100464,
      "step": 34760
    },
    {
      "epoch": 16.390853371051392,
      "grad_norm": 0.014970878139138222,
      "learning_rate": 0.012505792569431106,
      "loss": 0.1194,
      "num_input_tokens_seen": 30104960,
      "step": 34765
    },
    {
      "epoch": 16.393210749646393,
      "grad_norm": 0.016629910096526146,
      "learning_rate": 0.012482256471982422,
      "loss": 0.0894,
      "num_input_tokens_seen": 30109552,
      "step": 34770
    },
    {
      "epoch": 16.395568128241397,
      "grad_norm": 0.009407327510416508,
      "learning_rate": 0.012458741581497956,
      "loss": 0.0917,
      "num_input_tokens_seen": 30113648,
      "step": 34775
    },
    {
      "epoch": 16.397925506836398,
      "grad_norm": 0.008620122447609901,
      "learning_rate": 0.012435247901603974,
      "loss": 0.0731,
      "num_input_tokens_seen": 30118576,
      "step": 34780
    },
    {
      "epoch": 16.400282885431402,
      "grad_norm": 0.006362468469887972,
      "learning_rate": 0.012411775435923528,
      "loss": 0.1279,
      "num_input_tokens_seen": 30122752,
      "step": 34785
    },
    {
      "epoch": 16.402640264026402,
      "grad_norm": 0.016277378425002098,
      "learning_rate": 0.012388324188076354,
      "loss": 0.0719,
      "num_input_tokens_seen": 30127472,
      "step": 34790
    },
    {
      "epoch": 16.404997642621407,
      "grad_norm": 0.003680358175188303,
      "learning_rate": 0.012364894161678913,
      "loss": 0.1331,
      "num_input_tokens_seen": 30131008,
      "step": 34795
    },
    {
      "epoch": 16.407355021216407,
      "grad_norm": 0.01733982376754284,
      "learning_rate": 0.012341485360344445,
      "loss": 0.1382,
      "num_input_tokens_seen": 30134704,
      "step": 34800
    },
    {
      "epoch": 16.407355021216407,
      "eval_loss": 0.26583540439605713,
      "eval_runtime": 21.8943,
      "eval_samples_per_second": 43.071,
      "eval_steps_per_second": 21.558,
      "num_input_tokens_seen": 30134704,
      "step": 34800
    },
    {
      "epoch": 16.40971239981141,
      "grad_norm": 0.017474906519055367,
      "learning_rate": 0.01231809778768283,
      "loss": 0.0805,
      "num_input_tokens_seen": 30138832,
      "step": 34805
    },
    {
      "epoch": 16.412069778406412,
      "grad_norm": 0.006551632657647133,
      "learning_rate": 0.012294731447300799,
      "loss": 0.0908,
      "num_input_tokens_seen": 30143408,
      "step": 34810
    },
    {
      "epoch": 16.414427157001413,
      "grad_norm": 0.00411498686298728,
      "learning_rate": 0.012271386342801671,
      "loss": 0.0802,
      "num_input_tokens_seen": 30148176,
      "step": 34815
    },
    {
      "epoch": 16.416784535596417,
      "grad_norm": 0.0105154262855649,
      "learning_rate": 0.012248062477785565,
      "loss": 0.09,
      "num_input_tokens_seen": 30152240,
      "step": 34820
    },
    {
      "epoch": 16.419141914191417,
      "grad_norm": 0.008995122276246548,
      "learning_rate": 0.012224759855849305,
      "loss": 0.1283,
      "num_input_tokens_seen": 30156832,
      "step": 34825
    },
    {
      "epoch": 16.42149929278642,
      "grad_norm": 0.012466702610254288,
      "learning_rate": 0.012201478480586513,
      "loss": 0.1027,
      "num_input_tokens_seen": 30161280,
      "step": 34830
    },
    {
      "epoch": 16.423856671381422,
      "grad_norm": 0.006256289314478636,
      "learning_rate": 0.012178218355587389,
      "loss": 0.1305,
      "num_input_tokens_seen": 30166240,
      "step": 34835
    },
    {
      "epoch": 16.426214049976426,
      "grad_norm": 0.005141161382198334,
      "learning_rate": 0.01215497948443896,
      "loss": 0.0952,
      "num_input_tokens_seen": 30170624,
      "step": 34840
    },
    {
      "epoch": 16.428571428571427,
      "grad_norm": 0.006387734320014715,
      "learning_rate": 0.012131761870724993,
      "loss": 0.1236,
      "num_input_tokens_seen": 30174816,
      "step": 34845
    },
    {
      "epoch": 16.43092880716643,
      "grad_norm": 0.014991546981036663,
      "learning_rate": 0.012108565518025893,
      "loss": 0.1224,
      "num_input_tokens_seen": 30179520,
      "step": 34850
    },
    {
      "epoch": 16.433286185761432,
      "grad_norm": 0.011947426944971085,
      "learning_rate": 0.012085390429918862,
      "loss": 0.1436,
      "num_input_tokens_seen": 30183328,
      "step": 34855
    },
    {
      "epoch": 16.435643564356436,
      "grad_norm": 0.009861820377409458,
      "learning_rate": 0.012062236609977744,
      "loss": 0.1611,
      "num_input_tokens_seen": 30187344,
      "step": 34860
    },
    {
      "epoch": 16.438000942951437,
      "grad_norm": 0.013246297836303711,
      "learning_rate": 0.01203910406177318,
      "loss": 0.0905,
      "num_input_tokens_seen": 30191168,
      "step": 34865
    },
    {
      "epoch": 16.44035832154644,
      "grad_norm": 0.012861029244959354,
      "learning_rate": 0.01201599278887252,
      "loss": 0.1707,
      "num_input_tokens_seen": 30195776,
      "step": 34870
    },
    {
      "epoch": 16.44271570014144,
      "grad_norm": 0.011228815652430058,
      "learning_rate": 0.011992902794839744,
      "loss": 0.1348,
      "num_input_tokens_seen": 30199904,
      "step": 34875
    },
    {
      "epoch": 16.445073078736446,
      "grad_norm": 0.002383790910243988,
      "learning_rate": 0.011969834083235703,
      "loss": 0.0874,
      "num_input_tokens_seen": 30204144,
      "step": 34880
    },
    {
      "epoch": 16.447430457331446,
      "grad_norm": 0.014992046169936657,
      "learning_rate": 0.011946786657617836,
      "loss": 0.1404,
      "num_input_tokens_seen": 30208304,
      "step": 34885
    },
    {
      "epoch": 16.44978783592645,
      "grad_norm": 0.011577366851270199,
      "learning_rate": 0.011923760521540332,
      "loss": 0.1465,
      "num_input_tokens_seen": 30213600,
      "step": 34890
    },
    {
      "epoch": 16.45214521452145,
      "grad_norm": 0.020672926679253578,
      "learning_rate": 0.011900755678554153,
      "loss": 0.2086,
      "num_input_tokens_seen": 30217296,
      "step": 34895
    },
    {
      "epoch": 16.454502593116455,
      "grad_norm": 0.014294707216322422,
      "learning_rate": 0.011877772132206893,
      "loss": 0.1212,
      "num_input_tokens_seen": 30222096,
      "step": 34900
    },
    {
      "epoch": 16.456859971711456,
      "grad_norm": 0.003543161554262042,
      "learning_rate": 0.011854809886042915,
      "loss": 0.1296,
      "num_input_tokens_seen": 30226448,
      "step": 34905
    },
    {
      "epoch": 16.45921735030646,
      "grad_norm": 0.010926477611064911,
      "learning_rate": 0.011831868943603325,
      "loss": 0.1214,
      "num_input_tokens_seen": 30230176,
      "step": 34910
    },
    {
      "epoch": 16.46157472890146,
      "grad_norm": 0.014689477160573006,
      "learning_rate": 0.011808949308425836,
      "loss": 0.0857,
      "num_input_tokens_seen": 30234080,
      "step": 34915
    },
    {
      "epoch": 16.463932107496465,
      "grad_norm": 0.016087887808680534,
      "learning_rate": 0.01178605098404501,
      "loss": 0.1082,
      "num_input_tokens_seen": 30238112,
      "step": 34920
    },
    {
      "epoch": 16.466289486091465,
      "grad_norm": 0.006383042316883802,
      "learning_rate": 0.011763173973992002,
      "loss": 0.1542,
      "num_input_tokens_seen": 30242384,
      "step": 34925
    },
    {
      "epoch": 16.46864686468647,
      "grad_norm": 0.006607703864574432,
      "learning_rate": 0.011740318281794776,
      "loss": 0.1027,
      "num_input_tokens_seen": 30246880,
      "step": 34930
    },
    {
      "epoch": 16.47100424328147,
      "grad_norm": 0.005224226973950863,
      "learning_rate": 0.01171748391097796,
      "loss": 0.0954,
      "num_input_tokens_seen": 30251216,
      "step": 34935
    },
    {
      "epoch": 16.473361621876474,
      "grad_norm": 0.015584144741296768,
      "learning_rate": 0.011694670865062873,
      "loss": 0.1191,
      "num_input_tokens_seen": 30254464,
      "step": 34940
    },
    {
      "epoch": 16.475719000471475,
      "grad_norm": 0.0060554808005690575,
      "learning_rate": 0.011671879147567616,
      "loss": 0.1256,
      "num_input_tokens_seen": 30259136,
      "step": 34945
    },
    {
      "epoch": 16.47807637906648,
      "grad_norm": 0.008802007883787155,
      "learning_rate": 0.011649108762006893,
      "loss": 0.0973,
      "num_input_tokens_seen": 30263200,
      "step": 34950
    },
    {
      "epoch": 16.48043375766148,
      "grad_norm": 0.013043027371168137,
      "learning_rate": 0.011626359711892265,
      "loss": 0.0796,
      "num_input_tokens_seen": 30268096,
      "step": 34955
    },
    {
      "epoch": 16.482791136256484,
      "grad_norm": 0.0042444998398423195,
      "learning_rate": 0.01160363200073189,
      "loss": 0.1074,
      "num_input_tokens_seen": 30272480,
      "step": 34960
    },
    {
      "epoch": 16.485148514851485,
      "grad_norm": 0.01721043325960636,
      "learning_rate": 0.011580925632030614,
      "loss": 0.1518,
      "num_input_tokens_seen": 30276624,
      "step": 34965
    },
    {
      "epoch": 16.48750589344649,
      "grad_norm": 0.014687670394778252,
      "learning_rate": 0.011558240609290104,
      "loss": 0.1819,
      "num_input_tokens_seen": 30280544,
      "step": 34970
    },
    {
      "epoch": 16.48986327204149,
      "grad_norm": 0.0046437084674835205,
      "learning_rate": 0.011535576936008679,
      "loss": 0.1382,
      "num_input_tokens_seen": 30284112,
      "step": 34975
    },
    {
      "epoch": 16.492220650636494,
      "grad_norm": 0.020965280011296272,
      "learning_rate": 0.011512934615681309,
      "loss": 0.1359,
      "num_input_tokens_seen": 30288672,
      "step": 34980
    },
    {
      "epoch": 16.494578029231494,
      "grad_norm": 0.013096784241497517,
      "learning_rate": 0.011490313651799765,
      "loss": 0.0841,
      "num_input_tokens_seen": 30293120,
      "step": 34985
    },
    {
      "epoch": 16.4969354078265,
      "grad_norm": 0.04140080139040947,
      "learning_rate": 0.011467714047852512,
      "loss": 0.1904,
      "num_input_tokens_seen": 30297040,
      "step": 34990
    },
    {
      "epoch": 16.4992927864215,
      "grad_norm": 0.009016991592943668,
      "learning_rate": 0.011445135807324624,
      "loss": 0.0999,
      "num_input_tokens_seen": 30301328,
      "step": 34995
    },
    {
      "epoch": 16.501650165016503,
      "grad_norm": 0.015973450616002083,
      "learning_rate": 0.011422578933698002,
      "loss": 0.096,
      "num_input_tokens_seen": 30305200,
      "step": 35000
    },
    {
      "epoch": 16.501650165016503,
      "eval_loss": 0.26546040177345276,
      "eval_runtime": 21.8685,
      "eval_samples_per_second": 43.121,
      "eval_steps_per_second": 21.584,
      "num_input_tokens_seen": 30305200,
      "step": 35000
    },
    {
      "epoch": 16.504007543611504,
      "grad_norm": 0.0027341393288224936,
      "learning_rate": 0.011400043430451161,
      "loss": 0.1348,
      "num_input_tokens_seen": 30309360,
      "step": 35005
    },
    {
      "epoch": 16.506364922206508,
      "grad_norm": 0.013696285896003246,
      "learning_rate": 0.011377529301059392,
      "loss": 0.0945,
      "num_input_tokens_seen": 30314256,
      "step": 35010
    },
    {
      "epoch": 16.50872230080151,
      "grad_norm": 0.005790447350591421,
      "learning_rate": 0.011355036548994646,
      "loss": 0.0959,
      "num_input_tokens_seen": 30318544,
      "step": 35015
    },
    {
      "epoch": 16.51107967939651,
      "grad_norm": 0.01679118163883686,
      "learning_rate": 0.011332565177725584,
      "loss": 0.1827,
      "num_input_tokens_seen": 30322432,
      "step": 35020
    },
    {
      "epoch": 16.513437057991514,
      "grad_norm": 0.01996285282075405,
      "learning_rate": 0.011310115190717585,
      "loss": 0.1285,
      "num_input_tokens_seen": 30326176,
      "step": 35025
    },
    {
      "epoch": 16.515794436586514,
      "grad_norm": 0.016931995749473572,
      "learning_rate": 0.01128768659143271,
      "loss": 0.1119,
      "num_input_tokens_seen": 30330944,
      "step": 35030
    },
    {
      "epoch": 16.51815181518152,
      "grad_norm": 0.023037662729620934,
      "learning_rate": 0.011265279383329713,
      "loss": 0.1837,
      "num_input_tokens_seen": 30334944,
      "step": 35035
    },
    {
      "epoch": 16.52050919377652,
      "grad_norm": 0.0054449113085865974,
      "learning_rate": 0.01124289356986411,
      "loss": 0.1025,
      "num_input_tokens_seen": 30338688,
      "step": 35040
    },
    {
      "epoch": 16.522866572371523,
      "grad_norm": 0.0142629723995924,
      "learning_rate": 0.011220529154488023,
      "loss": 0.1037,
      "num_input_tokens_seen": 30342880,
      "step": 35045
    },
    {
      "epoch": 16.525223950966524,
      "grad_norm": 0.014100898988544941,
      "learning_rate": 0.011198186140650346,
      "loss": 0.0923,
      "num_input_tokens_seen": 30347168,
      "step": 35050
    },
    {
      "epoch": 16.527581329561528,
      "grad_norm": 0.019583038985729218,
      "learning_rate": 0.011175864531796685,
      "loss": 0.1557,
      "num_input_tokens_seen": 30351568,
      "step": 35055
    },
    {
      "epoch": 16.52993870815653,
      "grad_norm": 0.007302261423319578,
      "learning_rate": 0.011153564331369258,
      "loss": 0.1035,
      "num_input_tokens_seen": 30356048,
      "step": 35060
    },
    {
      "epoch": 16.532296086751533,
      "grad_norm": 0.014552667737007141,
      "learning_rate": 0.011131285542807078,
      "loss": 0.1251,
      "num_input_tokens_seen": 30360112,
      "step": 35065
    },
    {
      "epoch": 16.534653465346533,
      "grad_norm": 0.018458813428878784,
      "learning_rate": 0.011109028169545815,
      "loss": 0.1578,
      "num_input_tokens_seen": 30364288,
      "step": 35070
    },
    {
      "epoch": 16.537010843941538,
      "grad_norm": 0.004091160837560892,
      "learning_rate": 0.011086792215017804,
      "loss": 0.0844,
      "num_input_tokens_seen": 30369280,
      "step": 35075
    },
    {
      "epoch": 16.539368222536538,
      "grad_norm": 0.014887602999806404,
      "learning_rate": 0.011064577682652137,
      "loss": 0.1375,
      "num_input_tokens_seen": 30373584,
      "step": 35080
    },
    {
      "epoch": 16.541725601131542,
      "grad_norm": 0.017657272517681122,
      "learning_rate": 0.011042384575874559,
      "loss": 0.073,
      "num_input_tokens_seen": 30378016,
      "step": 35085
    },
    {
      "epoch": 16.544082979726543,
      "grad_norm": 0.00416394229978323,
      "learning_rate": 0.011020212898107512,
      "loss": 0.107,
      "num_input_tokens_seen": 30382080,
      "step": 35090
    },
    {
      "epoch": 16.546440358321547,
      "grad_norm": 0.01989159733057022,
      "learning_rate": 0.010998062652770197,
      "loss": 0.2031,
      "num_input_tokens_seen": 30386656,
      "step": 35095
    },
    {
      "epoch": 16.548797736916548,
      "grad_norm": 0.010066955350339413,
      "learning_rate": 0.010975933843278428,
      "loss": 0.1106,
      "num_input_tokens_seen": 30391568,
      "step": 35100
    },
    {
      "epoch": 16.551155115511552,
      "grad_norm": 0.026892730966210365,
      "learning_rate": 0.010953826473044714,
      "loss": 0.1546,
      "num_input_tokens_seen": 30395712,
      "step": 35105
    },
    {
      "epoch": 16.553512494106553,
      "grad_norm": 0.010290831327438354,
      "learning_rate": 0.010931740545478357,
      "loss": 0.1838,
      "num_input_tokens_seen": 30400624,
      "step": 35110
    },
    {
      "epoch": 16.555869872701557,
      "grad_norm": 0.020022213459014893,
      "learning_rate": 0.010909676063985218,
      "loss": 0.1727,
      "num_input_tokens_seen": 30404512,
      "step": 35115
    },
    {
      "epoch": 16.558227251296557,
      "grad_norm": 0.00794756505638361,
      "learning_rate": 0.010887633031967974,
      "loss": 0.1039,
      "num_input_tokens_seen": 30410432,
      "step": 35120
    },
    {
      "epoch": 16.56058462989156,
      "grad_norm": 0.025536654517054558,
      "learning_rate": 0.01086561145282589,
      "loss": 0.1242,
      "num_input_tokens_seen": 30415120,
      "step": 35125
    },
    {
      "epoch": 16.562942008486562,
      "grad_norm": 0.012409742921590805,
      "learning_rate": 0.010843611329954983,
      "loss": 0.1031,
      "num_input_tokens_seen": 30419056,
      "step": 35130
    },
    {
      "epoch": 16.565299387081566,
      "grad_norm": 0.008502207696437836,
      "learning_rate": 0.010821632666747988,
      "loss": 0.094,
      "num_input_tokens_seen": 30423296,
      "step": 35135
    },
    {
      "epoch": 16.567656765676567,
      "grad_norm": 0.00784034002572298,
      "learning_rate": 0.010799675466594244,
      "loss": 0.0763,
      "num_input_tokens_seen": 30427296,
      "step": 35140
    },
    {
      "epoch": 16.57001414427157,
      "grad_norm": 0.018319740891456604,
      "learning_rate": 0.010777739732879826,
      "loss": 0.1287,
      "num_input_tokens_seen": 30431152,
      "step": 35145
    },
    {
      "epoch": 16.572371522866572,
      "grad_norm": 0.009106172248721123,
      "learning_rate": 0.010755825468987562,
      "loss": 0.0909,
      "num_input_tokens_seen": 30435728,
      "step": 35150
    },
    {
      "epoch": 16.574728901461576,
      "grad_norm": 0.010628446005284786,
      "learning_rate": 0.010733932678296814,
      "loss": 0.1358,
      "num_input_tokens_seen": 30440368,
      "step": 35155
    },
    {
      "epoch": 16.577086280056577,
      "grad_norm": 0.016367647796869278,
      "learning_rate": 0.010712061364183817,
      "loss": 0.1205,
      "num_input_tokens_seen": 30444336,
      "step": 35160
    },
    {
      "epoch": 16.57944365865158,
      "grad_norm": 0.0246882364153862,
      "learning_rate": 0.010690211530021337,
      "loss": 0.0974,
      "num_input_tokens_seen": 30448768,
      "step": 35165
    },
    {
      "epoch": 16.58180103724658,
      "grad_norm": 0.011260773055255413,
      "learning_rate": 0.01066838317917893,
      "loss": 0.1205,
      "num_input_tokens_seen": 30453632,
      "step": 35170
    },
    {
      "epoch": 16.584158415841586,
      "grad_norm": 0.015185331925749779,
      "learning_rate": 0.010646576315022787,
      "loss": 0.1065,
      "num_input_tokens_seen": 30458192,
      "step": 35175
    },
    {
      "epoch": 16.586515794436586,
      "grad_norm": 0.019020820036530495,
      "learning_rate": 0.010624790940915785,
      "loss": 0.119,
      "num_input_tokens_seen": 30461632,
      "step": 35180
    },
    {
      "epoch": 16.58887317303159,
      "grad_norm": 0.015028572641313076,
      "learning_rate": 0.0106030270602175,
      "loss": 0.1548,
      "num_input_tokens_seen": 30466064,
      "step": 35185
    },
    {
      "epoch": 16.59123055162659,
      "grad_norm": 0.0103147579357028,
      "learning_rate": 0.010581284676284252,
      "loss": 0.1212,
      "num_input_tokens_seen": 30470256,
      "step": 35190
    },
    {
      "epoch": 16.593587930221595,
      "grad_norm": 0.01812771148979664,
      "learning_rate": 0.010559563792468923,
      "loss": 0.0839,
      "num_input_tokens_seen": 30474096,
      "step": 35195
    },
    {
      "epoch": 16.595945308816596,
      "grad_norm": 0.006949461996555328,
      "learning_rate": 0.010537864412121217,
      "loss": 0.1434,
      "num_input_tokens_seen": 30478576,
      "step": 35200
    },
    {
      "epoch": 16.595945308816596,
      "eval_loss": 0.27300241589546204,
      "eval_runtime": 21.9431,
      "eval_samples_per_second": 42.975,
      "eval_steps_per_second": 21.51,
      "num_input_tokens_seen": 30478576,
      "step": 35200
    },
    {
      "epoch": 16.5983026874116,
      "grad_norm": 0.014065522700548172,
      "learning_rate": 0.010516186538587357,
      "loss": 0.1572,
      "num_input_tokens_seen": 30483200,
      "step": 35205
    },
    {
      "epoch": 16.6006600660066,
      "grad_norm": 0.027471262961626053,
      "learning_rate": 0.01049453017521042,
      "loss": 0.1338,
      "num_input_tokens_seen": 30486960,
      "step": 35210
    },
    {
      "epoch": 16.603017444601605,
      "grad_norm": 0.009548719972372055,
      "learning_rate": 0.010472895325330083,
      "loss": 0.1447,
      "num_input_tokens_seen": 30491792,
      "step": 35215
    },
    {
      "epoch": 16.605374823196605,
      "grad_norm": 0.02331313118338585,
      "learning_rate": 0.010451281992282662,
      "loss": 0.1631,
      "num_input_tokens_seen": 30495712,
      "step": 35220
    },
    {
      "epoch": 16.607732201791606,
      "grad_norm": 0.006815020926296711,
      "learning_rate": 0.01042969017940124,
      "loss": 0.0761,
      "num_input_tokens_seen": 30500096,
      "step": 35225
    },
    {
      "epoch": 16.61008958038661,
      "grad_norm": 0.02551700361073017,
      "learning_rate": 0.01040811989001557,
      "loss": 0.1431,
      "num_input_tokens_seen": 30505152,
      "step": 35230
    },
    {
      "epoch": 16.61244695898161,
      "grad_norm": 0.014257312752306461,
      "learning_rate": 0.010386571127451992,
      "loss": 0.1543,
      "num_input_tokens_seen": 30509504,
      "step": 35235
    },
    {
      "epoch": 16.614804337576615,
      "grad_norm": 0.026753578335046768,
      "learning_rate": 0.010365043895033682,
      "loss": 0.1776,
      "num_input_tokens_seen": 30513616,
      "step": 35240
    },
    {
      "epoch": 16.617161716171616,
      "grad_norm": 0.021302232518792152,
      "learning_rate": 0.010343538196080365,
      "loss": 0.1102,
      "num_input_tokens_seen": 30517792,
      "step": 35245
    },
    {
      "epoch": 16.61951909476662,
      "grad_norm": 0.021725891157984734,
      "learning_rate": 0.010322054033908457,
      "loss": 0.1398,
      "num_input_tokens_seen": 30522896,
      "step": 35250
    },
    {
      "epoch": 16.62187647336162,
      "grad_norm": 0.004303533583879471,
      "learning_rate": 0.010300591411831156,
      "loss": 0.0753,
      "num_input_tokens_seen": 30527232,
      "step": 35255
    },
    {
      "epoch": 16.624233851956625,
      "grad_norm": 0.008826233446598053,
      "learning_rate": 0.010279150333158198,
      "loss": 0.0834,
      "num_input_tokens_seen": 30531872,
      "step": 35260
    },
    {
      "epoch": 16.626591230551625,
      "grad_norm": 0.0048581636510789394,
      "learning_rate": 0.010257730801196107,
      "loss": 0.1003,
      "num_input_tokens_seen": 30536256,
      "step": 35265
    },
    {
      "epoch": 16.62894860914663,
      "grad_norm": 0.005547201726585627,
      "learning_rate": 0.010236332819248056,
      "loss": 0.0819,
      "num_input_tokens_seen": 30540528,
      "step": 35270
    },
    {
      "epoch": 16.63130598774163,
      "grad_norm": 0.012261842377483845,
      "learning_rate": 0.010214956390613854,
      "loss": 0.221,
      "num_input_tokens_seen": 30544784,
      "step": 35275
    },
    {
      "epoch": 16.633663366336634,
      "grad_norm": 0.004975807853043079,
      "learning_rate": 0.010193601518590034,
      "loss": 0.1179,
      "num_input_tokens_seen": 30548768,
      "step": 35280
    },
    {
      "epoch": 16.636020744931635,
      "grad_norm": 0.003555319970473647,
      "learning_rate": 0.010172268206469758,
      "loss": 0.1024,
      "num_input_tokens_seen": 30552656,
      "step": 35285
    },
    {
      "epoch": 16.63837812352664,
      "grad_norm": 0.0189500842243433,
      "learning_rate": 0.010150956457542897,
      "loss": 0.0873,
      "num_input_tokens_seen": 30556960,
      "step": 35290
    },
    {
      "epoch": 16.64073550212164,
      "grad_norm": 0.019520742818713188,
      "learning_rate": 0.010129666275096054,
      "loss": 0.1037,
      "num_input_tokens_seen": 30561088,
      "step": 35295
    },
    {
      "epoch": 16.643092880716644,
      "grad_norm": 0.016071228310465813,
      "learning_rate": 0.010108397662412338,
      "loss": 0.1011,
      "num_input_tokens_seen": 30564768,
      "step": 35300
    },
    {
      "epoch": 16.645450259311644,
      "grad_norm": 0.020050635561347008,
      "learning_rate": 0.010087150622771707,
      "loss": 0.2454,
      "num_input_tokens_seen": 30569632,
      "step": 35305
    },
    {
      "epoch": 16.64780763790665,
      "grad_norm": 0.005001400131732225,
      "learning_rate": 0.010065925159450739,
      "loss": 0.0897,
      "num_input_tokens_seen": 30574224,
      "step": 35310
    },
    {
      "epoch": 16.65016501650165,
      "grad_norm": 0.006814736872911453,
      "learning_rate": 0.010044721275722618,
      "loss": 0.0678,
      "num_input_tokens_seen": 30578512,
      "step": 35315
    },
    {
      "epoch": 16.652522395096653,
      "grad_norm": 0.015536016784608364,
      "learning_rate": 0.01002353897485726,
      "loss": 0.1224,
      "num_input_tokens_seen": 30582864,
      "step": 35320
    },
    {
      "epoch": 16.654879773691654,
      "grad_norm": 0.012418797239661217,
      "learning_rate": 0.010002378260121236,
      "loss": 0.0846,
      "num_input_tokens_seen": 30587216,
      "step": 35325
    },
    {
      "epoch": 16.65723715228666,
      "grad_norm": 0.0077856313437223434,
      "learning_rate": 0.009981239134777786,
      "loss": 0.0882,
      "num_input_tokens_seen": 30591120,
      "step": 35330
    },
    {
      "epoch": 16.65959453088166,
      "grad_norm": 0.02454352378845215,
      "learning_rate": 0.009960121602086884,
      "loss": 0.18,
      "num_input_tokens_seen": 30595472,
      "step": 35335
    },
    {
      "epoch": 16.661951909476663,
      "grad_norm": 0.021875590085983276,
      "learning_rate": 0.009939025665305062,
      "loss": 0.1213,
      "num_input_tokens_seen": 30599776,
      "step": 35340
    },
    {
      "epoch": 16.664309288071664,
      "grad_norm": 0.01761482283473015,
      "learning_rate": 0.009917951327685597,
      "loss": 0.1064,
      "num_input_tokens_seen": 30603616,
      "step": 35345
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.02375260926783085,
      "learning_rate": 0.009896898592478425,
      "loss": 0.0884,
      "num_input_tokens_seen": 30607376,
      "step": 35350
    },
    {
      "epoch": 16.66902404526167,
      "grad_norm": 0.014958813786506653,
      "learning_rate": 0.009875867462930132,
      "loss": 0.1322,
      "num_input_tokens_seen": 30611936,
      "step": 35355
    },
    {
      "epoch": 16.671381423856673,
      "grad_norm": 0.01618078164756298,
      "learning_rate": 0.009854857942284006,
      "loss": 0.2136,
      "num_input_tokens_seen": 30615552,
      "step": 35360
    },
    {
      "epoch": 16.673738802451673,
      "grad_norm": 0.010000337846577168,
      "learning_rate": 0.009833870033779923,
      "loss": 0.0936,
      "num_input_tokens_seen": 30619312,
      "step": 35365
    },
    {
      "epoch": 16.676096181046677,
      "grad_norm": 0.004790876992046833,
      "learning_rate": 0.009812903740654527,
      "loss": 0.0983,
      "num_input_tokens_seen": 30623456,
      "step": 35370
    },
    {
      "epoch": 16.678453559641678,
      "grad_norm": 0.015369550324976444,
      "learning_rate": 0.009791959066141097,
      "loss": 0.1077,
      "num_input_tokens_seen": 30627392,
      "step": 35375
    },
    {
      "epoch": 16.680810938236682,
      "grad_norm": 0.01710900105535984,
      "learning_rate": 0.009771036013469537,
      "loss": 0.1505,
      "num_input_tokens_seen": 30630864,
      "step": 35380
    },
    {
      "epoch": 16.683168316831683,
      "grad_norm": 0.007363800425082445,
      "learning_rate": 0.00975013458586646,
      "loss": 0.0919,
      "num_input_tokens_seen": 30634592,
      "step": 35385
    },
    {
      "epoch": 16.685525695426687,
      "grad_norm": 0.015121519565582275,
      "learning_rate": 0.009729254786555107,
      "loss": 0.0972,
      "num_input_tokens_seen": 30639120,
      "step": 35390
    },
    {
      "epoch": 16.687883074021688,
      "grad_norm": 0.003069967729970813,
      "learning_rate": 0.009708396618755421,
      "loss": 0.1346,
      "num_input_tokens_seen": 30643104,
      "step": 35395
    },
    {
      "epoch": 16.690240452616692,
      "grad_norm": 0.010551847517490387,
      "learning_rate": 0.009687560085683994,
      "loss": 0.0859,
      "num_input_tokens_seen": 30647744,
      "step": 35400
    },
    {
      "epoch": 16.690240452616692,
      "eval_loss": 0.26213181018829346,
      "eval_runtime": 21.9246,
      "eval_samples_per_second": 43.011,
      "eval_steps_per_second": 21.528,
      "num_input_tokens_seen": 30647744,
      "step": 35400
    },
    {
      "epoch": 16.692597831211692,
      "grad_norm": 0.016356129199266434,
      "learning_rate": 0.009666745190554054,
      "loss": 0.1053,
      "num_input_tokens_seen": 30652144,
      "step": 35405
    },
    {
      "epoch": 16.694955209806697,
      "grad_norm": 0.013923811726272106,
      "learning_rate": 0.009645951936575553,
      "loss": 0.1144,
      "num_input_tokens_seen": 30656496,
      "step": 35410
    },
    {
      "epoch": 16.697312588401697,
      "grad_norm": 0.012098482809960842,
      "learning_rate": 0.00962518032695509,
      "loss": 0.0996,
      "num_input_tokens_seen": 30660528,
      "step": 35415
    },
    {
      "epoch": 16.6996699669967,
      "grad_norm": 0.0025649776216596365,
      "learning_rate": 0.009604430364895855,
      "loss": 0.0971,
      "num_input_tokens_seen": 30665744,
      "step": 35420
    },
    {
      "epoch": 16.702027345591702,
      "grad_norm": 0.021852515637874603,
      "learning_rate": 0.00958370205359777,
      "loss": 0.0825,
      "num_input_tokens_seen": 30669904,
      "step": 35425
    },
    {
      "epoch": 16.704384724186703,
      "grad_norm": 0.01130243856459856,
      "learning_rate": 0.009562995396257445,
      "loss": 0.0917,
      "num_input_tokens_seen": 30673920,
      "step": 35430
    },
    {
      "epoch": 16.706742102781707,
      "grad_norm": 0.01743270270526409,
      "learning_rate": 0.009542310396068026,
      "loss": 0.1056,
      "num_input_tokens_seen": 30678784,
      "step": 35435
    },
    {
      "epoch": 16.709099481376708,
      "grad_norm": 0.016530361026525497,
      "learning_rate": 0.009521647056219495,
      "loss": 0.1001,
      "num_input_tokens_seen": 30683168,
      "step": 35440
    },
    {
      "epoch": 16.71145685997171,
      "grad_norm": 0.023032763972878456,
      "learning_rate": 0.00950100537989832,
      "loss": 0.1294,
      "num_input_tokens_seen": 30687664,
      "step": 35445
    },
    {
      "epoch": 16.713814238566712,
      "grad_norm": 0.014949853532016277,
      "learning_rate": 0.00948038537028772,
      "loss": 0.1489,
      "num_input_tokens_seen": 30691904,
      "step": 35450
    },
    {
      "epoch": 16.716171617161717,
      "grad_norm": 0.004398577380925417,
      "learning_rate": 0.009459787030567617,
      "loss": 0.0698,
      "num_input_tokens_seen": 30695600,
      "step": 35455
    },
    {
      "epoch": 16.718528995756717,
      "grad_norm": 0.010042480193078518,
      "learning_rate": 0.00943921036391449,
      "loss": 0.0964,
      "num_input_tokens_seen": 30699968,
      "step": 35460
    },
    {
      "epoch": 16.72088637435172,
      "grad_norm": 0.009190640412271023,
      "learning_rate": 0.009418655373501483,
      "loss": 0.1267,
      "num_input_tokens_seen": 30703792,
      "step": 35465
    },
    {
      "epoch": 16.723243752946722,
      "grad_norm": 0.017326461151242256,
      "learning_rate": 0.00939812206249851,
      "loss": 0.1226,
      "num_input_tokens_seen": 30708096,
      "step": 35470
    },
    {
      "epoch": 16.725601131541726,
      "grad_norm": 0.018404478207230568,
      "learning_rate": 0.009377610434072004,
      "loss": 0.1431,
      "num_input_tokens_seen": 30713792,
      "step": 35475
    },
    {
      "epoch": 16.727958510136727,
      "grad_norm": 0.010921269655227661,
      "learning_rate": 0.009357120491385167,
      "loss": 0.1303,
      "num_input_tokens_seen": 30718288,
      "step": 35480
    },
    {
      "epoch": 16.73031588873173,
      "grad_norm": 0.024027390405535698,
      "learning_rate": 0.009336652237597743,
      "loss": 0.1221,
      "num_input_tokens_seen": 30722384,
      "step": 35485
    },
    {
      "epoch": 16.73267326732673,
      "grad_norm": 0.005403308197855949,
      "learning_rate": 0.009316205675866251,
      "loss": 0.1742,
      "num_input_tokens_seen": 30726592,
      "step": 35490
    },
    {
      "epoch": 16.735030645921736,
      "grad_norm": 0.015990082174539566,
      "learning_rate": 0.00929578080934379,
      "loss": 0.1412,
      "num_input_tokens_seen": 30730784,
      "step": 35495
    },
    {
      "epoch": 16.737388024516736,
      "grad_norm": 0.011869063600897789,
      "learning_rate": 0.00927537764118012,
      "loss": 0.1431,
      "num_input_tokens_seen": 30734608,
      "step": 35500
    },
    {
      "epoch": 16.73974540311174,
      "grad_norm": 0.010554048232734203,
      "learning_rate": 0.009254996174521678,
      "loss": 0.1018,
      "num_input_tokens_seen": 30738656,
      "step": 35505
    },
    {
      "epoch": 16.74210278170674,
      "grad_norm": 0.003269665874540806,
      "learning_rate": 0.009234636412511531,
      "loss": 0.0592,
      "num_input_tokens_seen": 30742720,
      "step": 35510
    },
    {
      "epoch": 16.744460160301745,
      "grad_norm": 0.01641368679702282,
      "learning_rate": 0.009214298358289418,
      "loss": 0.0878,
      "num_input_tokens_seen": 30747392,
      "step": 35515
    },
    {
      "epoch": 16.746817538896746,
      "grad_norm": 0.01102248951792717,
      "learning_rate": 0.00919398201499173,
      "loss": 0.0769,
      "num_input_tokens_seen": 30752608,
      "step": 35520
    },
    {
      "epoch": 16.74917491749175,
      "grad_norm": 0.0121776657178998,
      "learning_rate": 0.009173687385751495,
      "loss": 0.1128,
      "num_input_tokens_seen": 30757664,
      "step": 35525
    },
    {
      "epoch": 16.75153229608675,
      "grad_norm": 0.005921770818531513,
      "learning_rate": 0.009153414473698407,
      "loss": 0.1375,
      "num_input_tokens_seen": 30761872,
      "step": 35530
    },
    {
      "epoch": 16.753889674681755,
      "grad_norm": 0.01893548108637333,
      "learning_rate": 0.009133163281958784,
      "loss": 0.1208,
      "num_input_tokens_seen": 30766176,
      "step": 35535
    },
    {
      "epoch": 16.756247053276756,
      "grad_norm": 0.009933964349329472,
      "learning_rate": 0.009112933813655627,
      "loss": 0.2145,
      "num_input_tokens_seen": 30770704,
      "step": 35540
    },
    {
      "epoch": 16.75860443187176,
      "grad_norm": 0.01219438947737217,
      "learning_rate": 0.009092726071908573,
      "loss": 0.0722,
      "num_input_tokens_seen": 30775024,
      "step": 35545
    },
    {
      "epoch": 16.76096181046676,
      "grad_norm": 0.014504989609122276,
      "learning_rate": 0.0090725400598339,
      "loss": 0.1209,
      "num_input_tokens_seen": 30779200,
      "step": 35550
    },
    {
      "epoch": 16.763319189061765,
      "grad_norm": 0.012825874611735344,
      "learning_rate": 0.009052375780544563,
      "loss": 0.1016,
      "num_input_tokens_seen": 30783088,
      "step": 35555
    },
    {
      "epoch": 16.765676567656765,
      "grad_norm": 0.01118024718016386,
      "learning_rate": 0.009032233237150144,
      "loss": 0.0723,
      "num_input_tokens_seen": 30787376,
      "step": 35560
    },
    {
      "epoch": 16.76803394625177,
      "grad_norm": 0.0103830611333251,
      "learning_rate": 0.009012112432756875,
      "loss": 0.0592,
      "num_input_tokens_seen": 30791376,
      "step": 35565
    },
    {
      "epoch": 16.77039132484677,
      "grad_norm": 0.008351908065378666,
      "learning_rate": 0.008992013370467605,
      "loss": 0.0842,
      "num_input_tokens_seen": 30795456,
      "step": 35570
    },
    {
      "epoch": 16.772748703441774,
      "grad_norm": 0.019620757550001144,
      "learning_rate": 0.008971936053381924,
      "loss": 0.1991,
      "num_input_tokens_seen": 30800320,
      "step": 35575
    },
    {
      "epoch": 16.775106082036775,
      "grad_norm": 0.007302840705960989,
      "learning_rate": 0.008951880484595953,
      "loss": 0.0479,
      "num_input_tokens_seen": 30804560,
      "step": 35580
    },
    {
      "epoch": 16.77746346063178,
      "grad_norm": 0.005145778879523277,
      "learning_rate": 0.008931846667202552,
      "loss": 0.1362,
      "num_input_tokens_seen": 30809232,
      "step": 35585
    },
    {
      "epoch": 16.77982083922678,
      "grad_norm": 0.02542269416153431,
      "learning_rate": 0.008911834604291152,
      "loss": 0.1742,
      "num_input_tokens_seen": 30813760,
      "step": 35590
    },
    {
      "epoch": 16.782178217821784,
      "grad_norm": 0.0035323547199368477,
      "learning_rate": 0.008891844298947882,
      "loss": 0.1113,
      "num_input_tokens_seen": 30818512,
      "step": 35595
    },
    {
      "epoch": 16.784535596416784,
      "grad_norm": 0.016081709414720535,
      "learning_rate": 0.008871875754255508,
      "loss": 0.1035,
      "num_input_tokens_seen": 30823072,
      "step": 35600
    },
    {
      "epoch": 16.784535596416784,
      "eval_loss": 0.2634243071079254,
      "eval_runtime": 21.9317,
      "eval_samples_per_second": 42.997,
      "eval_steps_per_second": 21.521,
      "num_input_tokens_seen": 30823072,
      "step": 35600
    },
    {
      "epoch": 16.78689297501179,
      "grad_norm": 0.007513964083045721,
      "learning_rate": 0.008851928973293422,
      "loss": 0.155,
      "num_input_tokens_seen": 30827136,
      "step": 35605
    },
    {
      "epoch": 16.78925035360679,
      "grad_norm": 0.008744494058191776,
      "learning_rate": 0.00883200395913764,
      "loss": 0.1164,
      "num_input_tokens_seen": 30831248,
      "step": 35610
    },
    {
      "epoch": 16.79160773220179,
      "grad_norm": 0.020081497728824615,
      "learning_rate": 0.00881210071486091,
      "loss": 0.1239,
      "num_input_tokens_seen": 30837152,
      "step": 35615
    },
    {
      "epoch": 16.793965110796794,
      "grad_norm": 0.014561597257852554,
      "learning_rate": 0.008792219243532505,
      "loss": 0.13,
      "num_input_tokens_seen": 30841760,
      "step": 35620
    },
    {
      "epoch": 16.796322489391795,
      "grad_norm": 0.017971096560359,
      "learning_rate": 0.008772359548218428,
      "loss": 0.1277,
      "num_input_tokens_seen": 30846032,
      "step": 35625
    },
    {
      "epoch": 16.7986798679868,
      "grad_norm": 0.004171531647443771,
      "learning_rate": 0.008752521631981274,
      "loss": 0.1071,
      "num_input_tokens_seen": 30850496,
      "step": 35630
    },
    {
      "epoch": 16.8010372465818,
      "grad_norm": 0.011229939758777618,
      "learning_rate": 0.008732705497880315,
      "loss": 0.0896,
      "num_input_tokens_seen": 30854496,
      "step": 35635
    },
    {
      "epoch": 16.803394625176804,
      "grad_norm": 0.010215594433248043,
      "learning_rate": 0.008712911148971459,
      "loss": 0.121,
      "num_input_tokens_seen": 30858320,
      "step": 35640
    },
    {
      "epoch": 16.805752003771804,
      "grad_norm": 0.003503407584503293,
      "learning_rate": 0.008693138588307208,
      "loss": 0.0724,
      "num_input_tokens_seen": 30862992,
      "step": 35645
    },
    {
      "epoch": 16.80810938236681,
      "grad_norm": 0.004218435380607843,
      "learning_rate": 0.008673387818936762,
      "loss": 0.0946,
      "num_input_tokens_seen": 30867136,
      "step": 35650
    },
    {
      "epoch": 16.81046676096181,
      "grad_norm": 0.0130458427593112,
      "learning_rate": 0.008653658843905948,
      "loss": 0.1564,
      "num_input_tokens_seen": 30871216,
      "step": 35655
    },
    {
      "epoch": 16.812824139556813,
      "grad_norm": 0.023598428815603256,
      "learning_rate": 0.0086339516662572,
      "loss": 0.1514,
      "num_input_tokens_seen": 30875856,
      "step": 35660
    },
    {
      "epoch": 16.815181518151814,
      "grad_norm": 0.012036167085170746,
      "learning_rate": 0.008614266289029638,
      "loss": 0.0686,
      "num_input_tokens_seen": 30880016,
      "step": 35665
    },
    {
      "epoch": 16.817538896746818,
      "grad_norm": 0.006943972781300545,
      "learning_rate": 0.008594602715258965,
      "loss": 0.0945,
      "num_input_tokens_seen": 30884048,
      "step": 35670
    },
    {
      "epoch": 16.81989627534182,
      "grad_norm": 0.022596141323447227,
      "learning_rate": 0.008574960947977573,
      "loss": 0.2305,
      "num_input_tokens_seen": 30888448,
      "step": 35675
    },
    {
      "epoch": 16.822253653936823,
      "grad_norm": 0.0193775687366724,
      "learning_rate": 0.008555340990214438,
      "loss": 0.0987,
      "num_input_tokens_seen": 30892176,
      "step": 35680
    },
    {
      "epoch": 16.824611032531823,
      "grad_norm": 0.013672865927219391,
      "learning_rate": 0.008535742844995258,
      "loss": 0.11,
      "num_input_tokens_seen": 30896640,
      "step": 35685
    },
    {
      "epoch": 16.826968411126828,
      "grad_norm": 0.015574382618069649,
      "learning_rate": 0.008516166515342266,
      "loss": 0.084,
      "num_input_tokens_seen": 30900720,
      "step": 35690
    },
    {
      "epoch": 16.82932578972183,
      "grad_norm": 0.02158981002867222,
      "learning_rate": 0.008496612004274411,
      "loss": 0.2111,
      "num_input_tokens_seen": 30905616,
      "step": 35695
    },
    {
      "epoch": 16.831683168316832,
      "grad_norm": 0.01221825648099184,
      "learning_rate": 0.008477079314807201,
      "loss": 0.1449,
      "num_input_tokens_seen": 30909600,
      "step": 35700
    },
    {
      "epoch": 16.834040546911833,
      "grad_norm": 0.012631695717573166,
      "learning_rate": 0.008457568449952874,
      "loss": 0.2199,
      "num_input_tokens_seen": 30913568,
      "step": 35705
    },
    {
      "epoch": 16.836397925506837,
      "grad_norm": 0.015725458040833473,
      "learning_rate": 0.008438079412720189,
      "loss": 0.153,
      "num_input_tokens_seen": 30918080,
      "step": 35710
    },
    {
      "epoch": 16.838755304101838,
      "grad_norm": 0.0046507553197443485,
      "learning_rate": 0.00841861220611466,
      "loss": 0.0709,
      "num_input_tokens_seen": 30923120,
      "step": 35715
    },
    {
      "epoch": 16.841112682696842,
      "grad_norm": 0.01059041265398264,
      "learning_rate": 0.008399166833138355,
      "loss": 0.1082,
      "num_input_tokens_seen": 30927104,
      "step": 35720
    },
    {
      "epoch": 16.843470061291843,
      "grad_norm": 0.018893718719482422,
      "learning_rate": 0.008379743296789987,
      "loss": 0.0789,
      "num_input_tokens_seen": 30931920,
      "step": 35725
    },
    {
      "epoch": 16.845827439886847,
      "grad_norm": 0.010083929635584354,
      "learning_rate": 0.008360341600064896,
      "loss": 0.0878,
      "num_input_tokens_seen": 30936608,
      "step": 35730
    },
    {
      "epoch": 16.848184818481847,
      "grad_norm": 0.011624534614384174,
      "learning_rate": 0.008340961745955121,
      "loss": 0.1578,
      "num_input_tokens_seen": 30941024,
      "step": 35735
    },
    {
      "epoch": 16.85054219707685,
      "grad_norm": 0.007596930954605341,
      "learning_rate": 0.008321603737449224,
      "loss": 0.0673,
      "num_input_tokens_seen": 30945200,
      "step": 35740
    },
    {
      "epoch": 16.852899575671852,
      "grad_norm": 0.012930074706673622,
      "learning_rate": 0.008302267577532479,
      "loss": 0.1238,
      "num_input_tokens_seen": 30950000,
      "step": 35745
    },
    {
      "epoch": 16.855256954266856,
      "grad_norm": 0.005511323921382427,
      "learning_rate": 0.008282953269186771,
      "loss": 0.0535,
      "num_input_tokens_seen": 30953760,
      "step": 35750
    },
    {
      "epoch": 16.857614332861857,
      "grad_norm": 0.024114711210131645,
      "learning_rate": 0.008263660815390567,
      "loss": 0.1489,
      "num_input_tokens_seen": 30957376,
      "step": 35755
    },
    {
      "epoch": 16.85997171145686,
      "grad_norm": 0.006128882523626089,
      "learning_rate": 0.008244390219119069,
      "loss": 0.1692,
      "num_input_tokens_seen": 30961712,
      "step": 35760
    },
    {
      "epoch": 16.862329090051862,
      "grad_norm": 0.023384280502796173,
      "learning_rate": 0.008225141483343967,
      "loss": 0.167,
      "num_input_tokens_seen": 30966480,
      "step": 35765
    },
    {
      "epoch": 16.864686468646866,
      "grad_norm": 0.01856972463428974,
      "learning_rate": 0.00820591461103372,
      "loss": 0.1353,
      "num_input_tokens_seen": 30971152,
      "step": 35770
    },
    {
      "epoch": 16.867043847241867,
      "grad_norm": 0.009614518843591213,
      "learning_rate": 0.008186709605153358,
      "loss": 0.1783,
      "num_input_tokens_seen": 30975088,
      "step": 35775
    },
    {
      "epoch": 16.86940122583687,
      "grad_norm": 0.026372816413640976,
      "learning_rate": 0.008167526468664492,
      "loss": 0.1407,
      "num_input_tokens_seen": 30979264,
      "step": 35780
    },
    {
      "epoch": 16.87175860443187,
      "grad_norm": 0.007105548400431871,
      "learning_rate": 0.008148365204525443,
      "loss": 0.1258,
      "num_input_tokens_seen": 30982960,
      "step": 35785
    },
    {
      "epoch": 16.874115983026876,
      "grad_norm": 0.013276240788400173,
      "learning_rate": 0.00812922581569106,
      "loss": 0.1649,
      "num_input_tokens_seen": 30987056,
      "step": 35790
    },
    {
      "epoch": 16.876473361621876,
      "grad_norm": 0.009276241064071655,
      "learning_rate": 0.008110108305112934,
      "loss": 0.1078,
      "num_input_tokens_seen": 30991456,
      "step": 35795
    },
    {
      "epoch": 16.87883074021688,
      "grad_norm": 0.011670474894344807,
      "learning_rate": 0.008091012675739223,
      "loss": 0.1649,
      "num_input_tokens_seen": 30996032,
      "step": 35800
    },
    {
      "epoch": 16.87883074021688,
      "eval_loss": 0.26129886507987976,
      "eval_runtime": 22.0031,
      "eval_samples_per_second": 42.858,
      "eval_steps_per_second": 21.451,
      "num_input_tokens_seen": 30996032,
      "step": 35800
    },
    {
      "epoch": 16.88118811881188,
      "grad_norm": 0.01039219368249178,
      "learning_rate": 0.008071938930514671,
      "loss": 0.0599,
      "num_input_tokens_seen": 30999968,
      "step": 35805
    },
    {
      "epoch": 16.883545497406885,
      "grad_norm": 0.003808434819802642,
      "learning_rate": 0.008052887072380726,
      "loss": 0.1325,
      "num_input_tokens_seen": 31003792,
      "step": 35810
    },
    {
      "epoch": 16.885902876001886,
      "grad_norm": 0.013571172021329403,
      "learning_rate": 0.008033857104275437,
      "loss": 0.1751,
      "num_input_tokens_seen": 31008944,
      "step": 35815
    },
    {
      "epoch": 16.888260254596887,
      "grad_norm": 0.008013325743377209,
      "learning_rate": 0.008014849029133424,
      "loss": 0.0871,
      "num_input_tokens_seen": 31012976,
      "step": 35820
    },
    {
      "epoch": 16.89061763319189,
      "grad_norm": 0.008567462675273418,
      "learning_rate": 0.007995862849885975,
      "loss": 0.1596,
      "num_input_tokens_seen": 31017664,
      "step": 35825
    },
    {
      "epoch": 16.89297501178689,
      "grad_norm": 0.008565583266317844,
      "learning_rate": 0.007976898569461032,
      "loss": 0.1298,
      "num_input_tokens_seen": 31022704,
      "step": 35830
    },
    {
      "epoch": 16.895332390381895,
      "grad_norm": 0.009018138982355595,
      "learning_rate": 0.007957956190783088,
      "loss": 0.0979,
      "num_input_tokens_seen": 31026752,
      "step": 35835
    },
    {
      "epoch": 16.897689768976896,
      "grad_norm": 0.009172054007649422,
      "learning_rate": 0.007939035716773324,
      "loss": 0.1101,
      "num_input_tokens_seen": 31031136,
      "step": 35840
    },
    {
      "epoch": 16.9000471475719,
      "grad_norm": 0.013981117866933346,
      "learning_rate": 0.007920137150349487,
      "loss": 0.1484,
      "num_input_tokens_seen": 31036208,
      "step": 35845
    },
    {
      "epoch": 16.9024045261669,
      "grad_norm": 0.014014261774718761,
      "learning_rate": 0.007901260494425981,
      "loss": 0.0813,
      "num_input_tokens_seen": 31040816,
      "step": 35850
    },
    {
      "epoch": 16.904761904761905,
      "grad_norm": 0.014437993057072163,
      "learning_rate": 0.007882405751913861,
      "loss": 0.118,
      "num_input_tokens_seen": 31044928,
      "step": 35855
    },
    {
      "epoch": 16.907119283356906,
      "grad_norm": 0.012140852399170399,
      "learning_rate": 0.007863572925720702,
      "loss": 0.11,
      "num_input_tokens_seen": 31048736,
      "step": 35860
    },
    {
      "epoch": 16.90947666195191,
      "grad_norm": 0.005972868762910366,
      "learning_rate": 0.007844762018750827,
      "loss": 0.0821,
      "num_input_tokens_seen": 31053136,
      "step": 35865
    },
    {
      "epoch": 16.91183404054691,
      "grad_norm": 0.00921252928674221,
      "learning_rate": 0.007825973033905054,
      "loss": 0.1153,
      "num_input_tokens_seen": 31057216,
      "step": 35870
    },
    {
      "epoch": 16.914191419141915,
      "grad_norm": 0.007058992981910706,
      "learning_rate": 0.007807205974080927,
      "loss": 0.0655,
      "num_input_tokens_seen": 31061616,
      "step": 35875
    },
    {
      "epoch": 16.916548797736915,
      "grad_norm": 0.003664696356281638,
      "learning_rate": 0.007788460842172551,
      "loss": 0.0817,
      "num_input_tokens_seen": 31065200,
      "step": 35880
    },
    {
      "epoch": 16.91890617633192,
      "grad_norm": 0.006061555817723274,
      "learning_rate": 0.0077697376410706285,
      "loss": 0.089,
      "num_input_tokens_seen": 31069584,
      "step": 35885
    },
    {
      "epoch": 16.92126355492692,
      "grad_norm": 0.02002040669322014,
      "learning_rate": 0.007751036373662567,
      "loss": 0.1716,
      "num_input_tokens_seen": 31074464,
      "step": 35890
    },
    {
      "epoch": 16.923620933521924,
      "grad_norm": 0.01151377335190773,
      "learning_rate": 0.00773235704283231,
      "loss": 0.1197,
      "num_input_tokens_seen": 31079008,
      "step": 35895
    },
    {
      "epoch": 16.925978312116925,
      "grad_norm": 0.0063186693005263805,
      "learning_rate": 0.007713699651460437,
      "loss": 0.1289,
      "num_input_tokens_seen": 31083264,
      "step": 35900
    },
    {
      "epoch": 16.92833569071193,
      "grad_norm": 0.011716816574335098,
      "learning_rate": 0.007695064202424162,
      "loss": 0.0883,
      "num_input_tokens_seen": 31087792,
      "step": 35905
    },
    {
      "epoch": 16.93069306930693,
      "grad_norm": 0.014431536197662354,
      "learning_rate": 0.007676450698597286,
      "loss": 0.089,
      "num_input_tokens_seen": 31091840,
      "step": 35910
    },
    {
      "epoch": 16.933050447901934,
      "grad_norm": 0.013445715419948101,
      "learning_rate": 0.007657859142850265,
      "loss": 0.1364,
      "num_input_tokens_seen": 31095872,
      "step": 35915
    },
    {
      "epoch": 16.935407826496935,
      "grad_norm": 0.009531930088996887,
      "learning_rate": 0.0076392895380501535,
      "loss": 0.1752,
      "num_input_tokens_seen": 31100464,
      "step": 35920
    },
    {
      "epoch": 16.93776520509194,
      "grad_norm": 0.007896044291555882,
      "learning_rate": 0.007620741887060611,
      "loss": 0.1096,
      "num_input_tokens_seen": 31104544,
      "step": 35925
    },
    {
      "epoch": 16.94012258368694,
      "grad_norm": 0.001596235204488039,
      "learning_rate": 0.007602216192741901,
      "loss": 0.0922,
      "num_input_tokens_seen": 31109072,
      "step": 35930
    },
    {
      "epoch": 16.942479962281944,
      "grad_norm": 0.016933150589466095,
      "learning_rate": 0.007583712457950969,
      "loss": 0.1624,
      "num_input_tokens_seen": 31113824,
      "step": 35935
    },
    {
      "epoch": 16.944837340876944,
      "grad_norm": 0.018517998978495598,
      "learning_rate": 0.007565230685541269,
      "loss": 0.1173,
      "num_input_tokens_seen": 31117440,
      "step": 35940
    },
    {
      "epoch": 16.94719471947195,
      "grad_norm": 0.02614550292491913,
      "learning_rate": 0.007546770878362968,
      "loss": 0.1174,
      "num_input_tokens_seen": 31121584,
      "step": 35945
    },
    {
      "epoch": 16.94955209806695,
      "grad_norm": 0.0169843640178442,
      "learning_rate": 0.0075283330392627405,
      "loss": 0.0855,
      "num_input_tokens_seen": 31126080,
      "step": 35950
    },
    {
      "epoch": 16.951909476661953,
      "grad_norm": 0.02208230458199978,
      "learning_rate": 0.007509917171083979,
      "loss": 0.1417,
      "num_input_tokens_seen": 31130432,
      "step": 35955
    },
    {
      "epoch": 16.954266855256954,
      "grad_norm": 0.01811923459172249,
      "learning_rate": 0.007491523276666662,
      "loss": 0.1465,
      "num_input_tokens_seen": 31134464,
      "step": 35960
    },
    {
      "epoch": 16.956624233851958,
      "grad_norm": 0.01501999981701374,
      "learning_rate": 0.007473151358847318,
      "loss": 0.0654,
      "num_input_tokens_seen": 31138544,
      "step": 35965
    },
    {
      "epoch": 16.95898161244696,
      "grad_norm": 0.008531957864761353,
      "learning_rate": 0.007454801420459117,
      "loss": 0.0611,
      "num_input_tokens_seen": 31143408,
      "step": 35970
    },
    {
      "epoch": 16.961338991041963,
      "grad_norm": 0.006068440619856119,
      "learning_rate": 0.0074364734643319105,
      "loss": 0.1054,
      "num_input_tokens_seen": 31147120,
      "step": 35975
    },
    {
      "epoch": 16.963696369636963,
      "grad_norm": 0.0061237686313688755,
      "learning_rate": 0.007418167493292022,
      "loss": 0.109,
      "num_input_tokens_seen": 31150800,
      "step": 35980
    },
    {
      "epoch": 16.966053748231968,
      "grad_norm": 0.024726686999201775,
      "learning_rate": 0.0073998835101625245,
      "loss": 0.1689,
      "num_input_tokens_seen": 31155008,
      "step": 35985
    },
    {
      "epoch": 16.968411126826968,
      "grad_norm": 0.02818484976887703,
      "learning_rate": 0.007381621517762998,
      "loss": 0.3027,
      "num_input_tokens_seen": 31158960,
      "step": 35990
    },
    {
      "epoch": 16.970768505421972,
      "grad_norm": 0.005384627263993025,
      "learning_rate": 0.007363381518909689,
      "loss": 0.1312,
      "num_input_tokens_seen": 31163936,
      "step": 35995
    },
    {
      "epoch": 16.973125884016973,
      "grad_norm": 0.006879071239382029,
      "learning_rate": 0.007345163516415448,
      "loss": 0.0938,
      "num_input_tokens_seen": 31167328,
      "step": 36000
    },
    {
      "epoch": 16.973125884016973,
      "eval_loss": 0.27088505029678345,
      "eval_runtime": 21.9078,
      "eval_samples_per_second": 43.044,
      "eval_steps_per_second": 21.545,
      "num_input_tokens_seen": 31167328,
      "step": 36000
    },
    {
      "epoch": 16.975483262611977,
      "grad_norm": 0.007064170204102993,
      "learning_rate": 0.007326967513089693,
      "loss": 0.0866,
      "num_input_tokens_seen": 31171424,
      "step": 36005
    },
    {
      "epoch": 16.977840641206978,
      "grad_norm": 0.008873811922967434,
      "learning_rate": 0.0073087935117384815,
      "loss": 0.0857,
      "num_input_tokens_seen": 31175696,
      "step": 36010
    },
    {
      "epoch": 16.980198019801982,
      "grad_norm": 0.017500873655080795,
      "learning_rate": 0.007290641515164503,
      "loss": 0.1068,
      "num_input_tokens_seen": 31180064,
      "step": 36015
    },
    {
      "epoch": 16.982555398396983,
      "grad_norm": 0.009193282574415207,
      "learning_rate": 0.007272511526166986,
      "loss": 0.1669,
      "num_input_tokens_seen": 31184976,
      "step": 36020
    },
    {
      "epoch": 16.984912776991983,
      "grad_norm": 0.016908008605241776,
      "learning_rate": 0.0072544035475418265,
      "loss": 0.1368,
      "num_input_tokens_seen": 31189392,
      "step": 36025
    },
    {
      "epoch": 16.987270155586987,
      "grad_norm": 0.018625443801283836,
      "learning_rate": 0.007236317582081475,
      "loss": 0.2236,
      "num_input_tokens_seen": 31193456,
      "step": 36030
    },
    {
      "epoch": 16.989627534181988,
      "grad_norm": 0.004359911661595106,
      "learning_rate": 0.007218253632575066,
      "loss": 0.1539,
      "num_input_tokens_seen": 31198288,
      "step": 36035
    },
    {
      "epoch": 16.991984912776992,
      "grad_norm": 0.01650894433259964,
      "learning_rate": 0.007200211701808223,
      "loss": 0.0892,
      "num_input_tokens_seen": 31202336,
      "step": 36040
    },
    {
      "epoch": 16.994342291371993,
      "grad_norm": 0.001986259827390313,
      "learning_rate": 0.007182191792563286,
      "loss": 0.1659,
      "num_input_tokens_seen": 31206832,
      "step": 36045
    },
    {
      "epoch": 16.996699669966997,
      "grad_norm": 0.003989676013588905,
      "learning_rate": 0.0071641939076191145,
      "loss": 0.157,
      "num_input_tokens_seen": 31210880,
      "step": 36050
    },
    {
      "epoch": 16.999057048561998,
      "grad_norm": 0.0046168118715286255,
      "learning_rate": 0.007146218049751257,
      "loss": 0.154,
      "num_input_tokens_seen": 31215248,
      "step": 36055
    },
    {
      "epoch": 17.001414427157002,
      "grad_norm": 0.006854538805782795,
      "learning_rate": 0.0071282642217317775,
      "loss": 0.0742,
      "num_input_tokens_seen": 31219808,
      "step": 36060
    },
    {
      "epoch": 17.003771805752002,
      "grad_norm": 0.01015361025929451,
      "learning_rate": 0.007110332426329396,
      "loss": 0.1169,
      "num_input_tokens_seen": 31223520,
      "step": 36065
    },
    {
      "epoch": 17.006129184347007,
      "grad_norm": 0.013338311575353146,
      "learning_rate": 0.007092422666309417,
      "loss": 0.1183,
      "num_input_tokens_seen": 31227456,
      "step": 36070
    },
    {
      "epoch": 17.008486562942007,
      "grad_norm": 0.012907693162560463,
      "learning_rate": 0.0070745349444337295,
      "loss": 0.1056,
      "num_input_tokens_seen": 31231344,
      "step": 36075
    },
    {
      "epoch": 17.01084394153701,
      "grad_norm": 0.01949482597410679,
      "learning_rate": 0.007056669263460913,
      "loss": 0.1236,
      "num_input_tokens_seen": 31235552,
      "step": 36080
    },
    {
      "epoch": 17.013201320132012,
      "grad_norm": 0.0028664893470704556,
      "learning_rate": 0.007038825626145995,
      "loss": 0.1003,
      "num_input_tokens_seen": 31240016,
      "step": 36085
    },
    {
      "epoch": 17.015558698727016,
      "grad_norm": 0.01643681339919567,
      "learning_rate": 0.007021004035240724,
      "loss": 0.0792,
      "num_input_tokens_seen": 31244368,
      "step": 36090
    },
    {
      "epoch": 17.017916077322017,
      "grad_norm": 0.004617566708475351,
      "learning_rate": 0.007003204493493453,
      "loss": 0.098,
      "num_input_tokens_seen": 31248432,
      "step": 36095
    },
    {
      "epoch": 17.02027345591702,
      "grad_norm": 0.01235467940568924,
      "learning_rate": 0.006985427003649036,
      "loss": 0.1168,
      "num_input_tokens_seen": 31252800,
      "step": 36100
    },
    {
      "epoch": 17.02263083451202,
      "grad_norm": 0.014176882803440094,
      "learning_rate": 0.006967671568449013,
      "loss": 0.115,
      "num_input_tokens_seen": 31256992,
      "step": 36105
    },
    {
      "epoch": 17.024988213107026,
      "grad_norm": 0.006037629209458828,
      "learning_rate": 0.006949938190631511,
      "loss": 0.0715,
      "num_input_tokens_seen": 31262096,
      "step": 36110
    },
    {
      "epoch": 17.027345591702026,
      "grad_norm": 0.011909185908734798,
      "learning_rate": 0.0069322268729311905,
      "loss": 0.1197,
      "num_input_tokens_seen": 31265872,
      "step": 36115
    },
    {
      "epoch": 17.02970297029703,
      "grad_norm": 0.0054299538023769855,
      "learning_rate": 0.006914537618079403,
      "loss": 0.0802,
      "num_input_tokens_seen": 31270304,
      "step": 36120
    },
    {
      "epoch": 17.03206034889203,
      "grad_norm": 0.00811014510691166,
      "learning_rate": 0.006896870428804031,
      "loss": 0.0864,
      "num_input_tokens_seen": 31274336,
      "step": 36125
    },
    {
      "epoch": 17.034417727487035,
      "grad_norm": 0.011875255033373833,
      "learning_rate": 0.006879225307829595,
      "loss": 0.0917,
      "num_input_tokens_seen": 31278784,
      "step": 36130
    },
    {
      "epoch": 17.036775106082036,
      "grad_norm": 0.015378271229565144,
      "learning_rate": 0.00686160225787717,
      "loss": 0.0952,
      "num_input_tokens_seen": 31282736,
      "step": 36135
    },
    {
      "epoch": 17.03913248467704,
      "grad_norm": 0.017506049945950508,
      "learning_rate": 0.006844001281664463,
      "loss": 0.0727,
      "num_input_tokens_seen": 31288480,
      "step": 36140
    },
    {
      "epoch": 17.04148986327204,
      "grad_norm": 0.01013952773064375,
      "learning_rate": 0.006826422381905789,
      "loss": 0.0668,
      "num_input_tokens_seen": 31292400,
      "step": 36145
    },
    {
      "epoch": 17.043847241867045,
      "grad_norm": 0.01578451693058014,
      "learning_rate": 0.006808865561311994,
      "loss": 0.1547,
      "num_input_tokens_seen": 31296400,
      "step": 36150
    },
    {
      "epoch": 17.046204620462046,
      "grad_norm": 0.005914626177400351,
      "learning_rate": 0.00679133082259058,
      "loss": 0.1096,
      "num_input_tokens_seen": 31300304,
      "step": 36155
    },
    {
      "epoch": 17.04856199905705,
      "grad_norm": 0.01096775010228157,
      "learning_rate": 0.00677381816844565,
      "loss": 0.1028,
      "num_input_tokens_seen": 31305040,
      "step": 36160
    },
    {
      "epoch": 17.05091937765205,
      "grad_norm": 0.014520931988954544,
      "learning_rate": 0.0067563276015778434,
      "loss": 0.0705,
      "num_input_tokens_seen": 31310576,
      "step": 36165
    },
    {
      "epoch": 17.053276756247055,
      "grad_norm": 0.010059935040771961,
      "learning_rate": 0.006738859124684437,
      "loss": 0.0985,
      "num_input_tokens_seen": 31314544,
      "step": 36170
    },
    {
      "epoch": 17.055634134842055,
      "grad_norm": 0.010816458612680435,
      "learning_rate": 0.006721412740459259,
      "loss": 0.0632,
      "num_input_tokens_seen": 31320272,
      "step": 36175
    },
    {
      "epoch": 17.05799151343706,
      "grad_norm": 0.00584813579916954,
      "learning_rate": 0.006703988451592824,
      "loss": 0.0643,
      "num_input_tokens_seen": 31325104,
      "step": 36180
    },
    {
      "epoch": 17.06034889203206,
      "grad_norm": 0.016626687720417976,
      "learning_rate": 0.006686586260772114,
      "loss": 0.0701,
      "num_input_tokens_seen": 31329152,
      "step": 36185
    },
    {
      "epoch": 17.062706270627064,
      "grad_norm": 0.007504918146878481,
      "learning_rate": 0.006669206170680819,
      "loss": 0.0725,
      "num_input_tokens_seen": 31332896,
      "step": 36190
    },
    {
      "epoch": 17.065063649222065,
      "grad_norm": 0.005399907939136028,
      "learning_rate": 0.0066518481839991095,
      "loss": 0.0581,
      "num_input_tokens_seen": 31336656,
      "step": 36195
    },
    {
      "epoch": 17.06742102781707,
      "grad_norm": 0.006314376834779978,
      "learning_rate": 0.006634512303403861,
      "loss": 0.0595,
      "num_input_tokens_seen": 31341392,
      "step": 36200
    },
    {
      "epoch": 17.06742102781707,
      "eval_loss": 0.27376726269721985,
      "eval_runtime": 21.9028,
      "eval_samples_per_second": 43.054,
      "eval_steps_per_second": 21.55,
      "num_input_tokens_seen": 31341392,
      "step": 36200
    },
    {
      "epoch": 17.06977840641207,
      "grad_norm": 0.01618790626525879,
      "learning_rate": 0.0066171985315684355,
      "loss": 0.1527,
      "num_input_tokens_seen": 31346032,
      "step": 36205
    },
    {
      "epoch": 17.072135785007074,
      "grad_norm": 0.010968578048050404,
      "learning_rate": 0.0065999068711628806,
      "loss": 0.1294,
      "num_input_tokens_seen": 31350432,
      "step": 36210
    },
    {
      "epoch": 17.074493163602074,
      "grad_norm": 0.010473438538610935,
      "learning_rate": 0.0065826373248537295,
      "loss": 0.1055,
      "num_input_tokens_seen": 31354528,
      "step": 36215
    },
    {
      "epoch": 17.076850542197075,
      "grad_norm": 0.007763137575238943,
      "learning_rate": 0.006565389895304218,
      "loss": 0.1101,
      "num_input_tokens_seen": 31358544,
      "step": 36220
    },
    {
      "epoch": 17.07920792079208,
      "grad_norm": 0.011175322346389294,
      "learning_rate": 0.006548164585174104,
      "loss": 0.1229,
      "num_input_tokens_seen": 31363040,
      "step": 36225
    },
    {
      "epoch": 17.08156529938708,
      "grad_norm": 0.017585547640919685,
      "learning_rate": 0.006530961397119728,
      "loss": 0.1129,
      "num_input_tokens_seen": 31367280,
      "step": 36230
    },
    {
      "epoch": 17.083922677982084,
      "grad_norm": 0.013315394520759583,
      "learning_rate": 0.00651378033379405,
      "loss": 0.1171,
      "num_input_tokens_seen": 31371264,
      "step": 36235
    },
    {
      "epoch": 17.086280056577085,
      "grad_norm": 0.0194857195019722,
      "learning_rate": 0.006496621397846619,
      "loss": 0.1135,
      "num_input_tokens_seen": 31375488,
      "step": 36240
    },
    {
      "epoch": 17.08863743517209,
      "grad_norm": 0.009498114697635174,
      "learning_rate": 0.006479484591923518,
      "loss": 0.0885,
      "num_input_tokens_seen": 31381280,
      "step": 36245
    },
    {
      "epoch": 17.09099481376709,
      "grad_norm": 0.011644668877124786,
      "learning_rate": 0.006462369918667515,
      "loss": 0.1026,
      "num_input_tokens_seen": 31385120,
      "step": 36250
    },
    {
      "epoch": 17.093352192362094,
      "grad_norm": 0.010010889731347561,
      "learning_rate": 0.006445277380717851,
      "loss": 0.1565,
      "num_input_tokens_seen": 31389888,
      "step": 36255
    },
    {
      "epoch": 17.095709570957094,
      "grad_norm": 0.0033402724657207727,
      "learning_rate": 0.006428206980710466,
      "loss": 0.0593,
      "num_input_tokens_seen": 31393648,
      "step": 36260
    },
    {
      "epoch": 17.0980669495521,
      "grad_norm": 0.009152665734291077,
      "learning_rate": 0.006411158721277788,
      "loss": 0.0546,
      "num_input_tokens_seen": 31397888,
      "step": 36265
    },
    {
      "epoch": 17.1004243281471,
      "grad_norm": 0.013551130890846252,
      "learning_rate": 0.00639413260504888,
      "loss": 0.0776,
      "num_input_tokens_seen": 31402640,
      "step": 36270
    },
    {
      "epoch": 17.102781706742103,
      "grad_norm": 0.025572702288627625,
      "learning_rate": 0.006377128634649376,
      "loss": 0.1097,
      "num_input_tokens_seen": 31406928,
      "step": 36275
    },
    {
      "epoch": 17.105139085337104,
      "grad_norm": 0.010415523312985897,
      "learning_rate": 0.006360146812701528,
      "loss": 0.0728,
      "num_input_tokens_seen": 31411104,
      "step": 36280
    },
    {
      "epoch": 17.107496463932108,
      "grad_norm": 0.002753691514953971,
      "learning_rate": 0.006343187141824125,
      "loss": 0.0604,
      "num_input_tokens_seen": 31415632,
      "step": 36285
    },
    {
      "epoch": 17.10985384252711,
      "grad_norm": 0.002575148129835725,
      "learning_rate": 0.00632624962463259,
      "loss": 0.0488,
      "num_input_tokens_seen": 31420240,
      "step": 36290
    },
    {
      "epoch": 17.112211221122113,
      "grad_norm": 0.02011752687394619,
      "learning_rate": 0.006309334263738853,
      "loss": 0.1385,
      "num_input_tokens_seen": 31424592,
      "step": 36295
    },
    {
      "epoch": 17.114568599717114,
      "grad_norm": 0.004235713742673397,
      "learning_rate": 0.006292441061751508,
      "loss": 0.0985,
      "num_input_tokens_seen": 31429568,
      "step": 36300
    },
    {
      "epoch": 17.116925978312118,
      "grad_norm": 0.004790938459336758,
      "learning_rate": 0.0062755700212757054,
      "loss": 0.037,
      "num_input_tokens_seen": 31433520,
      "step": 36305
    },
    {
      "epoch": 17.11928335690712,
      "grad_norm": 0.005081883165985346,
      "learning_rate": 0.006258721144913148,
      "loss": 0.1048,
      "num_input_tokens_seen": 31439184,
      "step": 36310
    },
    {
      "epoch": 17.121640735502123,
      "grad_norm": 0.014255238696932793,
      "learning_rate": 0.0062418944352621575,
      "loss": 0.1235,
      "num_input_tokens_seen": 31443600,
      "step": 36315
    },
    {
      "epoch": 17.123998114097123,
      "grad_norm": 0.01183360442519188,
      "learning_rate": 0.0062250898949176405,
      "loss": 0.0711,
      "num_input_tokens_seen": 31448288,
      "step": 36320
    },
    {
      "epoch": 17.126355492692127,
      "grad_norm": 0.004065865650773048,
      "learning_rate": 0.006208307526471041,
      "loss": 0.0982,
      "num_input_tokens_seen": 31452864,
      "step": 36325
    },
    {
      "epoch": 17.128712871287128,
      "grad_norm": 0.013872957788407803,
      "learning_rate": 0.006191547332510405,
      "loss": 0.0717,
      "num_input_tokens_seen": 31457056,
      "step": 36330
    },
    {
      "epoch": 17.131070249882132,
      "grad_norm": 0.011521593667566776,
      "learning_rate": 0.006174809315620416,
      "loss": 0.1234,
      "num_input_tokens_seen": 31461584,
      "step": 36335
    },
    {
      "epoch": 17.133427628477133,
      "grad_norm": 0.015593203715980053,
      "learning_rate": 0.00615809347838221,
      "loss": 0.1182,
      "num_input_tokens_seen": 31465328,
      "step": 36340
    },
    {
      "epoch": 17.135785007072137,
      "grad_norm": 0.003055104287341237,
      "learning_rate": 0.006141399823373655,
      "loss": 0.125,
      "num_input_tokens_seen": 31469744,
      "step": 36345
    },
    {
      "epoch": 17.138142385667138,
      "grad_norm": 0.018332447856664658,
      "learning_rate": 0.0061247283531690455,
      "loss": 0.0922,
      "num_input_tokens_seen": 31473824,
      "step": 36350
    },
    {
      "epoch": 17.14049976426214,
      "grad_norm": 0.01814332976937294,
      "learning_rate": 0.0061080790703393895,
      "loss": 0.0632,
      "num_input_tokens_seen": 31477648,
      "step": 36355
    },
    {
      "epoch": 17.142857142857142,
      "grad_norm": 0.013918413780629635,
      "learning_rate": 0.006091451977452217,
      "loss": 0.1405,
      "num_input_tokens_seen": 31481808,
      "step": 36360
    },
    {
      "epoch": 17.145214521452147,
      "grad_norm": 0.018977725878357887,
      "learning_rate": 0.00607484707707161,
      "loss": 0.1416,
      "num_input_tokens_seen": 31486016,
      "step": 36365
    },
    {
      "epoch": 17.147571900047147,
      "grad_norm": 0.007560989819467068,
      "learning_rate": 0.006058264371758254,
      "loss": 0.1006,
      "num_input_tokens_seen": 31490192,
      "step": 36370
    },
    {
      "epoch": 17.14992927864215,
      "grad_norm": 0.01759473606944084,
      "learning_rate": 0.00604170386406942,
      "loss": 0.0929,
      "num_input_tokens_seen": 31494528,
      "step": 36375
    },
    {
      "epoch": 17.152286657237152,
      "grad_norm": 0.005675950087606907,
      "learning_rate": 0.006025165556558931,
      "loss": 0.072,
      "num_input_tokens_seen": 31498144,
      "step": 36380
    },
    {
      "epoch": 17.154644035832156,
      "grad_norm": 0.00676286406815052,
      "learning_rate": 0.006008649451777248,
      "loss": 0.0876,
      "num_input_tokens_seen": 31502688,
      "step": 36385
    },
    {
      "epoch": 17.157001414427157,
      "grad_norm": 0.012616056017577648,
      "learning_rate": 0.005992155552271283,
      "loss": 0.0496,
      "num_input_tokens_seen": 31507296,
      "step": 36390
    },
    {
      "epoch": 17.15935879302216,
      "grad_norm": 0.012391744181513786,
      "learning_rate": 0.005975683860584685,
      "loss": 0.1831,
      "num_input_tokens_seen": 31511680,
      "step": 36395
    },
    {
      "epoch": 17.16171617161716,
      "grad_norm": 0.011679720133543015,
      "learning_rate": 0.0059592343792575385,
      "loss": 0.1804,
      "num_input_tokens_seen": 31515648,
      "step": 36400
    },
    {
      "epoch": 17.16171617161716,
      "eval_loss": 0.2758013606071472,
      "eval_runtime": 21.9094,
      "eval_samples_per_second": 43.041,
      "eval_steps_per_second": 21.543,
      "num_input_tokens_seen": 31515648,
      "step": 36400
    },
    {
      "epoch": 17.164073550212166,
      "grad_norm": 0.01688370667397976,
      "learning_rate": 0.0059428071108265975,
      "loss": 0.1466,
      "num_input_tokens_seen": 31520352,
      "step": 36405
    },
    {
      "epoch": 17.166430928807166,
      "grad_norm": 0.021355103701353073,
      "learning_rate": 0.005926402057825136,
      "loss": 0.066,
      "num_input_tokens_seen": 31524528,
      "step": 36410
    },
    {
      "epoch": 17.16878830740217,
      "grad_norm": 0.00760915782302618,
      "learning_rate": 0.005910019222782997,
      "loss": 0.1206,
      "num_input_tokens_seen": 31529040,
      "step": 36415
    },
    {
      "epoch": 17.17114568599717,
      "grad_norm": 0.005009378306567669,
      "learning_rate": 0.005893658608226643,
      "loss": 0.0877,
      "num_input_tokens_seen": 31533392,
      "step": 36420
    },
    {
      "epoch": 17.173503064592172,
      "grad_norm": 0.0205161664634943,
      "learning_rate": 0.0058773202166791045,
      "loss": 0.1297,
      "num_input_tokens_seen": 31537488,
      "step": 36425
    },
    {
      "epoch": 17.175860443187176,
      "grad_norm": 0.005842353217303753,
      "learning_rate": 0.005861004050659918,
      "loss": 0.0474,
      "num_input_tokens_seen": 31541952,
      "step": 36430
    },
    {
      "epoch": 17.178217821782177,
      "grad_norm": 0.006939308252185583,
      "learning_rate": 0.005844710112685286,
      "loss": 0.1122,
      "num_input_tokens_seen": 31546480,
      "step": 36435
    },
    {
      "epoch": 17.18057520037718,
      "grad_norm": 0.016216833144426346,
      "learning_rate": 0.005828438405267933,
      "loss": 0.0868,
      "num_input_tokens_seen": 31551216,
      "step": 36440
    },
    {
      "epoch": 17.18293257897218,
      "grad_norm": 0.004711097106337547,
      "learning_rate": 0.00581218893091715,
      "loss": 0.0915,
      "num_input_tokens_seen": 31555744,
      "step": 36445
    },
    {
      "epoch": 17.185289957567186,
      "grad_norm": 0.006320019252598286,
      "learning_rate": 0.005795961692138801,
      "loss": 0.0638,
      "num_input_tokens_seen": 31559920,
      "step": 36450
    },
    {
      "epoch": 17.187647336162186,
      "grad_norm": 0.018396077677607536,
      "learning_rate": 0.00577975669143535,
      "loss": 0.0774,
      "num_input_tokens_seen": 31563328,
      "step": 36455
    },
    {
      "epoch": 17.19000471475719,
      "grad_norm": 0.007439781446009874,
      "learning_rate": 0.005763573931305782,
      "loss": 0.0857,
      "num_input_tokens_seen": 31567808,
      "step": 36460
    },
    {
      "epoch": 17.19236209335219,
      "grad_norm": 0.0021173127461224794,
      "learning_rate": 0.005747413414245733,
      "loss": 0.0778,
      "num_input_tokens_seen": 31572032,
      "step": 36465
    },
    {
      "epoch": 17.194719471947195,
      "grad_norm": 0.01380198821425438,
      "learning_rate": 0.005731275142747294,
      "loss": 0.085,
      "num_input_tokens_seen": 31576048,
      "step": 36470
    },
    {
      "epoch": 17.197076850542196,
      "grad_norm": 0.017338458448648453,
      "learning_rate": 0.005715159119299256,
      "loss": 0.116,
      "num_input_tokens_seen": 31580208,
      "step": 36475
    },
    {
      "epoch": 17.1994342291372,
      "grad_norm": 0.014047152362763882,
      "learning_rate": 0.005699065346386867,
      "loss": 0.1103,
      "num_input_tokens_seen": 31584976,
      "step": 36480
    },
    {
      "epoch": 17.2017916077322,
      "grad_norm": 0.009470172226428986,
      "learning_rate": 0.0056829938264919885,
      "loss": 0.091,
      "num_input_tokens_seen": 31588768,
      "step": 36485
    },
    {
      "epoch": 17.204148986327205,
      "grad_norm": 0.006561388727277517,
      "learning_rate": 0.005666944562093074,
      "loss": 0.0972,
      "num_input_tokens_seen": 31592832,
      "step": 36490
    },
    {
      "epoch": 17.206506364922205,
      "grad_norm": 0.001638320041820407,
      "learning_rate": 0.005650917555665108,
      "loss": 0.1218,
      "num_input_tokens_seen": 31596736,
      "step": 36495
    },
    {
      "epoch": 17.20886374351721,
      "grad_norm": 0.02003917470574379,
      "learning_rate": 0.005634912809679632,
      "loss": 0.1084,
      "num_input_tokens_seen": 31602352,
      "step": 36500
    },
    {
      "epoch": 17.21122112211221,
      "grad_norm": 0.018580028787255287,
      "learning_rate": 0.005618930326604854,
      "loss": 0.1284,
      "num_input_tokens_seen": 31606976,
      "step": 36505
    },
    {
      "epoch": 17.213578500707214,
      "grad_norm": 0.018456248566508293,
      "learning_rate": 0.005602970108905386,
      "loss": 0.1167,
      "num_input_tokens_seen": 31610592,
      "step": 36510
    },
    {
      "epoch": 17.215935879302215,
      "grad_norm": 0.012837285175919533,
      "learning_rate": 0.005587032159042543,
      "loss": 0.0669,
      "num_input_tokens_seen": 31614768,
      "step": 36515
    },
    {
      "epoch": 17.21829325789722,
      "grad_norm": 0.006123372353613377,
      "learning_rate": 0.005571116479474158,
      "loss": 0.0915,
      "num_input_tokens_seen": 31618688,
      "step": 36520
    },
    {
      "epoch": 17.22065063649222,
      "grad_norm": 0.016067981719970703,
      "learning_rate": 0.005555223072654619,
      "loss": 0.081,
      "num_input_tokens_seen": 31623152,
      "step": 36525
    },
    {
      "epoch": 17.223008015087224,
      "grad_norm": 0.0017877136124297976,
      "learning_rate": 0.005539351941034881,
      "loss": 0.1142,
      "num_input_tokens_seen": 31627040,
      "step": 36530
    },
    {
      "epoch": 17.225365393682225,
      "grad_norm": 0.01693052425980568,
      "learning_rate": 0.0055235030870624865,
      "loss": 0.1117,
      "num_input_tokens_seen": 31632656,
      "step": 36535
    },
    {
      "epoch": 17.22772277227723,
      "grad_norm": 0.01036886591464281,
      "learning_rate": 0.005507676513181514,
      "loss": 0.1088,
      "num_input_tokens_seen": 31636832,
      "step": 36540
    },
    {
      "epoch": 17.23008015087223,
      "grad_norm": 0.005394837353378534,
      "learning_rate": 0.005491872221832628,
      "loss": 0.0537,
      "num_input_tokens_seen": 31640992,
      "step": 36545
    },
    {
      "epoch": 17.232437529467234,
      "grad_norm": 0.0036144896876066923,
      "learning_rate": 0.005476090215453061,
      "loss": 0.0577,
      "num_input_tokens_seen": 31644992,
      "step": 36550
    },
    {
      "epoch": 17.234794908062234,
      "grad_norm": 0.011230967938899994,
      "learning_rate": 0.0054603304964765675,
      "loss": 0.1145,
      "num_input_tokens_seen": 31649024,
      "step": 36555
    },
    {
      "epoch": 17.23715228665724,
      "grad_norm": 0.013110796920955181,
      "learning_rate": 0.005444593067333519,
      "loss": 0.0993,
      "num_input_tokens_seen": 31652752,
      "step": 36560
    },
    {
      "epoch": 17.23950966525224,
      "grad_norm": 0.0106892054900527,
      "learning_rate": 0.00542887793045081,
      "loss": 0.0765,
      "num_input_tokens_seen": 31657472,
      "step": 36565
    },
    {
      "epoch": 17.241867043847243,
      "grad_norm": 0.007662664633244276,
      "learning_rate": 0.005413185088251932,
      "loss": 0.0818,
      "num_input_tokens_seen": 31661632,
      "step": 36570
    },
    {
      "epoch": 17.244224422442244,
      "grad_norm": 0.004474200773984194,
      "learning_rate": 0.005397514543156884,
      "loss": 0.1605,
      "num_input_tokens_seen": 31666256,
      "step": 36575
    },
    {
      "epoch": 17.246581801037248,
      "grad_norm": 0.008614257909357548,
      "learning_rate": 0.0053818662975822825,
      "loss": 0.0905,
      "num_input_tokens_seen": 31670368,
      "step": 36580
    },
    {
      "epoch": 17.24893917963225,
      "grad_norm": 0.01151672936975956,
      "learning_rate": 0.005366240353941315,
      "loss": 0.0907,
      "num_input_tokens_seen": 31675328,
      "step": 36585
    },
    {
      "epoch": 17.251296558227253,
      "grad_norm": 0.015184842981398106,
      "learning_rate": 0.005350636714643636,
      "loss": 0.113,
      "num_input_tokens_seen": 31680224,
      "step": 36590
    },
    {
      "epoch": 17.253653936822253,
      "grad_norm": 0.007778584025800228,
      "learning_rate": 0.005335055382095555,
      "loss": 0.1056,
      "num_input_tokens_seen": 31684864,
      "step": 36595
    },
    {
      "epoch": 17.256011315417258,
      "grad_norm": 0.004454773850739002,
      "learning_rate": 0.005319496358699915,
      "loss": 0.0988,
      "num_input_tokens_seen": 31690208,
      "step": 36600
    },
    {
      "epoch": 17.256011315417258,
      "eval_loss": 0.28145697712898254,
      "eval_runtime": 21.9155,
      "eval_samples_per_second": 43.029,
      "eval_steps_per_second": 21.537,
      "num_input_tokens_seen": 31690208,
      "step": 36600
    },
    {
      "epoch": 17.25836869401226,
      "grad_norm": 0.007907221093773842,
      "learning_rate": 0.005303959646856099,
      "loss": 0.0831,
      "num_input_tokens_seen": 31694384,
      "step": 36605
    },
    {
      "epoch": 17.260726072607262,
      "grad_norm": 0.0030305185355246067,
      "learning_rate": 0.005288445248960089,
      "loss": 0.0916,
      "num_input_tokens_seen": 31698480,
      "step": 36610
    },
    {
      "epoch": 17.263083451202263,
      "grad_norm": 0.009816888719797134,
      "learning_rate": 0.005272953167404354,
      "loss": 0.1345,
      "num_input_tokens_seen": 31702576,
      "step": 36615
    },
    {
      "epoch": 17.265440829797264,
      "grad_norm": 0.00972907803952694,
      "learning_rate": 0.005257483404578017,
      "loss": 0.0894,
      "num_input_tokens_seen": 31707184,
      "step": 36620
    },
    {
      "epoch": 17.267798208392268,
      "grad_norm": 0.017799345776438713,
      "learning_rate": 0.0052420359628666865,
      "loss": 0.0963,
      "num_input_tokens_seen": 31711552,
      "step": 36625
    },
    {
      "epoch": 17.27015558698727,
      "grad_norm": 0.018719591200351715,
      "learning_rate": 0.00522661084465254,
      "loss": 0.0842,
      "num_input_tokens_seen": 31716048,
      "step": 36630
    },
    {
      "epoch": 17.272512965582273,
      "grad_norm": 0.005683872848749161,
      "learning_rate": 0.005211208052314326,
      "loss": 0.1484,
      "num_input_tokens_seen": 31719872,
      "step": 36635
    },
    {
      "epoch": 17.274870344177273,
      "grad_norm": 0.007124336436390877,
      "learning_rate": 0.005195827588227391,
      "loss": 0.0869,
      "num_input_tokens_seen": 31724800,
      "step": 36640
    },
    {
      "epoch": 17.277227722772277,
      "grad_norm": 0.011073101311922073,
      "learning_rate": 0.0051804694547635255,
      "loss": 0.0945,
      "num_input_tokens_seen": 31729344,
      "step": 36645
    },
    {
      "epoch": 17.279585101367278,
      "grad_norm": 0.013655728660523891,
      "learning_rate": 0.005165133654291232,
      "loss": 0.0653,
      "num_input_tokens_seen": 31734240,
      "step": 36650
    },
    {
      "epoch": 17.281942479962282,
      "grad_norm": 0.013372094370424747,
      "learning_rate": 0.005149820189175402,
      "loss": 0.0794,
      "num_input_tokens_seen": 31738464,
      "step": 36655
    },
    {
      "epoch": 17.284299858557283,
      "grad_norm": 0.002947862260043621,
      "learning_rate": 0.005134529061777598,
      "loss": 0.062,
      "num_input_tokens_seen": 31742816,
      "step": 36660
    },
    {
      "epoch": 17.286657237152287,
      "grad_norm": 0.00851537100970745,
      "learning_rate": 0.005119260274455933,
      "loss": 0.0936,
      "num_input_tokens_seen": 31746960,
      "step": 36665
    },
    {
      "epoch": 17.289014615747288,
      "grad_norm": 0.003180151106789708,
      "learning_rate": 0.005104013829565007,
      "loss": 0.1269,
      "num_input_tokens_seen": 31752064,
      "step": 36670
    },
    {
      "epoch": 17.291371994342292,
      "grad_norm": 0.03190287947654724,
      "learning_rate": 0.005088789729456006,
      "loss": 0.2212,
      "num_input_tokens_seen": 31756368,
      "step": 36675
    },
    {
      "epoch": 17.293729372937293,
      "grad_norm": 0.00527909304946661,
      "learning_rate": 0.005073587976476735,
      "loss": 0.0618,
      "num_input_tokens_seen": 31761120,
      "step": 36680
    },
    {
      "epoch": 17.296086751532297,
      "grad_norm": 0.001020805211737752,
      "learning_rate": 0.005058408572971418,
      "loss": 0.0795,
      "num_input_tokens_seen": 31766048,
      "step": 36685
    },
    {
      "epoch": 17.298444130127297,
      "grad_norm": 0.009127741679549217,
      "learning_rate": 0.005043251521280983,
      "loss": 0.1165,
      "num_input_tokens_seen": 31770368,
      "step": 36690
    },
    {
      "epoch": 17.3008015087223,
      "grad_norm": 0.02393074706196785,
      "learning_rate": 0.005028116823742795,
      "loss": 0.1032,
      "num_input_tokens_seen": 31774096,
      "step": 36695
    },
    {
      "epoch": 17.303158887317302,
      "grad_norm": 0.025998570024967194,
      "learning_rate": 0.005013004482690819,
      "loss": 0.1377,
      "num_input_tokens_seen": 31778576,
      "step": 36700
    },
    {
      "epoch": 17.305516265912306,
      "grad_norm": 0.012036054395139217,
      "learning_rate": 0.0049979145004555746,
      "loss": 0.0424,
      "num_input_tokens_seen": 31783184,
      "step": 36705
    },
    {
      "epoch": 17.307873644507307,
      "grad_norm": 0.020190933719277382,
      "learning_rate": 0.004982846879364116,
      "loss": 0.1003,
      "num_input_tokens_seen": 31787840,
      "step": 36710
    },
    {
      "epoch": 17.31023102310231,
      "grad_norm": 0.010709051042795181,
      "learning_rate": 0.0049678016217400535,
      "loss": 0.0799,
      "num_input_tokens_seen": 31792320,
      "step": 36715
    },
    {
      "epoch": 17.31258840169731,
      "grad_norm": 0.015421408228576183,
      "learning_rate": 0.004952778729903595,
      "loss": 0.1102,
      "num_input_tokens_seen": 31796272,
      "step": 36720
    },
    {
      "epoch": 17.314945780292316,
      "grad_norm": 0.00610466580837965,
      "learning_rate": 0.004937778206171422,
      "loss": 0.0916,
      "num_input_tokens_seen": 31800640,
      "step": 36725
    },
    {
      "epoch": 17.317303158887317,
      "grad_norm": 0.00515762809664011,
      "learning_rate": 0.004922800052856835,
      "loss": 0.075,
      "num_input_tokens_seen": 31805872,
      "step": 36730
    },
    {
      "epoch": 17.31966053748232,
      "grad_norm": 0.010027341544628143,
      "learning_rate": 0.004907844272269602,
      "loss": 0.1794,
      "num_input_tokens_seen": 31810608,
      "step": 36735
    },
    {
      "epoch": 17.32201791607732,
      "grad_norm": 0.012927073054015636,
      "learning_rate": 0.004892910866716144,
      "loss": 0.0927,
      "num_input_tokens_seen": 31814688,
      "step": 36740
    },
    {
      "epoch": 17.324375294672326,
      "grad_norm": 0.02078653685748577,
      "learning_rate": 0.004877999838499369,
      "loss": 0.0788,
      "num_input_tokens_seen": 31819520,
      "step": 36745
    },
    {
      "epoch": 17.326732673267326,
      "grad_norm": 0.0125935859978199,
      "learning_rate": 0.0048631111899187065,
      "loss": 0.1159,
      "num_input_tokens_seen": 31824912,
      "step": 36750
    },
    {
      "epoch": 17.32909005186233,
      "grad_norm": 0.021209241822361946,
      "learning_rate": 0.0048482449232702335,
      "loss": 0.1389,
      "num_input_tokens_seen": 31829104,
      "step": 36755
    },
    {
      "epoch": 17.33144743045733,
      "grad_norm": 0.007866098545491695,
      "learning_rate": 0.004833401040846469,
      "loss": 0.0447,
      "num_input_tokens_seen": 31833824,
      "step": 36760
    },
    {
      "epoch": 17.333804809052335,
      "grad_norm": 0.029457172378897667,
      "learning_rate": 0.004818579544936546,
      "loss": 0.1531,
      "num_input_tokens_seen": 31837712,
      "step": 36765
    },
    {
      "epoch": 17.336162187647336,
      "grad_norm": 0.004547744523733854,
      "learning_rate": 0.004803780437826121,
      "loss": 0.077,
      "num_input_tokens_seen": 31841824,
      "step": 36770
    },
    {
      "epoch": 17.33851956624234,
      "grad_norm": 0.002559335669502616,
      "learning_rate": 0.004789003721797402,
      "loss": 0.1112,
      "num_input_tokens_seen": 31845952,
      "step": 36775
    },
    {
      "epoch": 17.34087694483734,
      "grad_norm": 0.02416328713297844,
      "learning_rate": 0.004774249399129132,
      "loss": 0.1575,
      "num_input_tokens_seen": 31850112,
      "step": 36780
    },
    {
      "epoch": 17.343234323432345,
      "grad_norm": 0.017543796449899673,
      "learning_rate": 0.004759517472096642,
      "loss": 0.1209,
      "num_input_tokens_seen": 31853936,
      "step": 36785
    },
    {
      "epoch": 17.345591702027345,
      "grad_norm": 0.013840384781360626,
      "learning_rate": 0.004744807942971746,
      "loss": 0.0726,
      "num_input_tokens_seen": 31858480,
      "step": 36790
    },
    {
      "epoch": 17.34794908062235,
      "grad_norm": 0.01507110521197319,
      "learning_rate": 0.004730120814022881,
      "loss": 0.1331,
      "num_input_tokens_seen": 31863376,
      "step": 36795
    },
    {
      "epoch": 17.35030645921735,
      "grad_norm": 0.016000894829630852,
      "learning_rate": 0.004715456087514935,
      "loss": 0.1109,
      "num_input_tokens_seen": 31868288,
      "step": 36800
    },
    {
      "epoch": 17.35030645921735,
      "eval_loss": 0.28111526370048523,
      "eval_runtime": 21.8597,
      "eval_samples_per_second": 43.139,
      "eval_steps_per_second": 21.592,
      "num_input_tokens_seen": 31868288,
      "step": 36800
    },
    {
      "epoch": 17.352663837812354,
      "grad_norm": 0.028043637052178383,
      "learning_rate": 0.004700813765709432,
      "loss": 0.0703,
      "num_input_tokens_seen": 31872656,
      "step": 36805
    },
    {
      "epoch": 17.355021216407355,
      "grad_norm": 0.012592381797730923,
      "learning_rate": 0.004686193850864401,
      "loss": 0.0941,
      "num_input_tokens_seen": 31877552,
      "step": 36810
    },
    {
      "epoch": 17.35737859500236,
      "grad_norm": 0.01725969836115837,
      "learning_rate": 0.004671596345234385,
      "loss": 0.1153,
      "num_input_tokens_seen": 31882400,
      "step": 36815
    },
    {
      "epoch": 17.35973597359736,
      "grad_norm": 0.015227221883833408,
      "learning_rate": 0.00465702125107052,
      "loss": 0.1542,
      "num_input_tokens_seen": 31886560,
      "step": 36820
    },
    {
      "epoch": 17.36209335219236,
      "grad_norm": 0.021076224744319916,
      "learning_rate": 0.004642468570620506,
      "loss": 0.1175,
      "num_input_tokens_seen": 31890512,
      "step": 36825
    },
    {
      "epoch": 17.364450730787365,
      "grad_norm": 0.01934572495520115,
      "learning_rate": 0.004627938306128482,
      "loss": 0.1155,
      "num_input_tokens_seen": 31895040,
      "step": 36830
    },
    {
      "epoch": 17.366808109382365,
      "grad_norm": 0.00551070598885417,
      "learning_rate": 0.004613430459835255,
      "loss": 0.0965,
      "num_input_tokens_seen": 31899312,
      "step": 36835
    },
    {
      "epoch": 17.36916548797737,
      "grad_norm": 0.01007936429232359,
      "learning_rate": 0.004598945033978085,
      "loss": 0.0826,
      "num_input_tokens_seen": 31904112,
      "step": 36840
    },
    {
      "epoch": 17.37152286657237,
      "grad_norm": 0.014137782156467438,
      "learning_rate": 0.004584482030790804,
      "loss": 0.0985,
      "num_input_tokens_seen": 31908048,
      "step": 36845
    },
    {
      "epoch": 17.373880245167374,
      "grad_norm": 0.007444418501108885,
      "learning_rate": 0.004570041452503826,
      "loss": 0.1,
      "num_input_tokens_seen": 31911696,
      "step": 36850
    },
    {
      "epoch": 17.376237623762375,
      "grad_norm": 0.005004009697586298,
      "learning_rate": 0.004555623301344003,
      "loss": 0.0356,
      "num_input_tokens_seen": 31916944,
      "step": 36855
    },
    {
      "epoch": 17.37859500235738,
      "grad_norm": 0.007047620601952076,
      "learning_rate": 0.004541227579534857,
      "loss": 0.1214,
      "num_input_tokens_seen": 31921264,
      "step": 36860
    },
    {
      "epoch": 17.38095238095238,
      "grad_norm": 0.018108729273080826,
      "learning_rate": 0.004526854289296378,
      "loss": 0.088,
      "num_input_tokens_seen": 31925520,
      "step": 36865
    },
    {
      "epoch": 17.383309759547384,
      "grad_norm": 0.01568642072379589,
      "learning_rate": 0.004512503432845078,
      "loss": 0.0701,
      "num_input_tokens_seen": 31929104,
      "step": 36870
    },
    {
      "epoch": 17.385667138142384,
      "grad_norm": 0.010210195556282997,
      "learning_rate": 0.004498175012394068,
      "loss": 0.0627,
      "num_input_tokens_seen": 31933136,
      "step": 36875
    },
    {
      "epoch": 17.38802451673739,
      "grad_norm": 0.019430244341492653,
      "learning_rate": 0.004483869030152965,
      "loss": 0.1142,
      "num_input_tokens_seen": 31936912,
      "step": 36880
    },
    {
      "epoch": 17.39038189533239,
      "grad_norm": 0.018688926473259926,
      "learning_rate": 0.004469585488327904,
      "loss": 0.1067,
      "num_input_tokens_seen": 31941040,
      "step": 36885
    },
    {
      "epoch": 17.392739273927393,
      "grad_norm": 0.00512363575398922,
      "learning_rate": 0.0044553243891216395,
      "loss": 0.0963,
      "num_input_tokens_seen": 31945024,
      "step": 36890
    },
    {
      "epoch": 17.395096652522394,
      "grad_norm": 0.004080599173903465,
      "learning_rate": 0.004441085734733363,
      "loss": 0.0656,
      "num_input_tokens_seen": 31949312,
      "step": 36895
    },
    {
      "epoch": 17.397454031117398,
      "grad_norm": 0.021944599226117134,
      "learning_rate": 0.004426869527358884,
      "loss": 0.1399,
      "num_input_tokens_seen": 31953536,
      "step": 36900
    },
    {
      "epoch": 17.3998114097124,
      "grad_norm": 0.017604153603315353,
      "learning_rate": 0.0044126757691905156,
      "loss": 0.0798,
      "num_input_tokens_seen": 31957680,
      "step": 36905
    },
    {
      "epoch": 17.402168788307403,
      "grad_norm": 0.006458121817559004,
      "learning_rate": 0.004398504462417107,
      "loss": 0.0651,
      "num_input_tokens_seen": 31962800,
      "step": 36910
    },
    {
      "epoch": 17.404526166902404,
      "grad_norm": 0.006085680797696114,
      "learning_rate": 0.0043843556092240605,
      "loss": 0.0947,
      "num_input_tokens_seen": 31967152,
      "step": 36915
    },
    {
      "epoch": 17.406883545497408,
      "grad_norm": 0.012713879346847534,
      "learning_rate": 0.004370229211793281,
      "loss": 0.1065,
      "num_input_tokens_seen": 31971056,
      "step": 36920
    },
    {
      "epoch": 17.40924092409241,
      "grad_norm": 0.01980450563132763,
      "learning_rate": 0.0043561252723032405,
      "loss": 0.0906,
      "num_input_tokens_seen": 31975152,
      "step": 36925
    },
    {
      "epoch": 17.411598302687413,
      "grad_norm": 0.00924790557473898,
      "learning_rate": 0.004342043792929001,
      "loss": 0.044,
      "num_input_tokens_seen": 31979184,
      "step": 36930
    },
    {
      "epoch": 17.413955681282413,
      "grad_norm": 0.017202438786625862,
      "learning_rate": 0.004327984775842025,
      "loss": 0.0663,
      "num_input_tokens_seen": 31982544,
      "step": 36935
    },
    {
      "epoch": 17.416313059877417,
      "grad_norm": 0.00657540000975132,
      "learning_rate": 0.004313948223210428,
      "loss": 0.0665,
      "num_input_tokens_seen": 31987136,
      "step": 36940
    },
    {
      "epoch": 17.418670438472418,
      "grad_norm": 0.011688865721225739,
      "learning_rate": 0.004299934137198846,
      "loss": 0.0376,
      "num_input_tokens_seen": 31992144,
      "step": 36945
    },
    {
      "epoch": 17.421027817067422,
      "grad_norm": 0.012884492054581642,
      "learning_rate": 0.004285942519968383,
      "loss": 0.1052,
      "num_input_tokens_seen": 31996064,
      "step": 36950
    },
    {
      "epoch": 17.423385195662423,
      "grad_norm": 0.007845276035368443,
      "learning_rate": 0.004271973373676746,
      "loss": 0.052,
      "num_input_tokens_seen": 32000448,
      "step": 36955
    },
    {
      "epoch": 17.425742574257427,
      "grad_norm": 0.008708599023520947,
      "learning_rate": 0.004258026700478146,
      "loss": 0.0847,
      "num_input_tokens_seen": 32005968,
      "step": 36960
    },
    {
      "epoch": 17.428099952852428,
      "grad_norm": 0.009054476395249367,
      "learning_rate": 0.004244102502523328,
      "loss": 0.0729,
      "num_input_tokens_seen": 32010976,
      "step": 36965
    },
    {
      "epoch": 17.430457331447432,
      "grad_norm": 0.0228553656488657,
      "learning_rate": 0.004230200781959592,
      "loss": 0.1265,
      "num_input_tokens_seen": 32016032,
      "step": 36970
    },
    {
      "epoch": 17.432814710042432,
      "grad_norm": 0.016372598707675934,
      "learning_rate": 0.004216321540930756,
      "loss": 0.0657,
      "num_input_tokens_seen": 32020304,
      "step": 36975
    },
    {
      "epoch": 17.435172088637437,
      "grad_norm": 0.0333210825920105,
      "learning_rate": 0.004202464781577175,
      "loss": 0.3029,
      "num_input_tokens_seen": 32024144,
      "step": 36980
    },
    {
      "epoch": 17.437529467232437,
      "grad_norm": 0.01952928863465786,
      "learning_rate": 0.00418863050603574,
      "loss": 0.1064,
      "num_input_tokens_seen": 32028864,
      "step": 36985
    },
    {
      "epoch": 17.43988684582744,
      "grad_norm": 0.018115228042006493,
      "learning_rate": 0.004174818716439843,
      "loss": 0.1168,
      "num_input_tokens_seen": 32032688,
      "step": 36990
    },
    {
      "epoch": 17.442244224422442,
      "grad_norm": 0.005769358482211828,
      "learning_rate": 0.004161029414919464,
      "loss": 0.0621,
      "num_input_tokens_seen": 32037008,
      "step": 36995
    },
    {
      "epoch": 17.444601603017446,
      "grad_norm": 0.019282463937997818,
      "learning_rate": 0.004147262603601071,
      "loss": 0.0748,
      "num_input_tokens_seen": 32041536,
      "step": 37000
    },
    {
      "epoch": 17.444601603017446,
      "eval_loss": 0.28019195795059204,
      "eval_runtime": 21.8982,
      "eval_samples_per_second": 43.063,
      "eval_steps_per_second": 21.554,
      "num_input_tokens_seen": 32041536,
      "step": 37000
    },
    {
      "epoch": 17.446958981612447,
      "grad_norm": 0.016871048137545586,
      "learning_rate": 0.004133518284607679,
      "loss": 0.0904,
      "num_input_tokens_seen": 32045568,
      "step": 37005
    },
    {
      "epoch": 17.44931636020745,
      "grad_norm": 0.025413470342755318,
      "learning_rate": 0.004119796460058861,
      "loss": 0.1198,
      "num_input_tokens_seen": 32050144,
      "step": 37010
    },
    {
      "epoch": 17.45167373880245,
      "grad_norm": 0.009033769369125366,
      "learning_rate": 0.00410609713207064,
      "loss": 0.0735,
      "num_input_tokens_seen": 32055296,
      "step": 37015
    },
    {
      "epoch": 17.454031117397456,
      "grad_norm": 0.023872872814536095,
      "learning_rate": 0.004092420302755678,
      "loss": 0.1033,
      "num_input_tokens_seen": 32059872,
      "step": 37020
    },
    {
      "epoch": 17.456388495992456,
      "grad_norm": 0.022432398051023483,
      "learning_rate": 0.004078765974223103,
      "loss": 0.0854,
      "num_input_tokens_seen": 32064128,
      "step": 37025
    },
    {
      "epoch": 17.458745874587457,
      "grad_norm": 0.01602267660200596,
      "learning_rate": 0.004065134148578564,
      "loss": 0.147,
      "num_input_tokens_seen": 32069232,
      "step": 37030
    },
    {
      "epoch": 17.46110325318246,
      "grad_norm": 0.008524245582520962,
      "learning_rate": 0.004051524827924279,
      "loss": 0.0823,
      "num_input_tokens_seen": 32073760,
      "step": 37035
    },
    {
      "epoch": 17.463460631777462,
      "grad_norm": 0.003973803948611021,
      "learning_rate": 0.004037938014358955,
      "loss": 0.0722,
      "num_input_tokens_seen": 32078944,
      "step": 37040
    },
    {
      "epoch": 17.465818010372466,
      "grad_norm": 0.00881690438836813,
      "learning_rate": 0.004024373709977863,
      "loss": 0.1095,
      "num_input_tokens_seen": 32082992,
      "step": 37045
    },
    {
      "epoch": 17.468175388967467,
      "grad_norm": 0.018982673063874245,
      "learning_rate": 0.004010831916872814,
      "loss": 0.0927,
      "num_input_tokens_seen": 32087584,
      "step": 37050
    },
    {
      "epoch": 17.47053276756247,
      "grad_norm": 0.011893784627318382,
      "learning_rate": 0.003997312637132089,
      "loss": 0.1159,
      "num_input_tokens_seen": 32091600,
      "step": 37055
    },
    {
      "epoch": 17.47289014615747,
      "grad_norm": 0.014219461008906364,
      "learning_rate": 0.003983815872840535,
      "loss": 0.122,
      "num_input_tokens_seen": 32096144,
      "step": 37060
    },
    {
      "epoch": 17.475247524752476,
      "grad_norm": 0.011056758463382721,
      "learning_rate": 0.003970341626079521,
      "loss": 0.0861,
      "num_input_tokens_seen": 32100624,
      "step": 37065
    },
    {
      "epoch": 17.477604903347476,
      "grad_norm": 0.007328271400183439,
      "learning_rate": 0.003956889898926952,
      "loss": 0.0917,
      "num_input_tokens_seen": 32104768,
      "step": 37070
    },
    {
      "epoch": 17.47996228194248,
      "grad_norm": 0.012774522416293621,
      "learning_rate": 0.0039434606934572675,
      "loss": 0.0815,
      "num_input_tokens_seen": 32108448,
      "step": 37075
    },
    {
      "epoch": 17.48231966053748,
      "grad_norm": 0.014154463075101376,
      "learning_rate": 0.003930054011741396,
      "loss": 0.0798,
      "num_input_tokens_seen": 32112784,
      "step": 37080
    },
    {
      "epoch": 17.484677039132485,
      "grad_norm": 0.005905671510845423,
      "learning_rate": 0.0039166698558468155,
      "loss": 0.0844,
      "num_input_tokens_seen": 32117104,
      "step": 37085
    },
    {
      "epoch": 17.487034417727486,
      "grad_norm": 0.007756276521831751,
      "learning_rate": 0.0039033082278375594,
      "loss": 0.0918,
      "num_input_tokens_seen": 32121984,
      "step": 37090
    },
    {
      "epoch": 17.48939179632249,
      "grad_norm": 0.017373360693454742,
      "learning_rate": 0.003889969129774112,
      "loss": 0.1276,
      "num_input_tokens_seen": 32126096,
      "step": 37095
    },
    {
      "epoch": 17.49174917491749,
      "grad_norm": 0.014663071371614933,
      "learning_rate": 0.0038766525637135784,
      "loss": 0.076,
      "num_input_tokens_seen": 32130016,
      "step": 37100
    },
    {
      "epoch": 17.494106553512495,
      "grad_norm": 0.0149783194065094,
      "learning_rate": 0.0038633585317095318,
      "loss": 0.1117,
      "num_input_tokens_seen": 32133632,
      "step": 37105
    },
    {
      "epoch": 17.496463932107496,
      "grad_norm": 0.017553722485899925,
      "learning_rate": 0.00385008703581205,
      "loss": 0.0894,
      "num_input_tokens_seen": 32137472,
      "step": 37110
    },
    {
      "epoch": 17.4988213107025,
      "grad_norm": 0.0034000156447291374,
      "learning_rate": 0.0038368380780677944,
      "loss": 0.0387,
      "num_input_tokens_seen": 32141472,
      "step": 37115
    },
    {
      "epoch": 17.5011786892975,
      "grad_norm": 0.010670600458979607,
      "learning_rate": 0.003823611660519882,
      "loss": 0.049,
      "num_input_tokens_seen": 32145840,
      "step": 37120
    },
    {
      "epoch": 17.503536067892504,
      "grad_norm": 0.012972015887498856,
      "learning_rate": 0.0038104077852080475,
      "loss": 0.0847,
      "num_input_tokens_seen": 32149664,
      "step": 37125
    },
    {
      "epoch": 17.505893446487505,
      "grad_norm": 0.01400159765034914,
      "learning_rate": 0.003797226454168462,
      "loss": 0.1395,
      "num_input_tokens_seen": 32154208,
      "step": 37130
    },
    {
      "epoch": 17.50825082508251,
      "grad_norm": 0.00900962669402361,
      "learning_rate": 0.003784067669433849,
      "loss": 0.0655,
      "num_input_tokens_seen": 32158688,
      "step": 37135
    },
    {
      "epoch": 17.51060820367751,
      "grad_norm": 0.010161323472857475,
      "learning_rate": 0.0037709314330334528,
      "loss": 0.0773,
      "num_input_tokens_seen": 32162512,
      "step": 37140
    },
    {
      "epoch": 17.512965582272514,
      "grad_norm": 0.009984783828258514,
      "learning_rate": 0.003757817746993086,
      "loss": 0.1263,
      "num_input_tokens_seen": 32166288,
      "step": 37145
    },
    {
      "epoch": 17.515322960867515,
      "grad_norm": 0.005558244418352842,
      "learning_rate": 0.0037447266133349977,
      "loss": 0.0796,
      "num_input_tokens_seen": 32171024,
      "step": 37150
    },
    {
      "epoch": 17.51768033946252,
      "grad_norm": 0.010638963431119919,
      "learning_rate": 0.003731658034078039,
      "loss": 0.0561,
      "num_input_tokens_seen": 32174496,
      "step": 37155
    },
    {
      "epoch": 17.52003771805752,
      "grad_norm": 0.012513309717178345,
      "learning_rate": 0.0037186120112375153,
      "loss": 0.1451,
      "num_input_tokens_seen": 32179120,
      "step": 37160
    },
    {
      "epoch": 17.522395096652524,
      "grad_norm": 0.011733729392290115,
      "learning_rate": 0.003705588546825317,
      "loss": 0.0608,
      "num_input_tokens_seen": 32182800,
      "step": 37165
    },
    {
      "epoch": 17.524752475247524,
      "grad_norm": 0.007248970679938793,
      "learning_rate": 0.0036925876428498205,
      "loss": 0.0571,
      "num_input_tokens_seen": 32187936,
      "step": 37170
    },
    {
      "epoch": 17.52710985384253,
      "grad_norm": 0.020274773240089417,
      "learning_rate": 0.0036796093013159057,
      "loss": 0.0995,
      "num_input_tokens_seen": 32191984,
      "step": 37175
    },
    {
      "epoch": 17.52946723243753,
      "grad_norm": 0.006151061039417982,
      "learning_rate": 0.0036666535242250217,
      "loss": 0.0581,
      "num_input_tokens_seen": 32196512,
      "step": 37180
    },
    {
      "epoch": 17.531824611032533,
      "grad_norm": 0.015256330370903015,
      "learning_rate": 0.003653720313575104,
      "loss": 0.1234,
      "num_input_tokens_seen": 32201168,
      "step": 37185
    },
    {
      "epoch": 17.534181989627534,
      "grad_norm": 0.011192667298018932,
      "learning_rate": 0.003640809671360623,
      "loss": 0.0523,
      "num_input_tokens_seen": 32205120,
      "step": 37190
    },
    {
      "epoch": 17.536539368222538,
      "grad_norm": 0.023285001516342163,
      "learning_rate": 0.003627921599572553,
      "loss": 0.0831,
      "num_input_tokens_seen": 32209440,
      "step": 37195
    },
    {
      "epoch": 17.53889674681754,
      "grad_norm": 0.028738277032971382,
      "learning_rate": 0.003615056100198405,
      "loss": 0.105,
      "num_input_tokens_seen": 32213584,
      "step": 37200
    },
    {
      "epoch": 17.53889674681754,
      "eval_loss": 0.28394991159439087,
      "eval_runtime": 21.9443,
      "eval_samples_per_second": 42.972,
      "eval_steps_per_second": 21.509,
      "num_input_tokens_seen": 32213584,
      "step": 37200
    },
    {
      "epoch": 17.541254125412543,
      "grad_norm": 0.006572001148015261,
      "learning_rate": 0.003602213175222174,
      "loss": 0.1017,
      "num_input_tokens_seen": 32217664,
      "step": 37205
    },
    {
      "epoch": 17.543611504007544,
      "grad_norm": 0.013822491280734539,
      "learning_rate": 0.0035893928266244432,
      "loss": 0.0583,
      "num_input_tokens_seen": 32222288,
      "step": 37210
    },
    {
      "epoch": 17.545968882602544,
      "grad_norm": 0.00720096193253994,
      "learning_rate": 0.003576595056382248,
      "loss": 0.0867,
      "num_input_tokens_seen": 32226720,
      "step": 37215
    },
    {
      "epoch": 17.54832626119755,
      "grad_norm": 0.0058188047260046005,
      "learning_rate": 0.0035638198664691423,
      "loss": 0.1175,
      "num_input_tokens_seen": 32230160,
      "step": 37220
    },
    {
      "epoch": 17.55068363979255,
      "grad_norm": 0.008298723958432674,
      "learning_rate": 0.003551067258855267,
      "loss": 0.1041,
      "num_input_tokens_seen": 32234784,
      "step": 37225
    },
    {
      "epoch": 17.553041018387553,
      "grad_norm": 0.016752494499087334,
      "learning_rate": 0.0035383372355071996,
      "loss": 0.1477,
      "num_input_tokens_seen": 32239920,
      "step": 37230
    },
    {
      "epoch": 17.555398396982554,
      "grad_norm": 0.025318287312984467,
      "learning_rate": 0.0035256297983881023,
      "loss": 0.2348,
      "num_input_tokens_seen": 32243888,
      "step": 37235
    },
    {
      "epoch": 17.557755775577558,
      "grad_norm": 0.01791287213563919,
      "learning_rate": 0.0035129449494575747,
      "loss": 0.0963,
      "num_input_tokens_seen": 32248480,
      "step": 37240
    },
    {
      "epoch": 17.56011315417256,
      "grad_norm": 0.002624685177579522,
      "learning_rate": 0.0035002826906718187,
      "loss": 0.0536,
      "num_input_tokens_seen": 32252848,
      "step": 37245
    },
    {
      "epoch": 17.562470532767563,
      "grad_norm": 0.01675652340054512,
      "learning_rate": 0.003487643023983522,
      "loss": 0.0984,
      "num_input_tokens_seen": 32257040,
      "step": 37250
    },
    {
      "epoch": 17.564827911362563,
      "grad_norm": 0.010680273175239563,
      "learning_rate": 0.003475025951341842,
      "loss": 0.1299,
      "num_input_tokens_seen": 32260768,
      "step": 37255
    },
    {
      "epoch": 17.567185289957568,
      "grad_norm": 0.028947947546839714,
      "learning_rate": 0.00346243147469249,
      "loss": 0.2178,
      "num_input_tokens_seen": 32264688,
      "step": 37260
    },
    {
      "epoch": 17.569542668552568,
      "grad_norm": 0.0035651070065796375,
      "learning_rate": 0.0034498595959777446,
      "loss": 0.1528,
      "num_input_tokens_seen": 32269792,
      "step": 37265
    },
    {
      "epoch": 17.571900047147572,
      "grad_norm": 0.014632598496973515,
      "learning_rate": 0.003437310317136305,
      "loss": 0.0711,
      "num_input_tokens_seen": 32274320,
      "step": 37270
    },
    {
      "epoch": 17.574257425742573,
      "grad_norm": 0.01866121217608452,
      "learning_rate": 0.0034247836401034236,
      "loss": 0.1379,
      "num_input_tokens_seen": 32278816,
      "step": 37275
    },
    {
      "epoch": 17.576614804337577,
      "grad_norm": 0.006002443376928568,
      "learning_rate": 0.003412279566810905,
      "loss": 0.0563,
      "num_input_tokens_seen": 32282816,
      "step": 37280
    },
    {
      "epoch": 17.578972182932578,
      "grad_norm": 0.02452298067510128,
      "learning_rate": 0.00339979809918699,
      "loss": 0.1561,
      "num_input_tokens_seen": 32286432,
      "step": 37285
    },
    {
      "epoch": 17.581329561527582,
      "grad_norm": 0.007970267906785011,
      "learning_rate": 0.0033873392391565228,
      "loss": 0.0857,
      "num_input_tokens_seen": 32290608,
      "step": 37290
    },
    {
      "epoch": 17.583686940122583,
      "grad_norm": 0.014925443567335606,
      "learning_rate": 0.003374902988640782,
      "loss": 0.1255,
      "num_input_tokens_seen": 32295344,
      "step": 37295
    },
    {
      "epoch": 17.586044318717587,
      "grad_norm": 0.013276524841785431,
      "learning_rate": 0.0033624893495576014,
      "loss": 0.0919,
      "num_input_tokens_seen": 32299728,
      "step": 37300
    },
    {
      "epoch": 17.588401697312587,
      "grad_norm": 0.0026049637235701084,
      "learning_rate": 0.0033500983238213323,
      "loss": 0.0503,
      "num_input_tokens_seen": 32303872,
      "step": 37305
    },
    {
      "epoch": 17.59075907590759,
      "grad_norm": 0.022344935685396194,
      "learning_rate": 0.0033377299133428126,
      "loss": 0.1128,
      "num_input_tokens_seen": 32307600,
      "step": 37310
    },
    {
      "epoch": 17.593116454502592,
      "grad_norm": 0.009955333545804024,
      "learning_rate": 0.003325384120029434,
      "loss": 0.0765,
      "num_input_tokens_seen": 32312016,
      "step": 37315
    },
    {
      "epoch": 17.595473833097596,
      "grad_norm": 0.006067886482924223,
      "learning_rate": 0.0033130609457850233,
      "loss": 0.0423,
      "num_input_tokens_seen": 32316160,
      "step": 37320
    },
    {
      "epoch": 17.597831211692597,
      "grad_norm": 0.01821281388401985,
      "learning_rate": 0.0033007603925100104,
      "loss": 0.1538,
      "num_input_tokens_seen": 32320480,
      "step": 37325
    },
    {
      "epoch": 17.6001885902876,
      "grad_norm": 0.025433285161852837,
      "learning_rate": 0.003288482462101294,
      "loss": 0.0863,
      "num_input_tokens_seen": 32325520,
      "step": 37330
    },
    {
      "epoch": 17.602545968882602,
      "grad_norm": 0.00830509327352047,
      "learning_rate": 0.0032762271564522605,
      "loss": 0.1245,
      "num_input_tokens_seen": 32330288,
      "step": 37335
    },
    {
      "epoch": 17.604903347477606,
      "grad_norm": 0.015903238207101822,
      "learning_rate": 0.003263994477452864,
      "loss": 0.1014,
      "num_input_tokens_seen": 32333824,
      "step": 37340
    },
    {
      "epoch": 17.607260726072607,
      "grad_norm": 0.023675723001360893,
      "learning_rate": 0.0032517844269895125,
      "loss": 0.0691,
      "num_input_tokens_seen": 32338240,
      "step": 37345
    },
    {
      "epoch": 17.60961810466761,
      "grad_norm": 0.009697126224637032,
      "learning_rate": 0.0032395970069451496,
      "loss": 0.096,
      "num_input_tokens_seen": 32342336,
      "step": 37350
    },
    {
      "epoch": 17.61197548326261,
      "grad_norm": 0.020435182377696037,
      "learning_rate": 0.0032274322191992388,
      "loss": 0.1761,
      "num_input_tokens_seen": 32346512,
      "step": 37355
    },
    {
      "epoch": 17.614332861857616,
      "grad_norm": 0.00927786622196436,
      "learning_rate": 0.0032152900656277294,
      "loss": 0.1245,
      "num_input_tokens_seen": 32350672,
      "step": 37360
    },
    {
      "epoch": 17.616690240452616,
      "grad_norm": 0.01766171306371689,
      "learning_rate": 0.0032031705481030902,
      "loss": 0.0984,
      "num_input_tokens_seen": 32355408,
      "step": 37365
    },
    {
      "epoch": 17.61904761904762,
      "grad_norm": 0.012787561863660812,
      "learning_rate": 0.0031910736684943428,
      "loss": 0.1077,
      "num_input_tokens_seen": 32359648,
      "step": 37370
    },
    {
      "epoch": 17.62140499764262,
      "grad_norm": 0.016698474064469337,
      "learning_rate": 0.0031789994286669453,
      "loss": 0.0796,
      "num_input_tokens_seen": 32363952,
      "step": 37375
    },
    {
      "epoch": 17.623762376237625,
      "grad_norm": 0.028920453041791916,
      "learning_rate": 0.003166947830482908,
      "loss": 0.0849,
      "num_input_tokens_seen": 32368544,
      "step": 37380
    },
    {
      "epoch": 17.626119754832626,
      "grad_norm": 0.005374092608690262,
      "learning_rate": 0.003154918875800727,
      "loss": 0.0817,
      "num_input_tokens_seen": 32373184,
      "step": 37385
    },
    {
      "epoch": 17.62847713342763,
      "grad_norm": 0.009427309036254883,
      "learning_rate": 0.00314291256647542,
      "loss": 0.0646,
      "num_input_tokens_seen": 32377232,
      "step": 37390
    },
    {
      "epoch": 17.63083451202263,
      "grad_norm": 0.0013187407748773694,
      "learning_rate": 0.0031309289043585375,
      "loss": 0.0657,
      "num_input_tokens_seen": 32381008,
      "step": 37395
    },
    {
      "epoch": 17.633191890617635,
      "grad_norm": 0.01300282496958971,
      "learning_rate": 0.003118967891298069,
      "loss": 0.0659,
      "num_input_tokens_seen": 32385920,
      "step": 37400
    },
    {
      "epoch": 17.633191890617635,
      "eval_loss": 0.2861294150352478,
      "eval_runtime": 21.9042,
      "eval_samples_per_second": 43.051,
      "eval_steps_per_second": 21.548,
      "num_input_tokens_seen": 32385920,
      "step": 37400
    },
    {
      "epoch": 17.635549269212635,
      "grad_norm": 0.011255573481321335,
      "learning_rate": 0.003107029529138572,
      "loss": 0.0893,
      "num_input_tokens_seen": 32390528,
      "step": 37405
    },
    {
      "epoch": 17.63790664780764,
      "grad_norm": 0.0017460254020988941,
      "learning_rate": 0.0030951138197211235,
      "loss": 0.1768,
      "num_input_tokens_seen": 32395040,
      "step": 37410
    },
    {
      "epoch": 17.64026402640264,
      "grad_norm": 0.012948630377650261,
      "learning_rate": 0.0030832207648832377,
      "loss": 0.0875,
      "num_input_tokens_seen": 32399440,
      "step": 37415
    },
    {
      "epoch": 17.64262140499764,
      "grad_norm": 0.008558126166462898,
      "learning_rate": 0.0030713503664589635,
      "loss": 0.0797,
      "num_input_tokens_seen": 32403184,
      "step": 37420
    },
    {
      "epoch": 17.644978783592645,
      "grad_norm": 0.004658039193600416,
      "learning_rate": 0.0030595026262788872,
      "loss": 0.1151,
      "num_input_tokens_seen": 32406848,
      "step": 37425
    },
    {
      "epoch": 17.647336162187646,
      "grad_norm": 0.016357524320483208,
      "learning_rate": 0.00304767754617008,
      "loss": 0.0802,
      "num_input_tokens_seen": 32411280,
      "step": 37430
    },
    {
      "epoch": 17.64969354078265,
      "grad_norm": 0.017679346725344658,
      "learning_rate": 0.003035875127956117,
      "loss": 0.0913,
      "num_input_tokens_seen": 32416000,
      "step": 37435
    },
    {
      "epoch": 17.65205091937765,
      "grad_norm": 0.007143513765186071,
      "learning_rate": 0.0030240953734570752,
      "loss": 0.12,
      "num_input_tokens_seen": 32419760,
      "step": 37440
    },
    {
      "epoch": 17.654408297972655,
      "grad_norm": 0.01890089362859726,
      "learning_rate": 0.003012338284489535,
      "loss": 0.0869,
      "num_input_tokens_seen": 32423408,
      "step": 37445
    },
    {
      "epoch": 17.656765676567655,
      "grad_norm": 0.009493334218859673,
      "learning_rate": 0.0030006038628665964,
      "loss": 0.0664,
      "num_input_tokens_seen": 32427728,
      "step": 37450
    },
    {
      "epoch": 17.65912305516266,
      "grad_norm": 0.019914016127586365,
      "learning_rate": 0.002988892110397845,
      "loss": 0.0883,
      "num_input_tokens_seen": 32431920,
      "step": 37455
    },
    {
      "epoch": 17.66148043375766,
      "grad_norm": 0.016373831778764725,
      "learning_rate": 0.0029772030288894025,
      "loss": 0.1019,
      "num_input_tokens_seen": 32435696,
      "step": 37460
    },
    {
      "epoch": 17.663837812352664,
      "grad_norm": 0.014126284047961235,
      "learning_rate": 0.0029655366201438438,
      "loss": 0.0993,
      "num_input_tokens_seen": 32440208,
      "step": 37465
    },
    {
      "epoch": 17.666195190947665,
      "grad_norm": 0.01077905111014843,
      "learning_rate": 0.0029538928859602965,
      "loss": 0.1074,
      "num_input_tokens_seen": 32444576,
      "step": 37470
    },
    {
      "epoch": 17.66855256954267,
      "grad_norm": 0.009381688199937344,
      "learning_rate": 0.002942271828134374,
      "loss": 0.1603,
      "num_input_tokens_seen": 32448864,
      "step": 37475
    },
    {
      "epoch": 17.67090994813767,
      "grad_norm": 0.016609903424978256,
      "learning_rate": 0.00293067344845816,
      "loss": 0.0819,
      "num_input_tokens_seen": 32452800,
      "step": 37480
    },
    {
      "epoch": 17.673267326732674,
      "grad_norm": 0.013592839241027832,
      "learning_rate": 0.0029190977487202896,
      "loss": 0.1395,
      "num_input_tokens_seen": 32456576,
      "step": 37485
    },
    {
      "epoch": 17.675624705327674,
      "grad_norm": 0.006814047694206238,
      "learning_rate": 0.0029075447307058853,
      "loss": 0.0666,
      "num_input_tokens_seen": 32461504,
      "step": 37490
    },
    {
      "epoch": 17.67798208392268,
      "grad_norm": 0.006972964387387037,
      "learning_rate": 0.0028960143961965722,
      "loss": 0.0777,
      "num_input_tokens_seen": 32465904,
      "step": 37495
    },
    {
      "epoch": 17.68033946251768,
      "grad_norm": 0.009058712050318718,
      "learning_rate": 0.002884506746970461,
      "loss": 0.1298,
      "num_input_tokens_seen": 32469888,
      "step": 37500
    },
    {
      "epoch": 17.682696841112683,
      "grad_norm": 0.011921214871108532,
      "learning_rate": 0.0028730217848021654,
      "loss": 0.075,
      "num_input_tokens_seen": 32473696,
      "step": 37505
    },
    {
      "epoch": 17.685054219707684,
      "grad_norm": 0.01218518428504467,
      "learning_rate": 0.0028615595114628188,
      "loss": 0.1105,
      "num_input_tokens_seen": 32478128,
      "step": 37510
    },
    {
      "epoch": 17.68741159830269,
      "grad_norm": 0.014437354169785976,
      "learning_rate": 0.002850119928720074,
      "loss": 0.1168,
      "num_input_tokens_seen": 32482352,
      "step": 37515
    },
    {
      "epoch": 17.68976897689769,
      "grad_norm": 0.005329999607056379,
      "learning_rate": 0.0028387030383380195,
      "loss": 0.0968,
      "num_input_tokens_seen": 32486432,
      "step": 37520
    },
    {
      "epoch": 17.692126355492693,
      "grad_norm": 0.011276536621153355,
      "learning_rate": 0.0028273088420772974,
      "loss": 0.0545,
      "num_input_tokens_seen": 32490784,
      "step": 37525
    },
    {
      "epoch": 17.694483734087694,
      "grad_norm": 0.01851036585867405,
      "learning_rate": 0.002815937341695068,
      "loss": 0.0982,
      "num_input_tokens_seen": 32495744,
      "step": 37530
    },
    {
      "epoch": 17.696841112682698,
      "grad_norm": 0.02714168280363083,
      "learning_rate": 0.0028045885389448963,
      "loss": 0.1136,
      "num_input_tokens_seen": 32499408,
      "step": 37535
    },
    {
      "epoch": 17.6991984912777,
      "grad_norm": 0.015347837470471859,
      "learning_rate": 0.002793262435576965,
      "loss": 0.1077,
      "num_input_tokens_seen": 32504096,
      "step": 37540
    },
    {
      "epoch": 17.701555869872703,
      "grad_norm": 0.0026802239008247852,
      "learning_rate": 0.0027819590333378772,
      "loss": 0.0792,
      "num_input_tokens_seen": 32508144,
      "step": 37545
    },
    {
      "epoch": 17.703913248467703,
      "grad_norm": 0.01572924666106701,
      "learning_rate": 0.002770678333970755,
      "loss": 0.1436,
      "num_input_tokens_seen": 32512352,
      "step": 37550
    },
    {
      "epoch": 17.706270627062707,
      "grad_norm": 0.004579825326800346,
      "learning_rate": 0.0027594203392152573,
      "loss": 0.0852,
      "num_input_tokens_seen": 32516576,
      "step": 37555
    },
    {
      "epoch": 17.708628005657708,
      "grad_norm": 0.017606457695364952,
      "learning_rate": 0.002748185050807478,
      "loss": 0.1346,
      "num_input_tokens_seen": 32521264,
      "step": 37560
    },
    {
      "epoch": 17.710985384252712,
      "grad_norm": 0.005874218884855509,
      "learning_rate": 0.002736972470480031,
      "loss": 0.1962,
      "num_input_tokens_seen": 32525456,
      "step": 37565
    },
    {
      "epoch": 17.713342762847713,
      "grad_norm": 0.015978962182998657,
      "learning_rate": 0.002725782599962068,
      "loss": 0.1892,
      "num_input_tokens_seen": 32529648,
      "step": 37570
    },
    {
      "epoch": 17.715700141442717,
      "grad_norm": 0.020400643348693848,
      "learning_rate": 0.0027146154409791734,
      "loss": 0.1237,
      "num_input_tokens_seen": 32534176,
      "step": 37575
    },
    {
      "epoch": 17.718057520037718,
      "grad_norm": 0.018990276381373405,
      "learning_rate": 0.002703470995253504,
      "loss": 0.1115,
      "num_input_tokens_seen": 32538080,
      "step": 37580
    },
    {
      "epoch": 17.720414898632722,
      "grad_norm": 0.006479684263467789,
      "learning_rate": 0.0026923492645036184,
      "loss": 0.0851,
      "num_input_tokens_seen": 32542848,
      "step": 37585
    },
    {
      "epoch": 17.722772277227723,
      "grad_norm": 0.003159410087391734,
      "learning_rate": 0.0026812502504446776,
      "loss": 0.0339,
      "num_input_tokens_seen": 32547456,
      "step": 37590
    },
    {
      "epoch": 17.725129655822727,
      "grad_norm": 0.029272079467773438,
      "learning_rate": 0.0026701739547882798,
      "loss": 0.1044,
      "num_input_tokens_seen": 32551168,
      "step": 37595
    },
    {
      "epoch": 17.727487034417727,
      "grad_norm": 0.019184965640306473,
      "learning_rate": 0.0026591203792425077,
      "loss": 0.094,
      "num_input_tokens_seen": 32555856,
      "step": 37600
    },
    {
      "epoch": 17.727487034417727,
      "eval_loss": 0.28277096152305603,
      "eval_runtime": 21.8781,
      "eval_samples_per_second": 43.102,
      "eval_steps_per_second": 21.574,
      "num_input_tokens_seen": 32555856,
      "step": 37600
    },
    {
      "epoch": 17.72984441301273,
      "grad_norm": 0.012827574275434017,
      "learning_rate": 0.0026480895255119818,
      "loss": 0.0603,
      "num_input_tokens_seen": 32560112,
      "step": 37605
    },
    {
      "epoch": 17.732201791607732,
      "grad_norm": 0.006732284557074308,
      "learning_rate": 0.002637081395297791,
      "loss": 0.0978,
      "num_input_tokens_seen": 32563312,
      "step": 37610
    },
    {
      "epoch": 17.734559170202736,
      "grad_norm": 0.0031669391319155693,
      "learning_rate": 0.0026260959902975113,
      "loss": 0.0983,
      "num_input_tokens_seen": 32567104,
      "step": 37615
    },
    {
      "epoch": 17.736916548797737,
      "grad_norm": 0.030518831685185432,
      "learning_rate": 0.00261513331220527,
      "loss": 0.1412,
      "num_input_tokens_seen": 32570896,
      "step": 37620
    },
    {
      "epoch": 17.739273927392738,
      "grad_norm": 0.02441331557929516,
      "learning_rate": 0.0026041933627116154,
      "loss": 0.0618,
      "num_input_tokens_seen": 32574768,
      "step": 37625
    },
    {
      "epoch": 17.74163130598774,
      "grad_norm": 0.005071558523923159,
      "learning_rate": 0.0025932761435036476,
      "loss": 0.0604,
      "num_input_tokens_seen": 32579024,
      "step": 37630
    },
    {
      "epoch": 17.743988684582742,
      "grad_norm": 0.012346313335001469,
      "learning_rate": 0.002582381656264904,
      "loss": 0.0949,
      "num_input_tokens_seen": 32584624,
      "step": 37635
    },
    {
      "epoch": 17.746346063177747,
      "grad_norm": 0.014568074606359005,
      "learning_rate": 0.0025715099026754895,
      "loss": 0.1068,
      "num_input_tokens_seen": 32589120,
      "step": 37640
    },
    {
      "epoch": 17.748703441772747,
      "grad_norm": 0.016098206862807274,
      "learning_rate": 0.002560660884411947,
      "loss": 0.0926,
      "num_input_tokens_seen": 32593872,
      "step": 37645
    },
    {
      "epoch": 17.75106082036775,
      "grad_norm": 0.00982583686709404,
      "learning_rate": 0.0025498346031473385,
      "loss": 0.0875,
      "num_input_tokens_seen": 32598016,
      "step": 37650
    },
    {
      "epoch": 17.753418198962752,
      "grad_norm": 0.012439287267625332,
      "learning_rate": 0.0025390310605511945,
      "loss": 0.1056,
      "num_input_tokens_seen": 32602240,
      "step": 37655
    },
    {
      "epoch": 17.755775577557756,
      "grad_norm": 0.01763077825307846,
      "learning_rate": 0.0025282502582895995,
      "loss": 0.0586,
      "num_input_tokens_seen": 32607216,
      "step": 37660
    },
    {
      "epoch": 17.758132956152757,
      "grad_norm": 0.025673573836684227,
      "learning_rate": 0.002517492198025023,
      "loss": 0.0939,
      "num_input_tokens_seen": 32611392,
      "step": 37665
    },
    {
      "epoch": 17.76049033474776,
      "grad_norm": 0.009384541772305965,
      "learning_rate": 0.0025067568814165554,
      "loss": 0.0787,
      "num_input_tokens_seen": 32614752,
      "step": 37670
    },
    {
      "epoch": 17.76284771334276,
      "grad_norm": 0.04071333259344101,
      "learning_rate": 0.0024960443101196884,
      "loss": 0.1187,
      "num_input_tokens_seen": 32618864,
      "step": 37675
    },
    {
      "epoch": 17.765205091937766,
      "grad_norm": 0.0332196019589901,
      "learning_rate": 0.002485354485786434,
      "loss": 0.1351,
      "num_input_tokens_seen": 32622656,
      "step": 37680
    },
    {
      "epoch": 17.767562470532766,
      "grad_norm": 0.01608564704656601,
      "learning_rate": 0.002474687410065307,
      "loss": 0.0966,
      "num_input_tokens_seen": 32626544,
      "step": 37685
    },
    {
      "epoch": 17.76991984912777,
      "grad_norm": 0.007058768067508936,
      "learning_rate": 0.002464043084601308,
      "loss": 0.0811,
      "num_input_tokens_seen": 32632688,
      "step": 37690
    },
    {
      "epoch": 17.77227722772277,
      "grad_norm": 0.019096404314041138,
      "learning_rate": 0.0024534215110358915,
      "loss": 0.0927,
      "num_input_tokens_seen": 32637120,
      "step": 37695
    },
    {
      "epoch": 17.774634606317775,
      "grad_norm": 0.008591579273343086,
      "learning_rate": 0.002442822691007096,
      "loss": 0.0939,
      "num_input_tokens_seen": 32641328,
      "step": 37700
    },
    {
      "epoch": 17.776991984912776,
      "grad_norm": 0.004804564174264669,
      "learning_rate": 0.002432246626149348,
      "loss": 0.0823,
      "num_input_tokens_seen": 32645216,
      "step": 37705
    },
    {
      "epoch": 17.77934936350778,
      "grad_norm": 0.009266759268939495,
      "learning_rate": 0.002421693318093626,
      "loss": 0.119,
      "num_input_tokens_seen": 32649456,
      "step": 37710
    },
    {
      "epoch": 17.78170674210278,
      "grad_norm": 0.0020745634101331234,
      "learning_rate": 0.0024111627684673784,
      "loss": 0.1856,
      "num_input_tokens_seen": 32654400,
      "step": 37715
    },
    {
      "epoch": 17.784064120697785,
      "grad_norm": 0.013148010708391666,
      "learning_rate": 0.0024006549788945395,
      "loss": 0.1064,
      "num_input_tokens_seen": 32659600,
      "step": 37720
    },
    {
      "epoch": 17.786421499292786,
      "grad_norm": 0.0368446446955204,
      "learning_rate": 0.0023901699509955463,
      "loss": 0.1446,
      "num_input_tokens_seen": 32663968,
      "step": 37725
    },
    {
      "epoch": 17.78877887788779,
      "grad_norm": 0.004968029912561178,
      "learning_rate": 0.0023797076863873554,
      "loss": 0.0615,
      "num_input_tokens_seen": 32668160,
      "step": 37730
    },
    {
      "epoch": 17.79113625648279,
      "grad_norm": 0.00618192320689559,
      "learning_rate": 0.0023692681866833262,
      "loss": 0.0591,
      "num_input_tokens_seen": 32672480,
      "step": 37735
    },
    {
      "epoch": 17.793493635077795,
      "grad_norm": 0.024555517360568047,
      "learning_rate": 0.0023588514534934046,
      "loss": 0.1072,
      "num_input_tokens_seen": 32676688,
      "step": 37740
    },
    {
      "epoch": 17.795851013672795,
      "grad_norm": 0.012568702921271324,
      "learning_rate": 0.002348457488423955,
      "loss": 0.0657,
      "num_input_tokens_seen": 32682080,
      "step": 37745
    },
    {
      "epoch": 17.7982083922678,
      "grad_norm": 0.011449404992163181,
      "learning_rate": 0.0023380862930778624,
      "loss": 0.1276,
      "num_input_tokens_seen": 32686432,
      "step": 37750
    },
    {
      "epoch": 17.8005657708628,
      "grad_norm": 0.025019392371177673,
      "learning_rate": 0.0023277378690545135,
      "loss": 0.0928,
      "num_input_tokens_seen": 32690384,
      "step": 37755
    },
    {
      "epoch": 17.802923149457804,
      "grad_norm": 0.003095651278272271,
      "learning_rate": 0.0023174122179497325,
      "loss": 0.1078,
      "num_input_tokens_seen": 32694912,
      "step": 37760
    },
    {
      "epoch": 17.805280528052805,
      "grad_norm": 0.0078053465113043785,
      "learning_rate": 0.0023071093413558784,
      "loss": 0.1156,
      "num_input_tokens_seen": 32699856,
      "step": 37765
    },
    {
      "epoch": 17.80763790664781,
      "grad_norm": 0.016472671180963516,
      "learning_rate": 0.002296829240861814,
      "loss": 0.1026,
      "num_input_tokens_seen": 32704544,
      "step": 37770
    },
    {
      "epoch": 17.80999528524281,
      "grad_norm": 0.007889578118920326,
      "learning_rate": 0.002286571918052821,
      "loss": 0.0774,
      "num_input_tokens_seen": 32708256,
      "step": 37775
    },
    {
      "epoch": 17.812352663837814,
      "grad_norm": 0.017487863078713417,
      "learning_rate": 0.0022763373745107174,
      "loss": 0.1092,
      "num_input_tokens_seen": 32712368,
      "step": 37780
    },
    {
      "epoch": 17.814710042432814,
      "grad_norm": 0.013089492917060852,
      "learning_rate": 0.0022661256118138074,
      "loss": 0.0532,
      "num_input_tokens_seen": 32717088,
      "step": 37785
    },
    {
      "epoch": 17.81706742102782,
      "grad_norm": 0.007651366759091616,
      "learning_rate": 0.0022559366315368645,
      "loss": 0.0673,
      "num_input_tokens_seen": 32720816,
      "step": 37790
    },
    {
      "epoch": 17.81942479962282,
      "grad_norm": 0.011137564666569233,
      "learning_rate": 0.002245770435251182,
      "loss": 0.0738,
      "num_input_tokens_seen": 32725120,
      "step": 37795
    },
    {
      "epoch": 17.821782178217823,
      "grad_norm": 0.011898873373866081,
      "learning_rate": 0.002235627024524456,
      "loss": 0.113,
      "num_input_tokens_seen": 32729024,
      "step": 37800
    },
    {
      "epoch": 17.821782178217823,
      "eval_loss": 0.2855503559112549,
      "eval_runtime": 21.9134,
      "eval_samples_per_second": 43.033,
      "eval_steps_per_second": 21.539,
      "num_input_tokens_seen": 32729024,
      "step": 37800
    },
    {
      "epoch": 17.824139556812824,
      "grad_norm": 0.013530240394175053,
      "learning_rate": 0.0022255064009209847,
      "loss": 0.0791,
      "num_input_tokens_seen": 32733504,
      "step": 37805
    },
    {
      "epoch": 17.826496935407828,
      "grad_norm": 0.0020145338494330645,
      "learning_rate": 0.0022154085660014864,
      "loss": 0.0682,
      "num_input_tokens_seen": 32738192,
      "step": 37810
    },
    {
      "epoch": 17.82885431400283,
      "grad_norm": 0.00931021198630333,
      "learning_rate": 0.0022053335213231494,
      "loss": 0.1012,
      "num_input_tokens_seen": 32742304,
      "step": 37815
    },
    {
      "epoch": 17.831211692597833,
      "grad_norm": 0.012183364480733871,
      "learning_rate": 0.002195281268439697,
      "loss": 0.115,
      "num_input_tokens_seen": 32747328,
      "step": 37820
    },
    {
      "epoch": 17.833569071192834,
      "grad_norm": 0.01888197660446167,
      "learning_rate": 0.002185251808901306,
      "loss": 0.0931,
      "num_input_tokens_seen": 32751552,
      "step": 37825
    },
    {
      "epoch": 17.835926449787834,
      "grad_norm": 0.020053604617714882,
      "learning_rate": 0.0021752451442546227,
      "loss": 0.0929,
      "num_input_tokens_seen": 32756144,
      "step": 37830
    },
    {
      "epoch": 17.83828382838284,
      "grad_norm": 0.03165505826473236,
      "learning_rate": 0.0021652612760428456,
      "loss": 0.0873,
      "num_input_tokens_seen": 32760208,
      "step": 37835
    },
    {
      "epoch": 17.84064120697784,
      "grad_norm": 0.024324428290128708,
      "learning_rate": 0.0021553002058055603,
      "loss": 0.1281,
      "num_input_tokens_seen": 32764960,
      "step": 37840
    },
    {
      "epoch": 17.842998585572843,
      "grad_norm": 0.0026885471306741238,
      "learning_rate": 0.0021453619350789376,
      "loss": 0.0637,
      "num_input_tokens_seen": 32769760,
      "step": 37845
    },
    {
      "epoch": 17.845355964167844,
      "grad_norm": 0.004285980481654406,
      "learning_rate": 0.0021354464653955516,
      "loss": 0.0907,
      "num_input_tokens_seen": 32774432,
      "step": 37850
    },
    {
      "epoch": 17.847713342762848,
      "grad_norm": 0.008758455514907837,
      "learning_rate": 0.002125553798284513,
      "loss": 0.045,
      "num_input_tokens_seen": 32779056,
      "step": 37855
    },
    {
      "epoch": 17.85007072135785,
      "grad_norm": 0.010427780449390411,
      "learning_rate": 0.002115683935271384,
      "loss": 0.1111,
      "num_input_tokens_seen": 32784080,
      "step": 37860
    },
    {
      "epoch": 17.852428099952853,
      "grad_norm": 0.016013622283935547,
      "learning_rate": 0.0021058368778782144,
      "loss": 0.165,
      "num_input_tokens_seen": 32788208,
      "step": 37865
    },
    {
      "epoch": 17.854785478547853,
      "grad_norm": 0.01217940729111433,
      "learning_rate": 0.002096012627623539,
      "loss": 0.0988,
      "num_input_tokens_seen": 32792576,
      "step": 37870
    },
    {
      "epoch": 17.857142857142858,
      "grad_norm": 0.007289068307727575,
      "learning_rate": 0.00208621118602243,
      "loss": 0.0632,
      "num_input_tokens_seen": 32796928,
      "step": 37875
    },
    {
      "epoch": 17.85950023573786,
      "grad_norm": 0.03092428669333458,
      "learning_rate": 0.002076432554586327,
      "loss": 0.1224,
      "num_input_tokens_seen": 32801008,
      "step": 37880
    },
    {
      "epoch": 17.861857614332862,
      "grad_norm": 0.019624466076493263,
      "learning_rate": 0.002066676734823258,
      "loss": 0.1862,
      "num_input_tokens_seen": 32805520,
      "step": 37885
    },
    {
      "epoch": 17.864214992927863,
      "grad_norm": 0.00908893533051014,
      "learning_rate": 0.0020569437282376866,
      "loss": 0.0858,
      "num_input_tokens_seen": 32809808,
      "step": 37890
    },
    {
      "epoch": 17.866572371522867,
      "grad_norm": 0.013917027972638607,
      "learning_rate": 0.002047233536330545,
      "loss": 0.0693,
      "num_input_tokens_seen": 32814624,
      "step": 37895
    },
    {
      "epoch": 17.868929750117868,
      "grad_norm": 0.0022596593480557203,
      "learning_rate": 0.0020375461605993015,
      "loss": 0.1049,
      "num_input_tokens_seen": 32818208,
      "step": 37900
    },
    {
      "epoch": 17.871287128712872,
      "grad_norm": 0.009653734974563122,
      "learning_rate": 0.002027881602537845,
      "loss": 0.0833,
      "num_input_tokens_seen": 32821760,
      "step": 37905
    },
    {
      "epoch": 17.873644507307873,
      "grad_norm": 0.038967959582805634,
      "learning_rate": 0.002018239863636567,
      "loss": 0.0915,
      "num_input_tokens_seen": 32825344,
      "step": 37910
    },
    {
      "epoch": 17.876001885902877,
      "grad_norm": 0.0022974100429564714,
      "learning_rate": 0.002008620945382378,
      "loss": 0.0888,
      "num_input_tokens_seen": 32829920,
      "step": 37915
    },
    {
      "epoch": 17.878359264497877,
      "grad_norm": 0.007134982384741306,
      "learning_rate": 0.001999024849258607,
      "loss": 0.1368,
      "num_input_tokens_seen": 32834112,
      "step": 37920
    },
    {
      "epoch": 17.88071664309288,
      "grad_norm": 0.013345448300242424,
      "learning_rate": 0.001989451576745105,
      "loss": 0.1075,
      "num_input_tokens_seen": 32838240,
      "step": 37925
    },
    {
      "epoch": 17.883074021687882,
      "grad_norm": 0.02551189623773098,
      "learning_rate": 0.00197990112931819,
      "loss": 0.2088,
      "num_input_tokens_seen": 32842432,
      "step": 37930
    },
    {
      "epoch": 17.885431400282886,
      "grad_norm": 0.012903011403977871,
      "learning_rate": 0.0019703735084506345,
      "loss": 0.0801,
      "num_input_tokens_seen": 32846864,
      "step": 37935
    },
    {
      "epoch": 17.887788778877887,
      "grad_norm": 0.01229048054665327,
      "learning_rate": 0.001960868715611763,
      "loss": 0.1194,
      "num_input_tokens_seen": 32851632,
      "step": 37940
    },
    {
      "epoch": 17.89014615747289,
      "grad_norm": 0.004790354520082474,
      "learning_rate": 0.0019513867522673034,
      "loss": 0.0713,
      "num_input_tokens_seen": 32856464,
      "step": 37945
    },
    {
      "epoch": 17.892503536067892,
      "grad_norm": 0.003282822435721755,
      "learning_rate": 0.001941927619879502,
      "loss": 0.0602,
      "num_input_tokens_seen": 32860912,
      "step": 37950
    },
    {
      "epoch": 17.894860914662896,
      "grad_norm": 0.009283834137022495,
      "learning_rate": 0.0019324913199070758,
      "loss": 0.1098,
      "num_input_tokens_seen": 32864816,
      "step": 37955
    },
    {
      "epoch": 17.897218293257897,
      "grad_norm": 0.003941444214433432,
      "learning_rate": 0.0019230778538052106,
      "loss": 0.0812,
      "num_input_tokens_seen": 32869056,
      "step": 37960
    },
    {
      "epoch": 17.8995756718529,
      "grad_norm": 0.02222757413983345,
      "learning_rate": 0.0019136872230255952,
      "loss": 0.1188,
      "num_input_tokens_seen": 32872816,
      "step": 37965
    },
    {
      "epoch": 17.9019330504479,
      "grad_norm": 0.030342984944581985,
      "learning_rate": 0.0019043194290164045,
      "loss": 0.1176,
      "num_input_tokens_seen": 32876128,
      "step": 37970
    },
    {
      "epoch": 17.904290429042906,
      "grad_norm": 0.0014170166105031967,
      "learning_rate": 0.0018949744732222162,
      "loss": 0.112,
      "num_input_tokens_seen": 32880960,
      "step": 37975
    },
    {
      "epoch": 17.906647807637906,
      "grad_norm": 0.012112445197999477,
      "learning_rate": 0.0018856523570841776,
      "loss": 0.0843,
      "num_input_tokens_seen": 32885552,
      "step": 37980
    },
    {
      "epoch": 17.90900518623291,
      "grad_norm": 0.012128538452088833,
      "learning_rate": 0.0018763530820398555,
      "loss": 0.1095,
      "num_input_tokens_seen": 32889664,
      "step": 37985
    },
    {
      "epoch": 17.91136256482791,
      "grad_norm": 0.018887383863329887,
      "learning_rate": 0.0018670766495233525,
      "loss": 0.0871,
      "num_input_tokens_seen": 32894832,
      "step": 37990
    },
    {
      "epoch": 17.913719943422915,
      "grad_norm": 0.002926361048594117,
      "learning_rate": 0.001857823060965158,
      "loss": 0.1675,
      "num_input_tokens_seen": 32898880,
      "step": 37995
    },
    {
      "epoch": 17.916077322017916,
      "grad_norm": 0.011991100385785103,
      "learning_rate": 0.0018485923177923467,
      "loss": 0.0887,
      "num_input_tokens_seen": 32902832,
      "step": 38000
    },
    {
      "epoch": 17.916077322017916,
      "eval_loss": 0.28853240609169006,
      "eval_runtime": 21.8656,
      "eval_samples_per_second": 43.127,
      "eval_steps_per_second": 21.586,
      "num_input_tokens_seen": 32902832,
      "step": 38000
    },
    {
      "epoch": 17.91843470061292,
      "grad_norm": 0.01959872432053089,
      "learning_rate": 0.001839384421428364,
      "loss": 0.1125,
      "num_input_tokens_seen": 32906896,
      "step": 38005
    },
    {
      "epoch": 17.92079207920792,
      "grad_norm": 0.012782145291566849,
      "learning_rate": 0.0018301993732932065,
      "loss": 0.1213,
      "num_input_tokens_seen": 32911168,
      "step": 38010
    },
    {
      "epoch": 17.92314945780292,
      "grad_norm": 0.012649822980165482,
      "learning_rate": 0.0018210371748033248,
      "loss": 0.0448,
      "num_input_tokens_seen": 32915664,
      "step": 38015
    },
    {
      "epoch": 17.925506836397926,
      "grad_norm": 0.00946555845439434,
      "learning_rate": 0.0018118978273716556,
      "loss": 0.0908,
      "num_input_tokens_seen": 32919824,
      "step": 38020
    },
    {
      "epoch": 17.927864214992926,
      "grad_norm": 0.014002281241118908,
      "learning_rate": 0.001802781332407588,
      "loss": 0.1055,
      "num_input_tokens_seen": 32925040,
      "step": 38025
    },
    {
      "epoch": 17.93022159358793,
      "grad_norm": 0.006152820307761431,
      "learning_rate": 0.0017936876913169806,
      "loss": 0.0682,
      "num_input_tokens_seen": 32929040,
      "step": 38030
    },
    {
      "epoch": 17.93257897218293,
      "grad_norm": 0.006718302145600319,
      "learning_rate": 0.0017846169055022287,
      "loss": 0.0528,
      "num_input_tokens_seen": 32933312,
      "step": 38035
    },
    {
      "epoch": 17.934936350777935,
      "grad_norm": 0.011343470774590969,
      "learning_rate": 0.0017755689763621295,
      "loss": 0.155,
      "num_input_tokens_seen": 32937648,
      "step": 38040
    },
    {
      "epoch": 17.937293729372936,
      "grad_norm": 0.025656158104538918,
      "learning_rate": 0.0017665439052920173,
      "loss": 0.1125,
      "num_input_tokens_seen": 32942048,
      "step": 38045
    },
    {
      "epoch": 17.93965110796794,
      "grad_norm": 0.008439374156296253,
      "learning_rate": 0.0017575416936836286,
      "loss": 0.0561,
      "num_input_tokens_seen": 32946272,
      "step": 38050
    },
    {
      "epoch": 17.94200848656294,
      "grad_norm": 0.02078508585691452,
      "learning_rate": 0.0017485623429252528,
      "loss": 0.1623,
      "num_input_tokens_seen": 32950704,
      "step": 38055
    },
    {
      "epoch": 17.944365865157945,
      "grad_norm": 0.022240743041038513,
      "learning_rate": 0.0017396058544016156,
      "loss": 0.1168,
      "num_input_tokens_seen": 32955328,
      "step": 38060
    },
    {
      "epoch": 17.946723243752945,
      "grad_norm": 0.024186881259083748,
      "learning_rate": 0.0017306722294938958,
      "loss": 0.1279,
      "num_input_tokens_seen": 32958928,
      "step": 38065
    },
    {
      "epoch": 17.94908062234795,
      "grad_norm": 0.021847285330295563,
      "learning_rate": 0.0017217614695798078,
      "loss": 0.0798,
      "num_input_tokens_seen": 32963472,
      "step": 38070
    },
    {
      "epoch": 17.95143800094295,
      "grad_norm": 0.007008741144090891,
      "learning_rate": 0.001712873576033469,
      "loss": 0.054,
      "num_input_tokens_seen": 32968128,
      "step": 38075
    },
    {
      "epoch": 17.953795379537954,
      "grad_norm": 0.0070619480684399605,
      "learning_rate": 0.0017040085502255163,
      "loss": 0.1337,
      "num_input_tokens_seen": 32973120,
      "step": 38080
    },
    {
      "epoch": 17.956152758132955,
      "grad_norm": 0.007161016575992107,
      "learning_rate": 0.0016951663935230565,
      "loss": 0.1014,
      "num_input_tokens_seen": 32978080,
      "step": 38085
    },
    {
      "epoch": 17.95851013672796,
      "grad_norm": 0.020568722859025,
      "learning_rate": 0.0016863471072896485,
      "loss": 0.0657,
      "num_input_tokens_seen": 32982240,
      "step": 38090
    },
    {
      "epoch": 17.96086751532296,
      "grad_norm": 0.002584498142823577,
      "learning_rate": 0.0016775506928853377,
      "loss": 0.1138,
      "num_input_tokens_seen": 32986400,
      "step": 38095
    },
    {
      "epoch": 17.963224893917964,
      "grad_norm": 0.003965984098613262,
      "learning_rate": 0.001668777151666656,
      "loss": 0.1601,
      "num_input_tokens_seen": 32990608,
      "step": 38100
    },
    {
      "epoch": 17.965582272512965,
      "grad_norm": 0.01800120621919632,
      "learning_rate": 0.0016600264849865709,
      "loss": 0.1195,
      "num_input_tokens_seen": 32994368,
      "step": 38105
    },
    {
      "epoch": 17.96793965110797,
      "grad_norm": 0.007404701318591833,
      "learning_rate": 0.0016512986941945695,
      "loss": 0.0976,
      "num_input_tokens_seen": 32998592,
      "step": 38110
    },
    {
      "epoch": 17.97029702970297,
      "grad_norm": 0.013168561272323132,
      "learning_rate": 0.0016425937806365753,
      "loss": 0.0402,
      "num_input_tokens_seen": 33003184,
      "step": 38115
    },
    {
      "epoch": 17.972654408297974,
      "grad_norm": 0.01746150106191635,
      "learning_rate": 0.0016339117456549979,
      "loss": 0.1067,
      "num_input_tokens_seen": 33006816,
      "step": 38120
    },
    {
      "epoch": 17.975011786892974,
      "grad_norm": 0.006270723883062601,
      "learning_rate": 0.0016252525905886995,
      "loss": 0.1301,
      "num_input_tokens_seen": 33011440,
      "step": 38125
    },
    {
      "epoch": 17.97736916548798,
      "grad_norm": 0.009296282194554806,
      "learning_rate": 0.0016166163167730617,
      "loss": 0.1074,
      "num_input_tokens_seen": 33015744,
      "step": 38130
    },
    {
      "epoch": 17.97972654408298,
      "grad_norm": 0.004519122187048197,
      "learning_rate": 0.0016080029255398864,
      "loss": 0.0748,
      "num_input_tokens_seen": 33019536,
      "step": 38135
    },
    {
      "epoch": 17.982083922677983,
      "grad_norm": 0.008839458227157593,
      "learning_rate": 0.0015994124182174606,
      "loss": 0.0579,
      "num_input_tokens_seen": 33024384,
      "step": 38140
    },
    {
      "epoch": 17.984441301272984,
      "grad_norm": 0.008435183204710484,
      "learning_rate": 0.001590844796130575,
      "loss": 0.0952,
      "num_input_tokens_seen": 33028816,
      "step": 38145
    },
    {
      "epoch": 17.986798679867988,
      "grad_norm": 0.012483861297369003,
      "learning_rate": 0.001582300060600439,
      "loss": 0.1705,
      "num_input_tokens_seen": 33033296,
      "step": 38150
    },
    {
      "epoch": 17.98915605846299,
      "grad_norm": 0.026790892705321312,
      "learning_rate": 0.0015737782129447652,
      "loss": 0.1218,
      "num_input_tokens_seen": 33037552,
      "step": 38155
    },
    {
      "epoch": 17.991513437057993,
      "grad_norm": 0.0038666073232889175,
      "learning_rate": 0.0015652792544777361,
      "loss": 0.095,
      "num_input_tokens_seen": 33041776,
      "step": 38160
    },
    {
      "epoch": 17.993870815652993,
      "grad_norm": 0.017153222113847733,
      "learning_rate": 0.0015568031865099863,
      "loss": 0.1281,
      "num_input_tokens_seen": 33046288,
      "step": 38165
    },
    {
      "epoch": 17.996228194247998,
      "grad_norm": 0.005000031553208828,
      "learning_rate": 0.0015483500103486369,
      "loss": 0.184,
      "num_input_tokens_seen": 33050224,
      "step": 38170
    },
    {
      "epoch": 17.998585572842998,
      "grad_norm": 0.029735101386904716,
      "learning_rate": 0.0015399197272972787,
      "loss": 0.1663,
      "num_input_tokens_seen": 33054960,
      "step": 38175
    },
    {
      "epoch": 18.000942951438002,
      "grad_norm": 0.01599138416349888,
      "learning_rate": 0.0015315123386559714,
      "loss": 0.1346,
      "num_input_tokens_seen": 33059456,
      "step": 38180
    },
    {
      "epoch": 18.003300330033003,
      "grad_norm": 0.010431255213916302,
      "learning_rate": 0.0015231278457212283,
      "loss": 0.093,
      "num_input_tokens_seen": 33065056,
      "step": 38185
    },
    {
      "epoch": 18.005657708628007,
      "grad_norm": 0.024192914366722107,
      "learning_rate": 0.001514766249786048,
      "loss": 0.1063,
      "num_input_tokens_seen": 33068992,
      "step": 38190
    },
    {
      "epoch": 18.008015087223008,
      "grad_norm": 0.007937935180962086,
      "learning_rate": 0.0015064275521398994,
      "loss": 0.0773,
      "num_input_tokens_seen": 33072944,
      "step": 38195
    },
    {
      "epoch": 18.010372465818012,
      "grad_norm": 0.007735913153737783,
      "learning_rate": 0.0014981117540686872,
      "loss": 0.1239,
      "num_input_tokens_seen": 33076912,
      "step": 38200
    },
    {
      "epoch": 18.010372465818012,
      "eval_loss": 0.2858264744281769,
      "eval_runtime": 21.9032,
      "eval_samples_per_second": 43.053,
      "eval_steps_per_second": 21.549,
      "num_input_tokens_seen": 33076912,
      "step": 38200
    },
    {
      "epoch": 18.012729844413013,
      "grad_norm": 0.022645853459835052,
      "learning_rate": 0.0014898188568548687,
      "loss": 0.0661,
      "num_input_tokens_seen": 33081536,
      "step": 38205
    },
    {
      "epoch": 18.015087223008017,
      "grad_norm": 0.030319707468152046,
      "learning_rate": 0.0014815488617772542,
      "loss": 0.0843,
      "num_input_tokens_seen": 33085600,
      "step": 38210
    },
    {
      "epoch": 18.017444601603017,
      "grad_norm": 0.024839386343955994,
      "learning_rate": 0.0014733017701112072,
      "loss": 0.1385,
      "num_input_tokens_seen": 33089824,
      "step": 38215
    },
    {
      "epoch": 18.019801980198018,
      "grad_norm": 0.011745426803827286,
      "learning_rate": 0.0014650775831285435,
      "loss": 0.0825,
      "num_input_tokens_seen": 33093936,
      "step": 38220
    },
    {
      "epoch": 18.022159358793022,
      "grad_norm": 0.010720506310462952,
      "learning_rate": 0.001456876302097515,
      "loss": 0.172,
      "num_input_tokens_seen": 33098208,
      "step": 38225
    },
    {
      "epoch": 18.024516737388023,
      "grad_norm": 0.005856699775904417,
      "learning_rate": 0.0014486979282828604,
      "loss": 0.0557,
      "num_input_tokens_seen": 33102528,
      "step": 38230
    },
    {
      "epoch": 18.026874115983027,
      "grad_norm": 0.003233224619179964,
      "learning_rate": 0.001440542462945804,
      "loss": 0.0857,
      "num_input_tokens_seen": 33106416,
      "step": 38235
    },
    {
      "epoch": 18.029231494578028,
      "grad_norm": 0.013712835498154163,
      "learning_rate": 0.0014324099073440232,
      "loss": 0.1145,
      "num_input_tokens_seen": 33110800,
      "step": 38240
    },
    {
      "epoch": 18.031588873173032,
      "grad_norm": 0.009180348366498947,
      "learning_rate": 0.0014243002627316482,
      "loss": 0.1125,
      "num_input_tokens_seen": 33115136,
      "step": 38245
    },
    {
      "epoch": 18.033946251768032,
      "grad_norm": 0.0068691992200911045,
      "learning_rate": 0.0014162135303592781,
      "loss": 0.0739,
      "num_input_tokens_seen": 33119488,
      "step": 38250
    },
    {
      "epoch": 18.036303630363037,
      "grad_norm": 0.010211888700723648,
      "learning_rate": 0.001408149711474016,
      "loss": 0.052,
      "num_input_tokens_seen": 33123568,
      "step": 38255
    },
    {
      "epoch": 18.038661008958037,
      "grad_norm": 0.0283662062138319,
      "learning_rate": 0.0014001088073193834,
      "loss": 0.1295,
      "num_input_tokens_seen": 33127952,
      "step": 38260
    },
    {
      "epoch": 18.04101838755304,
      "grad_norm": 0.015019943937659264,
      "learning_rate": 0.0013920908191354052,
      "loss": 0.0768,
      "num_input_tokens_seen": 33132976,
      "step": 38265
    },
    {
      "epoch": 18.043375766148042,
      "grad_norm": 0.002399677410721779,
      "learning_rate": 0.001384095748158526,
      "loss": 0.1325,
      "num_input_tokens_seen": 33137184,
      "step": 38270
    },
    {
      "epoch": 18.045733144743046,
      "grad_norm": 0.004782942123711109,
      "learning_rate": 0.0013761235956217255,
      "loss": 0.0619,
      "num_input_tokens_seen": 33141200,
      "step": 38275
    },
    {
      "epoch": 18.048090523338047,
      "grad_norm": 0.017798256129026413,
      "learning_rate": 0.0013681743627543873,
      "loss": 0.1282,
      "num_input_tokens_seen": 33146288,
      "step": 38280
    },
    {
      "epoch": 18.05044790193305,
      "grad_norm": 0.010104597546160221,
      "learning_rate": 0.001360248050782381,
      "loss": 0.077,
      "num_input_tokens_seen": 33150992,
      "step": 38285
    },
    {
      "epoch": 18.05280528052805,
      "grad_norm": 0.017527736723423004,
      "learning_rate": 0.001352344660928062,
      "loss": 0.1082,
      "num_input_tokens_seen": 33154768,
      "step": 38290
    },
    {
      "epoch": 18.055162659123056,
      "grad_norm": 0.011829888448119164,
      "learning_rate": 0.0013444641944102052,
      "loss": 0.064,
      "num_input_tokens_seen": 33158480,
      "step": 38295
    },
    {
      "epoch": 18.057520037718056,
      "grad_norm": 0.00579990865662694,
      "learning_rate": 0.0013366066524441056,
      "loss": 0.0422,
      "num_input_tokens_seen": 33162528,
      "step": 38300
    },
    {
      "epoch": 18.05987741631306,
      "grad_norm": 0.006139148026704788,
      "learning_rate": 0.0013287720362414768,
      "loss": 0.0921,
      "num_input_tokens_seen": 33166736,
      "step": 38305
    },
    {
      "epoch": 18.06223479490806,
      "grad_norm": 0.012056229636073112,
      "learning_rate": 0.0013209603470105025,
      "loss": 0.1282,
      "num_input_tokens_seen": 33170624,
      "step": 38310
    },
    {
      "epoch": 18.064592173503065,
      "grad_norm": 0.03021955117583275,
      "learning_rate": 0.0013131715859558857,
      "loss": 0.1081,
      "num_input_tokens_seen": 33176048,
      "step": 38315
    },
    {
      "epoch": 18.066949552098066,
      "grad_norm": 0.009102693758904934,
      "learning_rate": 0.001305405754278699,
      "loss": 0.0979,
      "num_input_tokens_seen": 33180768,
      "step": 38320
    },
    {
      "epoch": 18.06930693069307,
      "grad_norm": 0.003484122222289443,
      "learning_rate": 0.0012976628531765843,
      "loss": 0.0885,
      "num_input_tokens_seen": 33185248,
      "step": 38325
    },
    {
      "epoch": 18.07166430928807,
      "grad_norm": 0.004599242936819792,
      "learning_rate": 0.0012899428838435533,
      "loss": 0.059,
      "num_input_tokens_seen": 33189696,
      "step": 38330
    },
    {
      "epoch": 18.074021687883075,
      "grad_norm": 0.019366301596164703,
      "learning_rate": 0.001282245847470137,
      "loss": 0.0698,
      "num_input_tokens_seen": 33193312,
      "step": 38335
    },
    {
      "epoch": 18.076379066478076,
      "grad_norm": 0.013607708737254143,
      "learning_rate": 0.001274571745243319,
      "loss": 0.0964,
      "num_input_tokens_seen": 33198416,
      "step": 38340
    },
    {
      "epoch": 18.07873644507308,
      "grad_norm": 0.006907926872372627,
      "learning_rate": 0.0012669205783465364,
      "loss": 0.0819,
      "num_input_tokens_seen": 33202912,
      "step": 38345
    },
    {
      "epoch": 18.08109382366808,
      "grad_norm": 0.015390957705676556,
      "learning_rate": 0.001259292347959695,
      "loss": 0.0895,
      "num_input_tokens_seen": 33206768,
      "step": 38350
    },
    {
      "epoch": 18.083451202263085,
      "grad_norm": 0.014342605136334896,
      "learning_rate": 0.0012516870552591707,
      "loss": 0.0898,
      "num_input_tokens_seen": 33210896,
      "step": 38355
    },
    {
      "epoch": 18.085808580858085,
      "grad_norm": 0.005119668319821358,
      "learning_rate": 0.001244104701417792,
      "loss": 0.1223,
      "num_input_tokens_seen": 33215168,
      "step": 38360
    },
    {
      "epoch": 18.08816595945309,
      "grad_norm": 0.014188064262270927,
      "learning_rate": 0.0012365452876048565,
      "loss": 0.0436,
      "num_input_tokens_seen": 33218752,
      "step": 38365
    },
    {
      "epoch": 18.09052333804809,
      "grad_norm": 0.005307248793542385,
      "learning_rate": 0.001229008814986099,
      "loss": 0.0626,
      "num_input_tokens_seen": 33223248,
      "step": 38370
    },
    {
      "epoch": 18.092880716643094,
      "grad_norm": 0.014490621164441109,
      "learning_rate": 0.0012214952847237725,
      "loss": 0.092,
      "num_input_tokens_seen": 33227696,
      "step": 38375
    },
    {
      "epoch": 18.095238095238095,
      "grad_norm": 0.01472401525825262,
      "learning_rate": 0.0012140046979765339,
      "loss": 0.0652,
      "num_input_tokens_seen": 33231392,
      "step": 38380
    },
    {
      "epoch": 18.0975954738331,
      "grad_norm": 0.012282131239771843,
      "learning_rate": 0.0012065370558995258,
      "loss": 0.0879,
      "num_input_tokens_seen": 33235472,
      "step": 38385
    },
    {
      "epoch": 18.0999528524281,
      "grad_norm": 0.012077976949512959,
      "learning_rate": 0.0011990923596443602,
      "loss": 0.0714,
      "num_input_tokens_seen": 33239872,
      "step": 38390
    },
    {
      "epoch": 18.102310231023104,
      "grad_norm": 0.01181839220225811,
      "learning_rate": 0.001191670610359119,
      "loss": 0.0616,
      "num_input_tokens_seen": 33244768,
      "step": 38395
    },
    {
      "epoch": 18.104667609618105,
      "grad_norm": 0.016474559903144836,
      "learning_rate": 0.0011842718091882865,
      "loss": 0.0903,
      "num_input_tokens_seen": 33248832,
      "step": 38400
    },
    {
      "epoch": 18.104667609618105,
      "eval_loss": 0.28742265701293945,
      "eval_runtime": 21.9177,
      "eval_samples_per_second": 43.025,
      "eval_steps_per_second": 21.535,
      "num_input_tokens_seen": 33248832,
      "step": 38400
    },
    {
      "epoch": 18.10702498821311,
      "grad_norm": 0.019123561680316925,
      "learning_rate": 0.0011768959572729,
      "loss": 0.0999,
      "num_input_tokens_seen": 33253296,
      "step": 38405
    },
    {
      "epoch": 18.10938236680811,
      "grad_norm": 0.010630922392010689,
      "learning_rate": 0.001169543055750366,
      "loss": 0.1409,
      "num_input_tokens_seen": 33257136,
      "step": 38410
    },
    {
      "epoch": 18.111739745403113,
      "grad_norm": 0.01099251490086317,
      "learning_rate": 0.0011622131057546115,
      "loss": 0.0981,
      "num_input_tokens_seen": 33261504,
      "step": 38415
    },
    {
      "epoch": 18.114097123998114,
      "grad_norm": 0.015359519049525261,
      "learning_rate": 0.0011549061084160316,
      "loss": 0.1045,
      "num_input_tokens_seen": 33265600,
      "step": 38420
    },
    {
      "epoch": 18.116454502593115,
      "grad_norm": 0.030503666028380394,
      "learning_rate": 0.0011476220648614088,
      "loss": 0.1859,
      "num_input_tokens_seen": 33269104,
      "step": 38425
    },
    {
      "epoch": 18.11881188118812,
      "grad_norm": 0.014417004771530628,
      "learning_rate": 0.0011403609762140777,
      "loss": 0.1513,
      "num_input_tokens_seen": 33273632,
      "step": 38430
    },
    {
      "epoch": 18.12116925978312,
      "grad_norm": 0.018436715006828308,
      "learning_rate": 0.0011331228435937756,
      "loss": 0.1662,
      "num_input_tokens_seen": 33278704,
      "step": 38435
    },
    {
      "epoch": 18.123526638378124,
      "grad_norm": 0.008053172379732132,
      "learning_rate": 0.0011259076681166935,
      "loss": 0.0726,
      "num_input_tokens_seen": 33282768,
      "step": 38440
    },
    {
      "epoch": 18.125884016973124,
      "grad_norm": 0.012580829672515392,
      "learning_rate": 0.0011187154508955244,
      "loss": 0.0909,
      "num_input_tokens_seen": 33286416,
      "step": 38445
    },
    {
      "epoch": 18.12824139556813,
      "grad_norm": 0.010426346212625504,
      "learning_rate": 0.001111546193039381,
      "loss": 0.0684,
      "num_input_tokens_seen": 33290400,
      "step": 38450
    },
    {
      "epoch": 18.13059877416313,
      "grad_norm": 0.009272490628063679,
      "learning_rate": 0.0011043998956538792,
      "loss": 0.079,
      "num_input_tokens_seen": 33295120,
      "step": 38455
    },
    {
      "epoch": 18.132956152758133,
      "grad_norm": 0.018105091527104378,
      "learning_rate": 0.0010972765598410538,
      "loss": 0.0702,
      "num_input_tokens_seen": 33299280,
      "step": 38460
    },
    {
      "epoch": 18.135313531353134,
      "grad_norm": 0.01514426525682211,
      "learning_rate": 0.0010901761866993931,
      "loss": 0.1328,
      "num_input_tokens_seen": 33303856,
      "step": 38465
    },
    {
      "epoch": 18.137670909948138,
      "grad_norm": 0.005651990417391062,
      "learning_rate": 0.0010830987773238876,
      "loss": 0.064,
      "num_input_tokens_seen": 33308048,
      "step": 38470
    },
    {
      "epoch": 18.14002828854314,
      "grad_norm": 0.010926861315965652,
      "learning_rate": 0.0010760443328059644,
      "loss": 0.0569,
      "num_input_tokens_seen": 33312288,
      "step": 38475
    },
    {
      "epoch": 18.142385667138143,
      "grad_norm": 0.008650146424770355,
      "learning_rate": 0.001069012854233503,
      "loss": 0.0549,
      "num_input_tokens_seen": 33316864,
      "step": 38480
    },
    {
      "epoch": 18.144743045733144,
      "grad_norm": 0.0202487763017416,
      "learning_rate": 0.0010620043426908365,
      "loss": 0.1778,
      "num_input_tokens_seen": 33321184,
      "step": 38485
    },
    {
      "epoch": 18.147100424328148,
      "grad_norm": 0.016532903537154198,
      "learning_rate": 0.0010550187992587833,
      "loss": 0.1384,
      "num_input_tokens_seen": 33325168,
      "step": 38490
    },
    {
      "epoch": 18.14945780292315,
      "grad_norm": 0.01289845909923315,
      "learning_rate": 0.0010480562250145653,
      "loss": 0.0975,
      "num_input_tokens_seen": 33329104,
      "step": 38495
    },
    {
      "epoch": 18.151815181518153,
      "grad_norm": 0.016277363523840904,
      "learning_rate": 0.0010411166210319567,
      "loss": 0.1157,
      "num_input_tokens_seen": 33333424,
      "step": 38500
    },
    {
      "epoch": 18.154172560113153,
      "grad_norm": 0.0018365181749686599,
      "learning_rate": 0.0010341999883810848,
      "loss": 0.0574,
      "num_input_tokens_seen": 33337008,
      "step": 38505
    },
    {
      "epoch": 18.156529938708157,
      "grad_norm": 0.030944960191845894,
      "learning_rate": 0.0010273063281285965,
      "loss": 0.1482,
      "num_input_tokens_seen": 33341840,
      "step": 38510
    },
    {
      "epoch": 18.158887317303158,
      "grad_norm": 0.024335481226444244,
      "learning_rate": 0.0010204356413375747,
      "loss": 0.1128,
      "num_input_tokens_seen": 33346752,
      "step": 38515
    },
    {
      "epoch": 18.161244695898162,
      "grad_norm": 0.01192509476095438,
      "learning_rate": 0.001013587929067572,
      "loss": 0.1025,
      "num_input_tokens_seen": 33350400,
      "step": 38520
    },
    {
      "epoch": 18.163602074493163,
      "grad_norm": 0.00975954532623291,
      "learning_rate": 0.00100676319237461,
      "loss": 0.1119,
      "num_input_tokens_seen": 33354624,
      "step": 38525
    },
    {
      "epoch": 18.165959453088167,
      "grad_norm": 0.0026107397861778736,
      "learning_rate": 0.0009999614323110972,
      "loss": 0.1259,
      "num_input_tokens_seen": 33358816,
      "step": 38530
    },
    {
      "epoch": 18.168316831683168,
      "grad_norm": 0.00970445852726698,
      "learning_rate": 0.000993182649926011,
      "loss": 0.0637,
      "num_input_tokens_seen": 33363568,
      "step": 38535
    },
    {
      "epoch": 18.17067421027817,
      "grad_norm": 0.01386620756238699,
      "learning_rate": 0.000986426846264682,
      "loss": 0.0497,
      "num_input_tokens_seen": 33367920,
      "step": 38540
    },
    {
      "epoch": 18.173031588873172,
      "grad_norm": 0.015404759906232357,
      "learning_rate": 0.00097969402236896,
      "loss": 0.1391,
      "num_input_tokens_seen": 33372208,
      "step": 38545
    },
    {
      "epoch": 18.175388967468177,
      "grad_norm": 0.011873593553900719,
      "learning_rate": 0.0009729841792771143,
      "loss": 0.158,
      "num_input_tokens_seen": 33377040,
      "step": 38550
    },
    {
      "epoch": 18.177746346063177,
      "grad_norm": 0.008375647477805614,
      "learning_rate": 0.0009662973180239176,
      "loss": 0.0997,
      "num_input_tokens_seen": 33380976,
      "step": 38555
    },
    {
      "epoch": 18.18010372465818,
      "grad_norm": 0.012741712853312492,
      "learning_rate": 0.0009596334396405448,
      "loss": 0.1198,
      "num_input_tokens_seen": 33385504,
      "step": 38560
    },
    {
      "epoch": 18.182461103253182,
      "grad_norm": 0.0021736614871770144,
      "learning_rate": 0.0009529925451546406,
      "loss": 0.0653,
      "num_input_tokens_seen": 33389408,
      "step": 38565
    },
    {
      "epoch": 18.184818481848186,
      "grad_norm": 0.00871674157679081,
      "learning_rate": 0.0009463746355903357,
      "loss": 0.047,
      "num_input_tokens_seen": 33393344,
      "step": 38570
    },
    {
      "epoch": 18.187175860443187,
      "grad_norm": 0.00917226355522871,
      "learning_rate": 0.0009397797119681971,
      "loss": 0.1292,
      "num_input_tokens_seen": 33397856,
      "step": 38575
    },
    {
      "epoch": 18.18953323903819,
      "grad_norm": 0.014530804008245468,
      "learning_rate": 0.0009332077753052281,
      "loss": 0.0802,
      "num_input_tokens_seen": 33402176,
      "step": 38580
    },
    {
      "epoch": 18.19189061763319,
      "grad_norm": 0.013889393769204617,
      "learning_rate": 0.0009266588266149011,
      "loss": 0.1366,
      "num_input_tokens_seen": 33406656,
      "step": 38585
    },
    {
      "epoch": 18.194247996228196,
      "grad_norm": 0.006645447574555874,
      "learning_rate": 0.0009201328669071584,
      "loss": 0.102,
      "num_input_tokens_seen": 33411088,
      "step": 38590
    },
    {
      "epoch": 18.196605374823196,
      "grad_norm": 0.007802989333868027,
      "learning_rate": 0.0009136298971883949,
      "loss": 0.0591,
      "num_input_tokens_seen": 33416176,
      "step": 38595
    },
    {
      "epoch": 18.1989627534182,
      "grad_norm": 0.0016977130435407162,
      "learning_rate": 0.0009071499184614251,
      "loss": 0.0294,
      "num_input_tokens_seen": 33420800,
      "step": 38600
    },
    {
      "epoch": 18.1989627534182,
      "eval_loss": 0.28807637095451355,
      "eval_runtime": 21.919,
      "eval_samples_per_second": 43.022,
      "eval_steps_per_second": 21.534,
      "num_input_tokens_seen": 33420800,
      "step": 38600
    },
    {
      "epoch": 18.2013201320132,
      "grad_norm": 0.002798629691824317,
      "learning_rate": 0.0009006929317255663,
      "loss": 0.0974,
      "num_input_tokens_seen": 33424832,
      "step": 38605
    },
    {
      "epoch": 18.203677510608205,
      "grad_norm": 0.014146587811410427,
      "learning_rate": 0.0008942589379765387,
      "loss": 0.0799,
      "num_input_tokens_seen": 33428672,
      "step": 38610
    },
    {
      "epoch": 18.206034889203206,
      "grad_norm": 0.01277405396103859,
      "learning_rate": 0.0008878479382065817,
      "loss": 0.115,
      "num_input_tokens_seen": 33432528,
      "step": 38615
    },
    {
      "epoch": 18.208392267798207,
      "grad_norm": 0.017665691673755646,
      "learning_rate": 0.0008814599334043215,
      "loss": 0.1295,
      "num_input_tokens_seen": 33436816,
      "step": 38620
    },
    {
      "epoch": 18.21074964639321,
      "grad_norm": 0.01662038080394268,
      "learning_rate": 0.0008750949245548866,
      "loss": 0.0778,
      "num_input_tokens_seen": 33441536,
      "step": 38625
    },
    {
      "epoch": 18.21310702498821,
      "grad_norm": 0.0032254511024802923,
      "learning_rate": 0.0008687529126398252,
      "loss": 0.0592,
      "num_input_tokens_seen": 33445664,
      "step": 38630
    },
    {
      "epoch": 18.215464403583216,
      "grad_norm": 0.00875080842524767,
      "learning_rate": 0.0008624338986371715,
      "loss": 0.0471,
      "num_input_tokens_seen": 33450240,
      "step": 38635
    },
    {
      "epoch": 18.217821782178216,
      "grad_norm": 0.019642548635601997,
      "learning_rate": 0.0008561378835213962,
      "loss": 0.0872,
      "num_input_tokens_seen": 33454448,
      "step": 38640
    },
    {
      "epoch": 18.22017916077322,
      "grad_norm": 0.009133032523095608,
      "learning_rate": 0.0008498648682634058,
      "loss": 0.0613,
      "num_input_tokens_seen": 33458720,
      "step": 38645
    },
    {
      "epoch": 18.22253653936822,
      "grad_norm": 0.01762106828391552,
      "learning_rate": 0.0008436148538306099,
      "loss": 0.1159,
      "num_input_tokens_seen": 33462880,
      "step": 38650
    },
    {
      "epoch": 18.224893917963225,
      "grad_norm": 0.005270893685519695,
      "learning_rate": 0.0008373878411868041,
      "loss": 0.0545,
      "num_input_tokens_seen": 33466944,
      "step": 38655
    },
    {
      "epoch": 18.227251296558226,
      "grad_norm": 0.012603041715919971,
      "learning_rate": 0.000831183831292287,
      "loss": 0.1236,
      "num_input_tokens_seen": 33470752,
      "step": 38660
    },
    {
      "epoch": 18.22960867515323,
      "grad_norm": 0.024409467354416847,
      "learning_rate": 0.0008250028251037933,
      "loss": 0.1279,
      "num_input_tokens_seen": 33475504,
      "step": 38665
    },
    {
      "epoch": 18.23196605374823,
      "grad_norm": 0.0029145460575819016,
      "learning_rate": 0.0008188448235745271,
      "loss": 0.0601,
      "num_input_tokens_seen": 33480128,
      "step": 38670
    },
    {
      "epoch": 18.234323432343235,
      "grad_norm": 0.007935485802590847,
      "learning_rate": 0.0008127098276541122,
      "loss": 0.0464,
      "num_input_tokens_seen": 33484144,
      "step": 38675
    },
    {
      "epoch": 18.236680810938235,
      "grad_norm": 0.015196803025901318,
      "learning_rate": 0.0008065978382886418,
      "loss": 0.117,
      "num_input_tokens_seen": 33488608,
      "step": 38680
    },
    {
      "epoch": 18.23903818953324,
      "grad_norm": 0.018609551712870598,
      "learning_rate": 0.0008005088564206785,
      "loss": 0.0589,
      "num_input_tokens_seen": 33492864,
      "step": 38685
    },
    {
      "epoch": 18.24139556812824,
      "grad_norm": 0.002322407905012369,
      "learning_rate": 0.0007944428829891881,
      "loss": 0.048,
      "num_input_tokens_seen": 33496512,
      "step": 38690
    },
    {
      "epoch": 18.243752946723244,
      "grad_norm": 0.0090817641466856,
      "learning_rate": 0.0007883999189296386,
      "loss": 0.08,
      "num_input_tokens_seen": 33502368,
      "step": 38695
    },
    {
      "epoch": 18.246110325318245,
      "grad_norm": 0.009893090464174747,
      "learning_rate": 0.0007823799651739515,
      "loss": 0.124,
      "num_input_tokens_seen": 33506272,
      "step": 38700
    },
    {
      "epoch": 18.24846770391325,
      "grad_norm": 0.004182750359177589,
      "learning_rate": 0.0007763830226504509,
      "loss": 0.0946,
      "num_input_tokens_seen": 33511584,
      "step": 38705
    },
    {
      "epoch": 18.25082508250825,
      "grad_norm": 0.004378094337880611,
      "learning_rate": 0.0007704090922839468,
      "loss": 0.0733,
      "num_input_tokens_seen": 33516128,
      "step": 38710
    },
    {
      "epoch": 18.253182461103254,
      "grad_norm": 0.012883923947811127,
      "learning_rate": 0.0007644581749957025,
      "loss": 0.1308,
      "num_input_tokens_seen": 33520928,
      "step": 38715
    },
    {
      "epoch": 18.255539839698255,
      "grad_norm": 0.012955124489963055,
      "learning_rate": 0.000758530271703417,
      "loss": 0.0949,
      "num_input_tokens_seen": 33527792,
      "step": 38720
    },
    {
      "epoch": 18.25789721829326,
      "grad_norm": 0.010785473510622978,
      "learning_rate": 0.0007526253833212426,
      "loss": 0.1339,
      "num_input_tokens_seen": 33531728,
      "step": 38725
    },
    {
      "epoch": 18.26025459688826,
      "grad_norm": 0.0064979540184140205,
      "learning_rate": 0.0007467435107598008,
      "loss": 0.0825,
      "num_input_tokens_seen": 33535664,
      "step": 38730
    },
    {
      "epoch": 18.262611975483264,
      "grad_norm": 0.02031964249908924,
      "learning_rate": 0.0007408846549261328,
      "loss": 0.0981,
      "num_input_tokens_seen": 33539664,
      "step": 38735
    },
    {
      "epoch": 18.264969354078264,
      "grad_norm": 0.012543831020593643,
      "learning_rate": 0.0007350488167237656,
      "loss": 0.0655,
      "num_input_tokens_seen": 33543664,
      "step": 38740
    },
    {
      "epoch": 18.26732673267327,
      "grad_norm": 0.012430505827069283,
      "learning_rate": 0.0007292359970526629,
      "loss": 0.1133,
      "num_input_tokens_seen": 33547456,
      "step": 38745
    },
    {
      "epoch": 18.26968411126827,
      "grad_norm": 0.017103560268878937,
      "learning_rate": 0.0007234461968092076,
      "loss": 0.0955,
      "num_input_tokens_seen": 33551024,
      "step": 38750
    },
    {
      "epoch": 18.272041489863273,
      "grad_norm": 0.011629187501966953,
      "learning_rate": 0.0007176794168862854,
      "loss": 0.0532,
      "num_input_tokens_seen": 33555792,
      "step": 38755
    },
    {
      "epoch": 18.274398868458274,
      "grad_norm": 0.007442745845764875,
      "learning_rate": 0.000711935658173185,
      "loss": 0.0611,
      "num_input_tokens_seen": 33559872,
      "step": 38760
    },
    {
      "epoch": 18.276756247053278,
      "grad_norm": 0.013606538064777851,
      "learning_rate": 0.0007062149215556812,
      "loss": 0.097,
      "num_input_tokens_seen": 33564096,
      "step": 38765
    },
    {
      "epoch": 18.27911362564828,
      "grad_norm": 0.008590647019445896,
      "learning_rate": 0.0007005172079159849,
      "loss": 0.1599,
      "num_input_tokens_seen": 33567840,
      "step": 38770
    },
    {
      "epoch": 18.281471004243283,
      "grad_norm": 0.02148733288049698,
      "learning_rate": 0.0006948425181327267,
      "loss": 0.1307,
      "num_input_tokens_seen": 33573280,
      "step": 38775
    },
    {
      "epoch": 18.283828382838283,
      "grad_norm": 0.005320890340954065,
      "learning_rate": 0.000689190853081073,
      "loss": 0.0971,
      "num_input_tokens_seen": 33577424,
      "step": 38780
    },
    {
      "epoch": 18.286185761433288,
      "grad_norm": 0.023171983659267426,
      "learning_rate": 0.000683562213632527,
      "loss": 0.1271,
      "num_input_tokens_seen": 33581184,
      "step": 38785
    },
    {
      "epoch": 18.28854314002829,
      "grad_norm": 0.002641556551679969,
      "learning_rate": 0.0006779566006551108,
      "loss": 0.0439,
      "num_input_tokens_seen": 33585248,
      "step": 38790
    },
    {
      "epoch": 18.290900518623292,
      "grad_norm": 0.006319163367152214,
      "learning_rate": 0.0006723740150132995,
      "loss": 0.097,
      "num_input_tokens_seen": 33589664,
      "step": 38795
    },
    {
      "epoch": 18.293257897218293,
      "grad_norm": 0.005997419822961092,
      "learning_rate": 0.0006668144575679713,
      "loss": 0.1241,
      "num_input_tokens_seen": 33594000,
      "step": 38800
    },
    {
      "epoch": 18.293257897218293,
      "eval_loss": 0.2859748601913452,
      "eval_runtime": 21.9404,
      "eval_samples_per_second": 42.98,
      "eval_steps_per_second": 21.513,
      "num_input_tokens_seen": 33594000,
      "step": 38800
    },
    {
      "epoch": 18.295615275813297,
      "grad_norm": 0.002386396750807762,
      "learning_rate": 0.0006612779291765069,
      "loss": 0.1636,
      "num_input_tokens_seen": 33598000,
      "step": 38805
    },
    {
      "epoch": 18.297972654408298,
      "grad_norm": 0.014595330692827702,
      "learning_rate": 0.0006557644306926736,
      "loss": 0.0827,
      "num_input_tokens_seen": 33602512,
      "step": 38810
    },
    {
      "epoch": 18.300330033003302,
      "grad_norm": 0.013365499675273895,
      "learning_rate": 0.0006502739629667575,
      "loss": 0.0478,
      "num_input_tokens_seen": 33607040,
      "step": 38815
    },
    {
      "epoch": 18.302687411598303,
      "grad_norm": 0.017048876732587814,
      "learning_rate": 0.0006448065268454317,
      "loss": 0.1279,
      "num_input_tokens_seen": 33611728,
      "step": 38820
    },
    {
      "epoch": 18.305044790193303,
      "grad_norm": 0.008202873170375824,
      "learning_rate": 0.0006393621231718549,
      "loss": 0.0375,
      "num_input_tokens_seen": 33615488,
      "step": 38825
    },
    {
      "epoch": 18.307402168788308,
      "grad_norm": 0.004397777374833822,
      "learning_rate": 0.0006339407527856389,
      "loss": 0.0996,
      "num_input_tokens_seen": 33619968,
      "step": 38830
    },
    {
      "epoch": 18.309759547383308,
      "grad_norm": 0.017157312482595444,
      "learning_rate": 0.0006285424165227982,
      "loss": 0.0706,
      "num_input_tokens_seen": 33623424,
      "step": 38835
    },
    {
      "epoch": 18.312116925978312,
      "grad_norm": 0.008174706250429153,
      "learning_rate": 0.0006231671152158169,
      "loss": 0.0586,
      "num_input_tokens_seen": 33627392,
      "step": 38840
    },
    {
      "epoch": 18.314474304573313,
      "grad_norm": 0.010457228869199753,
      "learning_rate": 0.0006178148496936819,
      "loss": 0.0758,
      "num_input_tokens_seen": 33631392,
      "step": 38845
    },
    {
      "epoch": 18.316831683168317,
      "grad_norm": 0.01212691143155098,
      "learning_rate": 0.000612485620781733,
      "loss": 0.0619,
      "num_input_tokens_seen": 33636064,
      "step": 38850
    },
    {
      "epoch": 18.319189061763318,
      "grad_norm": 0.008400465361773968,
      "learning_rate": 0.0006071794293018296,
      "loss": 0.0908,
      "num_input_tokens_seen": 33639936,
      "step": 38855
    },
    {
      "epoch": 18.321546440358322,
      "grad_norm": 0.017973627895116806,
      "learning_rate": 0.0006018962760722501,
      "loss": 0.0725,
      "num_input_tokens_seen": 33644160,
      "step": 38860
    },
    {
      "epoch": 18.323903818953323,
      "grad_norm": 0.01630215346813202,
      "learning_rate": 0.0005966361619077098,
      "loss": 0.0601,
      "num_input_tokens_seen": 33647936,
      "step": 38865
    },
    {
      "epoch": 18.326261197548327,
      "grad_norm": 0.010439412668347359,
      "learning_rate": 0.000591399087619393,
      "loss": 0.0801,
      "num_input_tokens_seen": 33652208,
      "step": 38870
    },
    {
      "epoch": 18.328618576143327,
      "grad_norm": 0.008847573772072792,
      "learning_rate": 0.0005861850540149371,
      "loss": 0.0699,
      "num_input_tokens_seen": 33655792,
      "step": 38875
    },
    {
      "epoch": 18.33097595473833,
      "grad_norm": 0.023052891716361046,
      "learning_rate": 0.0005809940618983822,
      "loss": 0.0634,
      "num_input_tokens_seen": 33660176,
      "step": 38880
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.02688494697213173,
      "learning_rate": 0.0005758261120702712,
      "loss": 0.0729,
      "num_input_tokens_seen": 33664768,
      "step": 38885
    },
    {
      "epoch": 18.335690711928336,
      "grad_norm": 0.02152164839208126,
      "learning_rate": 0.0005706812053275501,
      "loss": 0.1185,
      "num_input_tokens_seen": 33668832,
      "step": 38890
    },
    {
      "epoch": 18.338048090523337,
      "grad_norm": 0.004761905875056982,
      "learning_rate": 0.0005655593424636173,
      "loss": 0.0747,
      "num_input_tokens_seen": 33673408,
      "step": 38895
    },
    {
      "epoch": 18.34040546911834,
      "grad_norm": 0.014462285675108433,
      "learning_rate": 0.0005604605242683746,
      "loss": 0.0388,
      "num_input_tokens_seen": 33678672,
      "step": 38900
    },
    {
      "epoch": 18.34276284771334,
      "grad_norm": 0.015734845772385597,
      "learning_rate": 0.0005553847515280596,
      "loss": 0.0972,
      "num_input_tokens_seen": 33683120,
      "step": 38905
    },
    {
      "epoch": 18.345120226308346,
      "grad_norm": 0.007995336316525936,
      "learning_rate": 0.0005503320250254795,
      "loss": 0.1078,
      "num_input_tokens_seen": 33687504,
      "step": 38910
    },
    {
      "epoch": 18.347477604903347,
      "grad_norm": 0.0031270708423107862,
      "learning_rate": 0.0005453023455397943,
      "loss": 0.0559,
      "num_input_tokens_seen": 33691952,
      "step": 38915
    },
    {
      "epoch": 18.34983498349835,
      "grad_norm": 0.005457921419292688,
      "learning_rate": 0.0005402957138466502,
      "loss": 0.1132,
      "num_input_tokens_seen": 33696016,
      "step": 38920
    },
    {
      "epoch": 18.35219236209335,
      "grad_norm": 0.010574528016149998,
      "learning_rate": 0.0005353121307181463,
      "loss": 0.0572,
      "num_input_tokens_seen": 33700144,
      "step": 38925
    },
    {
      "epoch": 18.354549740688356,
      "grad_norm": 0.01068148948252201,
      "learning_rate": 0.0005303515969227845,
      "loss": 0.0801,
      "num_input_tokens_seen": 33704144,
      "step": 38930
    },
    {
      "epoch": 18.356907119283356,
      "grad_norm": 0.006465582642704248,
      "learning_rate": 0.0005254141132255862,
      "loss": 0.1394,
      "num_input_tokens_seen": 33710192,
      "step": 38935
    },
    {
      "epoch": 18.35926449787836,
      "grad_norm": 0.013842878863215446,
      "learning_rate": 0.0005204996803879258,
      "loss": 0.0779,
      "num_input_tokens_seen": 33714960,
      "step": 38940
    },
    {
      "epoch": 18.36162187647336,
      "grad_norm": 0.01643979363143444,
      "learning_rate": 0.0005156082991676969,
      "loss": 0.0877,
      "num_input_tokens_seen": 33719136,
      "step": 38945
    },
    {
      "epoch": 18.363979255068365,
      "grad_norm": 0.014679396525025368,
      "learning_rate": 0.0005107399703192127,
      "loss": 0.0943,
      "num_input_tokens_seen": 33722768,
      "step": 38950
    },
    {
      "epoch": 18.366336633663366,
      "grad_norm": 0.009714308194816113,
      "learning_rate": 0.0005058946945932063,
      "loss": 0.0871,
      "num_input_tokens_seen": 33727312,
      "step": 38955
    },
    {
      "epoch": 18.36869401225837,
      "grad_norm": 0.011285386979579926,
      "learning_rate": 0.0005010724727369131,
      "loss": 0.0735,
      "num_input_tokens_seen": 33731296,
      "step": 38960
    },
    {
      "epoch": 18.37105139085337,
      "grad_norm": 0.010397360660135746,
      "learning_rate": 0.000496273305493955,
      "loss": 0.0726,
      "num_input_tokens_seen": 33735200,
      "step": 38965
    },
    {
      "epoch": 18.373408769448375,
      "grad_norm": 0.005112814251333475,
      "learning_rate": 0.0004914971936044399,
      "loss": 0.1086,
      "num_input_tokens_seen": 33740064,
      "step": 38970
    },
    {
      "epoch": 18.375766148043375,
      "grad_norm": 0.004232257138937712,
      "learning_rate": 0.00048674413780491196,
      "loss": 0.0759,
      "num_input_tokens_seen": 33744464,
      "step": 38975
    },
    {
      "epoch": 18.37812352663838,
      "grad_norm": 0.02125425450503826,
      "learning_rate": 0.0004820141388283183,
      "loss": 0.1586,
      "num_input_tokens_seen": 33749792,
      "step": 38980
    },
    {
      "epoch": 18.38048090523338,
      "grad_norm": 0.009054761379957199,
      "learning_rate": 0.00047730719740410874,
      "loss": 0.0652,
      "num_input_tokens_seen": 33754096,
      "step": 38985
    },
    {
      "epoch": 18.382838283828384,
      "grad_norm": 0.01772669330239296,
      "learning_rate": 0.00047262331425816927,
      "loss": 0.0602,
      "num_input_tokens_seen": 33758160,
      "step": 38990
    },
    {
      "epoch": 18.385195662423385,
      "grad_norm": 0.010442155413329601,
      "learning_rate": 0.00046796249011277213,
      "loss": 0.0911,
      "num_input_tokens_seen": 33761584,
      "step": 38995
    },
    {
      "epoch": 18.38755304101839,
      "grad_norm": 0.010579949244856834,
      "learning_rate": 0.00046332472568669236,
      "loss": 0.0538,
      "num_input_tokens_seen": 33765936,
      "step": 39000
    },
    {
      "epoch": 18.38755304101839,
      "eval_loss": 0.28802064061164856,
      "eval_runtime": 21.9082,
      "eval_samples_per_second": 43.043,
      "eval_steps_per_second": 21.544,
      "num_input_tokens_seen": 33765936,
      "step": 39000
    },
    {
      "epoch": 18.38991041961339,
      "grad_norm": 0.00810024980455637,
      "learning_rate": 0.0004587100216951578,
      "loss": 0.0772,
      "num_input_tokens_seen": 33770512,
      "step": 39005
    },
    {
      "epoch": 18.392267798208394,
      "grad_norm": 0.010942600667476654,
      "learning_rate": 0.00045411837884978265,
      "loss": 0.0615,
      "num_input_tokens_seen": 33773904,
      "step": 39010
    },
    {
      "epoch": 18.394625176803395,
      "grad_norm": 0.023365259170532227,
      "learning_rate": 0.00044954979785865045,
      "loss": 0.0819,
      "num_input_tokens_seen": 33777920,
      "step": 39015
    },
    {
      "epoch": 18.396982555398395,
      "grad_norm": 0.021406477317214012,
      "learning_rate": 0.00044500427942631426,
      "loss": 0.0769,
      "num_input_tokens_seen": 33781344,
      "step": 39020
    },
    {
      "epoch": 18.3993399339934,
      "grad_norm": 0.00809391401708126,
      "learning_rate": 0.0004404818242537467,
      "loss": 0.0583,
      "num_input_tokens_seen": 33785200,
      "step": 39025
    },
    {
      "epoch": 18.4016973125884,
      "grad_norm": 0.015666499733924866,
      "learning_rate": 0.00043598243303837324,
      "loss": 0.0897,
      "num_input_tokens_seen": 33790320,
      "step": 39030
    },
    {
      "epoch": 18.404054691183404,
      "grad_norm": 0.004839923698455095,
      "learning_rate": 0.00043150610647403885,
      "loss": 0.0712,
      "num_input_tokens_seen": 33794992,
      "step": 39035
    },
    {
      "epoch": 18.406412069778405,
      "grad_norm": 0.006537731271237135,
      "learning_rate": 0.00042705284525104134,
      "loss": 0.0926,
      "num_input_tokens_seen": 33799728,
      "step": 39040
    },
    {
      "epoch": 18.40876944837341,
      "grad_norm": 0.017217909917235374,
      "learning_rate": 0.0004226226500561647,
      "loss": 0.1121,
      "num_input_tokens_seen": 33803936,
      "step": 39045
    },
    {
      "epoch": 18.41112682696841,
      "grad_norm": 0.0037774937227368355,
      "learning_rate": 0.0004182155215725791,
      "loss": 0.0644,
      "num_input_tokens_seen": 33808144,
      "step": 39050
    },
    {
      "epoch": 18.413484205563414,
      "grad_norm": 0.0020505981519818306,
      "learning_rate": 0.00041383146047992424,
      "loss": 0.079,
      "num_input_tokens_seen": 33813280,
      "step": 39055
    },
    {
      "epoch": 18.415841584158414,
      "grad_norm": 0.003849760629236698,
      "learning_rate": 0.00040947046745427597,
      "loss": 0.0409,
      "num_input_tokens_seen": 33817440,
      "step": 39060
    },
    {
      "epoch": 18.41819896275342,
      "grad_norm": 0.006973329931497574,
      "learning_rate": 0.00040513254316814625,
      "loss": 0.0398,
      "num_input_tokens_seen": 33822800,
      "step": 39065
    },
    {
      "epoch": 18.42055634134842,
      "grad_norm": 0.0318143405020237,
      "learning_rate": 0.0004008176882905168,
      "loss": 0.1382,
      "num_input_tokens_seen": 33827376,
      "step": 39070
    },
    {
      "epoch": 18.422913719943423,
      "grad_norm": 0.0020538035314530134,
      "learning_rate": 0.00039652590348677184,
      "loss": 0.0646,
      "num_input_tokens_seen": 33831248,
      "step": 39075
    },
    {
      "epoch": 18.425271098538424,
      "grad_norm": 0.020781999453902245,
      "learning_rate": 0.00039225718941878206,
      "loss": 0.0765,
      "num_input_tokens_seen": 33835056,
      "step": 39080
    },
    {
      "epoch": 18.427628477133428,
      "grad_norm": 0.010085498914122581,
      "learning_rate": 0.00038801154674480417,
      "loss": 0.1413,
      "num_input_tokens_seen": 33839040,
      "step": 39085
    },
    {
      "epoch": 18.42998585572843,
      "grad_norm": 0.016100183129310608,
      "learning_rate": 0.00038378897611959784,
      "loss": 0.0943,
      "num_input_tokens_seen": 33843024,
      "step": 39090
    },
    {
      "epoch": 18.432343234323433,
      "grad_norm": 0.01310497522354126,
      "learning_rate": 0.00037958947819430875,
      "loss": 0.0866,
      "num_input_tokens_seen": 33847232,
      "step": 39095
    },
    {
      "epoch": 18.434700612918434,
      "grad_norm": 0.01717170886695385,
      "learning_rate": 0.0003754130536165856,
      "loss": 0.0897,
      "num_input_tokens_seen": 33851760,
      "step": 39100
    },
    {
      "epoch": 18.437057991513438,
      "grad_norm": 0.013882094994187355,
      "learning_rate": 0.0003712597030304632,
      "loss": 0.0592,
      "num_input_tokens_seen": 33855824,
      "step": 39105
    },
    {
      "epoch": 18.43941537010844,
      "grad_norm": 0.010565144941210747,
      "learning_rate": 0.00036712942707646247,
      "loss": 0.0508,
      "num_input_tokens_seen": 33860128,
      "step": 39110
    },
    {
      "epoch": 18.441772748703443,
      "grad_norm": 0.01672755368053913,
      "learning_rate": 0.00036302222639149063,
      "loss": 0.0826,
      "num_input_tokens_seen": 33864352,
      "step": 39115
    },
    {
      "epoch": 18.444130127298443,
      "grad_norm": 0.017068596556782722,
      "learning_rate": 0.000358938101608941,
      "loss": 0.0858,
      "num_input_tokens_seen": 33869280,
      "step": 39120
    },
    {
      "epoch": 18.446487505893447,
      "grad_norm": 0.025402406230568886,
      "learning_rate": 0.0003548770533586598,
      "loss": 0.0797,
      "num_input_tokens_seen": 33873472,
      "step": 39125
    },
    {
      "epoch": 18.448844884488448,
      "grad_norm": 0.017581013962626457,
      "learning_rate": 0.0003508390822668961,
      "loss": 0.0654,
      "num_input_tokens_seen": 33877344,
      "step": 39130
    },
    {
      "epoch": 18.451202263083452,
      "grad_norm": 0.016756225377321243,
      "learning_rate": 0.00034682418895633503,
      "loss": 0.067,
      "num_input_tokens_seen": 33881600,
      "step": 39135
    },
    {
      "epoch": 18.453559641678453,
      "grad_norm": 0.019894467666745186,
      "learning_rate": 0.0003428323740461647,
      "loss": 0.0974,
      "num_input_tokens_seen": 33885920,
      "step": 39140
    },
    {
      "epoch": 18.455917020273457,
      "grad_norm": 0.012173019349575043,
      "learning_rate": 0.00033886363815194276,
      "loss": 0.0806,
      "num_input_tokens_seen": 33890192,
      "step": 39145
    },
    {
      "epoch": 18.458274398868458,
      "grad_norm": 0.008509778417646885,
      "learning_rate": 0.0003349179818857129,
      "loss": 0.1277,
      "num_input_tokens_seen": 33895168,
      "step": 39150
    },
    {
      "epoch": 18.460631777463462,
      "grad_norm": 0.022791439667344093,
      "learning_rate": 0.0003309954058559383,
      "loss": 0.0956,
      "num_input_tokens_seen": 33899120,
      "step": 39155
    },
    {
      "epoch": 18.462989156058462,
      "grad_norm": 0.023058075457811356,
      "learning_rate": 0.0003270959106675186,
      "loss": 0.1027,
      "num_input_tokens_seen": 33902800,
      "step": 39160
    },
    {
      "epoch": 18.465346534653467,
      "grad_norm": 0.0047293771058321,
      "learning_rate": 0.0003232194969218227,
      "loss": 0.0686,
      "num_input_tokens_seen": 33906704,
      "step": 39165
    },
    {
      "epoch": 18.467703913248467,
      "grad_norm": 0.013443795032799244,
      "learning_rate": 0.00031936616521663905,
      "loss": 0.0996,
      "num_input_tokens_seen": 33911184,
      "step": 39170
    },
    {
      "epoch": 18.47006129184347,
      "grad_norm": 0.003442867659032345,
      "learning_rate": 0.00031553591614619236,
      "loss": 0.044,
      "num_input_tokens_seen": 33915376,
      "step": 39175
    },
    {
      "epoch": 18.472418670438472,
      "grad_norm": 0.008020518347620964,
      "learning_rate": 0.00031172875030117676,
      "loss": 0.0681,
      "num_input_tokens_seen": 33919600,
      "step": 39180
    },
    {
      "epoch": 18.474776049033476,
      "grad_norm": 0.012869726866483688,
      "learning_rate": 0.0003079446682686726,
      "loss": 0.0754,
      "num_input_tokens_seen": 33923520,
      "step": 39185
    },
    {
      "epoch": 18.477133427628477,
      "grad_norm": 0.00506545789539814,
      "learning_rate": 0.0003041836706322465,
      "loss": 0.0601,
      "num_input_tokens_seen": 33927808,
      "step": 39190
    },
    {
      "epoch": 18.47949080622348,
      "grad_norm": 0.006123701110482216,
      "learning_rate": 0.0003004457579719011,
      "loss": 0.0746,
      "num_input_tokens_seen": 33932240,
      "step": 39195
    },
    {
      "epoch": 18.48184818481848,
      "grad_norm": 0.003660484217107296,
      "learning_rate": 0.00029673093086405867,
      "loss": 0.0482,
      "num_input_tokens_seen": 33936896,
      "step": 39200
    },
    {
      "epoch": 18.48184818481848,
      "eval_loss": 0.2888562083244324,
      "eval_runtime": 21.9014,
      "eval_samples_per_second": 43.057,
      "eval_steps_per_second": 21.551,
      "num_input_tokens_seen": 33936896,
      "step": 39200
    },
    {
      "epoch": 18.484205563413486,
      "grad_norm": 0.01670616865158081,
      "learning_rate": 0.00029303918988159426,
      "loss": 0.0942,
      "num_input_tokens_seen": 33940832,
      "step": 39205
    },
    {
      "epoch": 18.486562942008486,
      "grad_norm": 0.02119067683815956,
      "learning_rate": 0.0002893705355938192,
      "loss": 0.0776,
      "num_input_tokens_seen": 33944624,
      "step": 39210
    },
    {
      "epoch": 18.48892032060349,
      "grad_norm": 0.010340698063373566,
      "learning_rate": 0.0002857249685664975,
      "loss": 0.09,
      "num_input_tokens_seen": 33948960,
      "step": 39215
    },
    {
      "epoch": 18.49127769919849,
      "grad_norm": 0.01002565398812294,
      "learning_rate": 0.0002821024893618129,
      "loss": 0.0819,
      "num_input_tokens_seen": 33954144,
      "step": 39220
    },
    {
      "epoch": 18.493635077793492,
      "grad_norm": 0.024105137214064598,
      "learning_rate": 0.0002785030985383852,
      "loss": 0.1095,
      "num_input_tokens_seen": 33958944,
      "step": 39225
    },
    {
      "epoch": 18.495992456388496,
      "grad_norm": 0.010402134619653225,
      "learning_rate": 0.00027492679665130356,
      "loss": 0.0413,
      "num_input_tokens_seen": 33963248,
      "step": 39230
    },
    {
      "epoch": 18.498349834983497,
      "grad_norm": 0.01282298844307661,
      "learning_rate": 0.000271373584252077,
      "loss": 0.0862,
      "num_input_tokens_seen": 33967392,
      "step": 39235
    },
    {
      "epoch": 18.5007072135785,
      "grad_norm": 0.02711041457951069,
      "learning_rate": 0.00026784346188865046,
      "loss": 0.1415,
      "num_input_tokens_seen": 33972096,
      "step": 39240
    },
    {
      "epoch": 18.5030645921735,
      "grad_norm": 0.0017261919565498829,
      "learning_rate": 0.0002643364301054218,
      "loss": 0.0836,
      "num_input_tokens_seen": 33977808,
      "step": 39245
    },
    {
      "epoch": 18.505421970768506,
      "grad_norm": 0.015207303687930107,
      "learning_rate": 0.0002608524894431918,
      "loss": 0.1021,
      "num_input_tokens_seen": 33982192,
      "step": 39250
    },
    {
      "epoch": 18.507779349363506,
      "grad_norm": 0.017558734863996506,
      "learning_rate": 0.000257391640439264,
      "loss": 0.0991,
      "num_input_tokens_seen": 33986544,
      "step": 39255
    },
    {
      "epoch": 18.51013672795851,
      "grad_norm": 0.016638997942209244,
      "learning_rate": 0.00025395388362732806,
      "loss": 0.1141,
      "num_input_tokens_seen": 33991456,
      "step": 39260
    },
    {
      "epoch": 18.51249410655351,
      "grad_norm": 0.025268755853176117,
      "learning_rate": 0.00025053921953751,
      "loss": 0.1638,
      "num_input_tokens_seen": 33995264,
      "step": 39265
    },
    {
      "epoch": 18.514851485148515,
      "grad_norm": 0.010673661716282368,
      "learning_rate": 0.00024714764869643855,
      "loss": 0.0827,
      "num_input_tokens_seen": 33999584,
      "step": 39270
    },
    {
      "epoch": 18.517208863743516,
      "grad_norm": 0.018715115264058113,
      "learning_rate": 0.0002437791716270954,
      "loss": 0.1168,
      "num_input_tokens_seen": 34002752,
      "step": 39275
    },
    {
      "epoch": 18.51956624233852,
      "grad_norm": 0.008257167413830757,
      "learning_rate": 0.00024043378884896493,
      "loss": 0.0574,
      "num_input_tokens_seen": 34007200,
      "step": 39280
    },
    {
      "epoch": 18.52192362093352,
      "grad_norm": 0.022653305903077126,
      "learning_rate": 0.00023711150087793453,
      "loss": 0.0848,
      "num_input_tokens_seen": 34011072,
      "step": 39285
    },
    {
      "epoch": 18.524280999528525,
      "grad_norm": 0.002676167991012335,
      "learning_rate": 0.000233812308226361,
      "loss": 0.0625,
      "num_input_tokens_seen": 34015216,
      "step": 39290
    },
    {
      "epoch": 18.526638378123526,
      "grad_norm": 0.011457450687885284,
      "learning_rate": 0.00023053621140300406,
      "loss": 0.177,
      "num_input_tokens_seen": 34020048,
      "step": 39295
    },
    {
      "epoch": 18.52899575671853,
      "grad_norm": 0.024237683042883873,
      "learning_rate": 0.00022728321091307623,
      "loss": 0.1379,
      "num_input_tokens_seen": 34024352,
      "step": 39300
    },
    {
      "epoch": 18.53135313531353,
      "grad_norm": 0.016124477609992027,
      "learning_rate": 0.0002240533072582429,
      "loss": 0.0943,
      "num_input_tokens_seen": 34027952,
      "step": 39305
    },
    {
      "epoch": 18.533710513908535,
      "grad_norm": 0.017075084149837494,
      "learning_rate": 0.00022084650093658897,
      "loss": 0.1281,
      "num_input_tokens_seen": 34031856,
      "step": 39310
    },
    {
      "epoch": 18.536067892503535,
      "grad_norm": 0.0207653921097517,
      "learning_rate": 0.0002176627924426522,
      "loss": 0.1554,
      "num_input_tokens_seen": 34036608,
      "step": 39315
    },
    {
      "epoch": 18.53842527109854,
      "grad_norm": 0.013970598578453064,
      "learning_rate": 0.0002145021822673898,
      "loss": 0.0782,
      "num_input_tokens_seen": 34040656,
      "step": 39320
    },
    {
      "epoch": 18.54078264969354,
      "grad_norm": 0.005550181493163109,
      "learning_rate": 0.00021136467089822862,
      "loss": 0.0913,
      "num_input_tokens_seen": 34044832,
      "step": 39325
    },
    {
      "epoch": 18.543140028288544,
      "grad_norm": 0.0037104832008481026,
      "learning_rate": 0.00020825025881898162,
      "loss": 0.0655,
      "num_input_tokens_seen": 34049808,
      "step": 39330
    },
    {
      "epoch": 18.545497406883545,
      "grad_norm": 0.0077162547968328,
      "learning_rate": 0.0002051589465099479,
      "loss": 0.0722,
      "num_input_tokens_seen": 34054528,
      "step": 39335
    },
    {
      "epoch": 18.54785478547855,
      "grad_norm": 0.013427382335066795,
      "learning_rate": 0.0002020907344478462,
      "loss": 0.0727,
      "num_input_tokens_seen": 34058912,
      "step": 39340
    },
    {
      "epoch": 18.55021216407355,
      "grad_norm": 0.004311059135943651,
      "learning_rate": 0.0001990456231058313,
      "loss": 0.068,
      "num_input_tokens_seen": 34063504,
      "step": 39345
    },
    {
      "epoch": 18.552569542668554,
      "grad_norm": 0.01876750960946083,
      "learning_rate": 0.00019602361295349423,
      "loss": 0.0627,
      "num_input_tokens_seen": 34067856,
      "step": 39350
    },
    {
      "epoch": 18.554926921263554,
      "grad_norm": 0.02497601881623268,
      "learning_rate": 0.0001930247044568789,
      "loss": 0.1235,
      "num_input_tokens_seen": 34071872,
      "step": 39355
    },
    {
      "epoch": 18.55728429985856,
      "grad_norm": 0.009049373678863049,
      "learning_rate": 0.00019004889807843205,
      "loss": 0.0594,
      "num_input_tokens_seen": 34076736,
      "step": 39360
    },
    {
      "epoch": 18.55964167845356,
      "grad_norm": 0.01647106185555458,
      "learning_rate": 0.00018709619427708656,
      "loss": 0.1006,
      "num_input_tokens_seen": 34081408,
      "step": 39365
    },
    {
      "epoch": 18.561999057048563,
      "grad_norm": 0.008814402855932713,
      "learning_rate": 0.00018416659350817822,
      "loss": 0.0695,
      "num_input_tokens_seen": 34085552,
      "step": 39370
    },
    {
      "epoch": 18.564356435643564,
      "grad_norm": 0.017760606482625008,
      "learning_rate": 0.00018126009622346229,
      "loss": 0.079,
      "num_input_tokens_seen": 34090320,
      "step": 39375
    },
    {
      "epoch": 18.566713814238568,
      "grad_norm": 0.01971689984202385,
      "learning_rate": 0.00017837670287119687,
      "loss": 0.1752,
      "num_input_tokens_seen": 34094208,
      "step": 39380
    },
    {
      "epoch": 18.56907119283357,
      "grad_norm": 0.009442835114896297,
      "learning_rate": 0.00017551641389602633,
      "loss": 0.09,
      "num_input_tokens_seen": 34098960,
      "step": 39385
    },
    {
      "epoch": 18.571428571428573,
      "grad_norm": 0.010458395816385746,
      "learning_rate": 0.00017267922973903115,
      "loss": 0.056,
      "num_input_tokens_seen": 34102944,
      "step": 39390
    },
    {
      "epoch": 18.573785950023574,
      "grad_norm": 0.009192125871777534,
      "learning_rate": 0.00016986515083774467,
      "loss": 0.1425,
      "num_input_tokens_seen": 34106752,
      "step": 39395
    },
    {
      "epoch": 18.576143328618578,
      "grad_norm": 0.009287843480706215,
      "learning_rate": 0.00016707417762611975,
      "loss": 0.0872,
      "num_input_tokens_seen": 34110592,
      "step": 39400
    },
    {
      "epoch": 18.576143328618578,
      "eval_loss": 0.28899332880973816,
      "eval_runtime": 21.8862,
      "eval_samples_per_second": 43.086,
      "eval_steps_per_second": 21.566,
      "num_input_tokens_seen": 34110592,
      "step": 39400
    },
    {
      "epoch": 18.57850070721358,
      "grad_norm": 0.017889028415083885,
      "learning_rate": 0.00016430631053459543,
      "loss": 0.0842,
      "num_input_tokens_seen": 34114256,
      "step": 39405
    },
    {
      "epoch": 18.580858085808583,
      "grad_norm": 0.015459234826266766,
      "learning_rate": 0.0001615615499899803,
      "loss": 0.1131,
      "num_input_tokens_seen": 34118240,
      "step": 39410
    },
    {
      "epoch": 18.583215464403583,
      "grad_norm": 0.008754379115998745,
      "learning_rate": 0.00015883989641556905,
      "loss": 0.1322,
      "num_input_tokens_seen": 34122896,
      "step": 39415
    },
    {
      "epoch": 18.585572842998587,
      "grad_norm": 0.012524116784334183,
      "learning_rate": 0.00015614135023105934,
      "loss": 0.1273,
      "num_input_tokens_seen": 34127008,
      "step": 39420
    },
    {
      "epoch": 18.587930221593588,
      "grad_norm": 0.008522243238985538,
      "learning_rate": 0.00015346591185261827,
      "loss": 0.0663,
      "num_input_tokens_seen": 34131632,
      "step": 39425
    },
    {
      "epoch": 18.59028760018859,
      "grad_norm": 0.01596761867403984,
      "learning_rate": 0.00015081358169281576,
      "loss": 0.0841,
      "num_input_tokens_seen": 34136000,
      "step": 39430
    },
    {
      "epoch": 18.592644978783593,
      "grad_norm": 0.017875174060463905,
      "learning_rate": 0.00014818436016069135,
      "loss": 0.0996,
      "num_input_tokens_seen": 34140672,
      "step": 39435
    },
    {
      "epoch": 18.595002357378593,
      "grad_norm": 0.020815974101424217,
      "learning_rate": 0.00014557824766168735,
      "loss": 0.0791,
      "num_input_tokens_seen": 34144512,
      "step": 39440
    },
    {
      "epoch": 18.597359735973598,
      "grad_norm": 0.01341498177498579,
      "learning_rate": 0.00014299524459769896,
      "loss": 0.0811,
      "num_input_tokens_seen": 34148288,
      "step": 39445
    },
    {
      "epoch": 18.599717114568598,
      "grad_norm": 0.008876568637788296,
      "learning_rate": 0.0001404353513670742,
      "loss": 0.0695,
      "num_input_tokens_seen": 34152128,
      "step": 39450
    },
    {
      "epoch": 18.602074493163602,
      "grad_norm": 0.0228035319596529,
      "learning_rate": 0.0001378985683645806,
      "loss": 0.0976,
      "num_input_tokens_seen": 34156192,
      "step": 39455
    },
    {
      "epoch": 18.604431871758603,
      "grad_norm": 0.01861092634499073,
      "learning_rate": 0.0001353848959813886,
      "loss": 0.1109,
      "num_input_tokens_seen": 34160064,
      "step": 39460
    },
    {
      "epoch": 18.606789250353607,
      "grad_norm": 0.006291537545621395,
      "learning_rate": 0.00013289433460517142,
      "loss": 0.0594,
      "num_input_tokens_seen": 34164160,
      "step": 39465
    },
    {
      "epoch": 18.609146628948608,
      "grad_norm": 0.007573523558676243,
      "learning_rate": 0.00013042688462000518,
      "loss": 0.055,
      "num_input_tokens_seen": 34168944,
      "step": 39470
    },
    {
      "epoch": 18.611504007543612,
      "grad_norm": 0.014548714272677898,
      "learning_rate": 0.0001279825464063855,
      "loss": 0.0792,
      "num_input_tokens_seen": 34173856,
      "step": 39475
    },
    {
      "epoch": 18.613861386138613,
      "grad_norm": 0.009897487238049507,
      "learning_rate": 0.00012556132034126087,
      "loss": 0.0692,
      "num_input_tokens_seen": 34178624,
      "step": 39480
    },
    {
      "epoch": 18.616218764733617,
      "grad_norm": 0.005154917947947979,
      "learning_rate": 0.0001231632067980326,
      "loss": 0.0747,
      "num_input_tokens_seen": 34184048,
      "step": 39485
    },
    {
      "epoch": 18.618576143328617,
      "grad_norm": 0.015009189024567604,
      "learning_rate": 0.00012078820614650486,
      "loss": 0.1342,
      "num_input_tokens_seen": 34188256,
      "step": 39490
    },
    {
      "epoch": 18.62093352192362,
      "grad_norm": 0.006844956427812576,
      "learning_rate": 0.00011843631875291804,
      "loss": 0.0606,
      "num_input_tokens_seen": 34192544,
      "step": 39495
    },
    {
      "epoch": 18.623290900518622,
      "grad_norm": 0.012442706152796745,
      "learning_rate": 0.00011610754497999863,
      "loss": 0.0645,
      "num_input_tokens_seen": 34196896,
      "step": 39500
    },
    {
      "epoch": 18.625648279113626,
      "grad_norm": 0.019810406491160393,
      "learning_rate": 0.0001138018851868594,
      "loss": 0.0687,
      "num_input_tokens_seen": 34200512,
      "step": 39505
    },
    {
      "epoch": 18.628005657708627,
      "grad_norm": 0.005730709992349148,
      "learning_rate": 0.0001115193397290326,
      "loss": 0.1299,
      "num_input_tokens_seen": 34204672,
      "step": 39510
    },
    {
      "epoch": 18.63036303630363,
      "grad_norm": 0.01674485206604004,
      "learning_rate": 0.00010925990895856996,
      "loss": 0.1067,
      "num_input_tokens_seen": 34209472,
      "step": 39515
    },
    {
      "epoch": 18.632720414898632,
      "grad_norm": 0.0068301050923764706,
      "learning_rate": 0.00010702359322385946,
      "loss": 0.0784,
      "num_input_tokens_seen": 34214064,
      "step": 39520
    },
    {
      "epoch": 18.635077793493636,
      "grad_norm": 0.006076378747820854,
      "learning_rate": 0.00010481039286977523,
      "loss": 0.0817,
      "num_input_tokens_seen": 34218864,
      "step": 39525
    },
    {
      "epoch": 18.637435172088637,
      "grad_norm": 0.016974911093711853,
      "learning_rate": 0.00010262030823764423,
      "loss": 0.0899,
      "num_input_tokens_seen": 34222768,
      "step": 39530
    },
    {
      "epoch": 18.63979255068364,
      "grad_norm": 0.0067319804802536964,
      "learning_rate": 0.00010045333966517966,
      "loss": 0.0666,
      "num_input_tokens_seen": 34226896,
      "step": 39535
    },
    {
      "epoch": 18.64214992927864,
      "grad_norm": 0.02635984681546688,
      "learning_rate": 9.83094874865642e-05,
      "loss": 0.09,
      "num_input_tokens_seen": 34231472,
      "step": 39540
    },
    {
      "epoch": 18.644507307873646,
      "grad_norm": 0.01712076924741268,
      "learning_rate": 9.618875203241672e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 34235696,
      "step": 39545
    },
    {
      "epoch": 18.646864686468646,
      "grad_norm": 0.027750233188271523,
      "learning_rate": 9.409113362977561e-05,
      "loss": 0.1756,
      "num_input_tokens_seen": 34240256,
      "step": 39550
    },
    {
      "epoch": 18.64922206506365,
      "grad_norm": 0.00917709618806839,
      "learning_rate": 9.20166326020988e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 34243760,
      "step": 39555
    },
    {
      "epoch": 18.65157944365865,
      "grad_norm": 0.009253164753317833,
      "learning_rate": 8.996524926933035e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 34249200,
      "step": 39560
    },
    {
      "epoch": 18.653936822253655,
      "grad_norm": 0.008726068772375584,
      "learning_rate": 8.793698394781723e-05,
      "loss": 0.1132,
      "num_input_tokens_seen": 34252656,
      "step": 39565
    },
    {
      "epoch": 18.656294200848656,
      "grad_norm": 0.02091221511363983,
      "learning_rate": 8.593183695030926e-05,
      "loss": 0.102,
      "num_input_tokens_seen": 34256688,
      "step": 39570
    },
    {
      "epoch": 18.65865157944366,
      "grad_norm": 0.003981017041951418,
      "learning_rate": 8.39498085860757e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 34261440,
      "step": 39575
    },
    {
      "epoch": 18.66100895803866,
      "grad_norm": 0.015008358284831047,
      "learning_rate": 8.199089916072211e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 34265776,
      "step": 39580
    },
    {
      "epoch": 18.663366336633665,
      "grad_norm": 0.0066367038525640965,
      "learning_rate": 8.005510897637346e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 34270784,
      "step": 39585
    },
    {
      "epoch": 18.665723715228665,
      "grad_norm": 0.02477482333779335,
      "learning_rate": 7.8142438331541e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 34274736,
      "step": 39590
    },
    {
      "epoch": 18.66808109382367,
      "grad_norm": 0.009845632128417492,
      "learning_rate": 7.625288752117209e-05,
      "loss": 0.1083,
      "num_input_tokens_seen": 34280352,
      "step": 39595
    },
    {
      "epoch": 18.67043847241867,
      "grad_norm": 0.018831059336662292,
      "learning_rate": 7.4386456836667e-05,
      "loss": 0.1452,
      "num_input_tokens_seen": 34284208,
      "step": 39600
    },
    {
      "epoch": 18.67043847241867,
      "eval_loss": 0.2887762784957886,
      "eval_runtime": 21.8799,
      "eval_samples_per_second": 43.099,
      "eval_steps_per_second": 21.572,
      "num_input_tokens_seen": 34284208,
      "step": 39600
    },
    {
      "epoch": 18.672795851013674,
      "grad_norm": 0.004413051530718803,
      "learning_rate": 7.254314656586214e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 34288928,
      "step": 39605
    },
    {
      "epoch": 18.675153229608675,
      "grad_norm": 0.022885063663125038,
      "learning_rate": 7.07229569929968e-05,
      "loss": 0.0919,
      "num_input_tokens_seen": 34293136,
      "step": 39610
    },
    {
      "epoch": 18.677510608203676,
      "grad_norm": 0.005017272662371397,
      "learning_rate": 6.892588839879643e-05,
      "loss": 0.0891,
      "num_input_tokens_seen": 34297984,
      "step": 39615
    },
    {
      "epoch": 18.67986798679868,
      "grad_norm": 0.01828698441386223,
      "learning_rate": 6.71519410603727e-05,
      "loss": 0.0809,
      "num_input_tokens_seen": 34301904,
      "step": 39620
    },
    {
      "epoch": 18.68222536539368,
      "grad_norm": 0.007361389696598053,
      "learning_rate": 6.540111525129011e-05,
      "loss": 0.0947,
      "num_input_tokens_seen": 34306368,
      "step": 39625
    },
    {
      "epoch": 18.684582743988685,
      "grad_norm": 0.01193610206246376,
      "learning_rate": 6.367341124154934e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 34310880,
      "step": 39630
    },
    {
      "epoch": 18.686940122583685,
      "grad_norm": 0.008383829146623611,
      "learning_rate": 6.19688292975873e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 34314592,
      "step": 39635
    },
    {
      "epoch": 18.68929750117869,
      "grad_norm": 0.016446270048618317,
      "learning_rate": 6.0287369682260336e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 34318864,
      "step": 39640
    },
    {
      "epoch": 18.69165487977369,
      "grad_norm": 0.0070274299941957,
      "learning_rate": 5.8629032654894384e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 34324000,
      "step": 39645
    },
    {
      "epoch": 18.694012258368694,
      "grad_norm": 0.01622469164431095,
      "learning_rate": 5.699381847120155e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 34327840,
      "step": 39650
    },
    {
      "epoch": 18.696369636963695,
      "grad_norm": 0.016749674454331398,
      "learning_rate": 5.5381727383380094e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 34331664,
      "step": 39655
    },
    {
      "epoch": 18.6987270155587,
      "grad_norm": 0.008495217189192772,
      "learning_rate": 5.379275964001451e-05,
      "loss": 0.0253,
      "num_input_tokens_seen": 34336016,
      "step": 39660
    },
    {
      "epoch": 18.7010843941537,
      "grad_norm": 0.025003831833600998,
      "learning_rate": 5.222691548614211e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 34341072,
      "step": 39665
    },
    {
      "epoch": 18.703441772748704,
      "grad_norm": 0.013879910111427307,
      "learning_rate": 5.068419516323641e-05,
      "loss": 0.0991,
      "num_input_tokens_seen": 34345824,
      "step": 39670
    },
    {
      "epoch": 18.705799151343705,
      "grad_norm": 0.031011074781417847,
      "learning_rate": 4.91645989092071e-05,
      "loss": 0.1449,
      "num_input_tokens_seen": 34350784,
      "step": 39675
    },
    {
      "epoch": 18.70815652993871,
      "grad_norm": 0.014465633779764175,
      "learning_rate": 4.7668126958400056e-05,
      "loss": 0.1564,
      "num_input_tokens_seen": 34354624,
      "step": 39680
    },
    {
      "epoch": 18.71051390853371,
      "grad_norm": 0.01345904916524887,
      "learning_rate": 4.619477954159734e-05,
      "loss": 0.0976,
      "num_input_tokens_seen": 34359088,
      "step": 39685
    },
    {
      "epoch": 18.712871287128714,
      "grad_norm": 0.013596734963357449,
      "learning_rate": 4.4744556885983884e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 34363008,
      "step": 39690
    },
    {
      "epoch": 18.715228665723714,
      "grad_norm": 0.021314403042197227,
      "learning_rate": 4.331745921523078e-05,
      "loss": 0.1558,
      "num_input_tokens_seen": 34367504,
      "step": 39695
    },
    {
      "epoch": 18.71758604431872,
      "grad_norm": 0.015984216704964638,
      "learning_rate": 4.191348674937867e-05,
      "loss": 0.0977,
      "num_input_tokens_seen": 34371760,
      "step": 39700
    },
    {
      "epoch": 18.71994342291372,
      "grad_norm": 0.005226494744420052,
      "learning_rate": 4.0532639704971006e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 34375440,
      "step": 39705
    },
    {
      "epoch": 18.722300801508723,
      "grad_norm": 0.009761794470250607,
      "learning_rate": 3.917491829493747e-05,
      "loss": 0.109,
      "num_input_tokens_seen": 34379440,
      "step": 39710
    },
    {
      "epoch": 18.724658180103724,
      "grad_norm": 0.02526105009019375,
      "learning_rate": 3.78403227286439e-05,
      "loss": 0.0891,
      "num_input_tokens_seen": 34383552,
      "step": 39715
    },
    {
      "epoch": 18.727015558698728,
      "grad_norm": 0.0034098774194717407,
      "learning_rate": 3.652885321192567e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 34387136,
      "step": 39720
    },
    {
      "epoch": 18.72937293729373,
      "grad_norm": 0.0038812693674117327,
      "learning_rate": 3.524050994702099e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 34391536,
      "step": 39725
    },
    {
      "epoch": 18.731730315888733,
      "grad_norm": 0.017734356224536896,
      "learning_rate": 3.3975293132604276e-05,
      "loss": 0.1053,
      "num_input_tokens_seen": 34395200,
      "step": 39730
    },
    {
      "epoch": 18.734087694483733,
      "grad_norm": 0.007793187629431486,
      "learning_rate": 3.2733202963786125e-05,
      "loss": 0.1325,
      "num_input_tokens_seen": 34399584,
      "step": 39735
    },
    {
      "epoch": 18.736445073078738,
      "grad_norm": 0.015179858542978764,
      "learning_rate": 3.15142396321133e-05,
      "loss": 0.1031,
      "num_input_tokens_seen": 34405296,
      "step": 39740
    },
    {
      "epoch": 18.738802451673738,
      "grad_norm": 0.0029584902804344893,
      "learning_rate": 3.0318403325552132e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 34409360,
      "step": 39745
    },
    {
      "epoch": 18.741159830268742,
      "grad_norm": 0.03023880161345005,
      "learning_rate": 2.914569422855506e-05,
      "loss": 0.1374,
      "num_input_tokens_seen": 34413136,
      "step": 39750
    },
    {
      "epoch": 18.743517208863743,
      "grad_norm": 0.011132280342280865,
      "learning_rate": 2.7996112521927462e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 34417488,
      "step": 39755
    },
    {
      "epoch": 18.745874587458747,
      "grad_norm": 0.010531708598136902,
      "learning_rate": 2.68696583829775e-05,
      "loss": 0.0466,
      "num_input_tokens_seen": 34421680,
      "step": 39760
    },
    {
      "epoch": 18.748231966053748,
      "grad_norm": 0.010450134985148907,
      "learning_rate": 2.576633198539957e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 34425296,
      "step": 39765
    },
    {
      "epoch": 18.750589344648752,
      "grad_norm": 0.013927256688475609,
      "learning_rate": 2.46861334993409e-05,
      "loss": 0.1092,
      "num_input_tokens_seen": 34430256,
      "step": 39770
    },
    {
      "epoch": 18.752946723243753,
      "grad_norm": 0.014284822158515453,
      "learning_rate": 2.3629063091384903e-05,
      "loss": 0.0797,
      "num_input_tokens_seen": 34435024,
      "step": 39775
    },
    {
      "epoch": 18.755304101838757,
      "grad_norm": 0.01127520203590393,
      "learning_rate": 2.2595120924567834e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 34440256,
      "step": 39780
    },
    {
      "epoch": 18.757661480433757,
      "grad_norm": 0.0074231550097465515,
      "learning_rate": 2.158430715829551e-05,
      "loss": 0.1007,
      "num_input_tokens_seen": 34444752,
      "step": 39785
    },
    {
      "epoch": 18.76001885902876,
      "grad_norm": 0.008589277043938637,
      "learning_rate": 2.059662194849321e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 34449952,
      "step": 39790
    },
    {
      "epoch": 18.762376237623762,
      "grad_norm": 0.030056290328502655,
      "learning_rate": 1.9632065447422463e-05,
      "loss": 0.1302,
      "num_input_tokens_seen": 34454112,
      "step": 39795
    },
    {
      "epoch": 18.764733616218766,
      "grad_norm": 0.008744222111999989,
      "learning_rate": 1.8690637803880916e-05,
      "loss": 0.1472,
      "num_input_tokens_seen": 34458576,
      "step": 39800
    },
    {
      "epoch": 18.764733616218766,
      "eval_loss": 0.28791555762290955,
      "eval_runtime": 21.9114,
      "eval_samples_per_second": 43.037,
      "eval_steps_per_second": 21.541,
      "num_input_tokens_seen": 34458576,
      "step": 39800
    },
    {
      "epoch": 18.767090994813767,
      "grad_norm": 0.0026055576745420694,
      "learning_rate": 1.7772339163019123e-05,
      "loss": 0.0218,
      "num_input_tokens_seen": 34462688,
      "step": 39805
    },
    {
      "epoch": 18.76944837340877,
      "grad_norm": 0.022913895547389984,
      "learning_rate": 1.6877169666457138e-05,
      "loss": 0.1083,
      "num_input_tokens_seen": 34466688,
      "step": 39810
    },
    {
      "epoch": 18.77180575200377,
      "grad_norm": 0.018014898523688316,
      "learning_rate": 1.6005129452234532e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 34471056,
      "step": 39815
    },
    {
      "epoch": 18.774163130598772,
      "grad_norm": 0.0030772502068430185,
      "learning_rate": 1.5156218654843733e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 34476880,
      "step": 39820
    },
    {
      "epoch": 18.776520509193777,
      "grad_norm": 0.017863677814602852,
      "learning_rate": 1.4330437405196683e-05,
      "loss": 0.091,
      "num_input_tokens_seen": 34480576,
      "step": 39825
    },
    {
      "epoch": 18.778877887788777,
      "grad_norm": 0.02051074430346489,
      "learning_rate": 1.352778583062486e-05,
      "loss": 0.09,
      "num_input_tokens_seen": 34484960,
      "step": 39830
    },
    {
      "epoch": 18.78123526638378,
      "grad_norm": 0.015396296046674252,
      "learning_rate": 1.2748264054929237e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 34489392,
      "step": 39835
    },
    {
      "epoch": 18.783592644978782,
      "grad_norm": 0.008194101974368095,
      "learning_rate": 1.1991872198297004e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 34494096,
      "step": 39840
    },
    {
      "epoch": 18.785950023573786,
      "grad_norm": 0.013768205419182777,
      "learning_rate": 1.1258610377384847e-05,
      "loss": 0.1238,
      "num_input_tokens_seen": 34499232,
      "step": 39845
    },
    {
      "epoch": 18.788307402168787,
      "grad_norm": 0.013058130629360676,
      "learning_rate": 1.0548478705268982e-05,
      "loss": 0.1082,
      "num_input_tokens_seen": 34503120,
      "step": 39850
    },
    {
      "epoch": 18.79066478076379,
      "grad_norm": 0.014570710249245167,
      "learning_rate": 9.86147729147846e-06,
      "loss": 0.1095,
      "num_input_tokens_seen": 34506656,
      "step": 39855
    },
    {
      "epoch": 18.79302215935879,
      "grad_norm": 0.014433410950005054,
      "learning_rate": 9.197606241928557e-06,
      "loss": 0.0733,
      "num_input_tokens_seen": 34509840,
      "step": 39860
    },
    {
      "epoch": 18.795379537953796,
      "grad_norm": 0.003787419991567731,
      "learning_rate": 8.556865659004042e-06,
      "loss": 0.0696,
      "num_input_tokens_seen": 34514320,
      "step": 39865
    },
    {
      "epoch": 18.797736916548796,
      "grad_norm": 0.012462849728763103,
      "learning_rate": 7.939255641525867e-06,
      "loss": 0.1178,
      "num_input_tokens_seen": 34518480,
      "step": 39870
    },
    {
      "epoch": 18.8000942951438,
      "grad_norm": 0.013562124222517014,
      "learning_rate": 7.344776284751164e-06,
      "loss": 0.0991,
      "num_input_tokens_seen": 34522592,
      "step": 39875
    },
    {
      "epoch": 18.8024516737388,
      "grad_norm": 0.01157116424292326,
      "learning_rate": 6.773427680323296e-06,
      "loss": 0.0594,
      "num_input_tokens_seen": 34527728,
      "step": 39880
    },
    {
      "epoch": 18.804809052333805,
      "grad_norm": 0.023624740540981293,
      "learning_rate": 6.225209916355112e-06,
      "loss": 0.1066,
      "num_input_tokens_seen": 34531584,
      "step": 39885
    },
    {
      "epoch": 18.807166430928806,
      "grad_norm": 0.0068284086883068085,
      "learning_rate": 5.7001230774123e-06,
      "loss": 0.0671,
      "num_input_tokens_seen": 34535776,
      "step": 39890
    },
    {
      "epoch": 18.80952380952381,
      "grad_norm": 0.008424722589552402,
      "learning_rate": 5.198167244446772e-06,
      "loss": 0.0672,
      "num_input_tokens_seen": 34540704,
      "step": 39895
    },
    {
      "epoch": 18.81188118811881,
      "grad_norm": 0.02570182830095291,
      "learning_rate": 4.71934249487993e-06,
      "loss": 0.1173,
      "num_input_tokens_seen": 34545504,
      "step": 39900
    },
    {
      "epoch": 18.814238566713815,
      "grad_norm": 0.0018481033621355891,
      "learning_rate": 4.2636489025527075e-06,
      "loss": 0.1157,
      "num_input_tokens_seen": 34549600,
      "step": 39905
    },
    {
      "epoch": 18.816595945308816,
      "grad_norm": 0.019510336220264435,
      "learning_rate": 3.831086537742223e-06,
      "loss": 0.1069,
      "num_input_tokens_seen": 34553872,
      "step": 39910
    },
    {
      "epoch": 18.81895332390382,
      "grad_norm": 0.009640456177294254,
      "learning_rate": 3.4216554671451236e-06,
      "loss": 0.1426,
      "num_input_tokens_seen": 34558400,
      "step": 39915
    },
    {
      "epoch": 18.82131070249882,
      "grad_norm": 0.021989544853568077,
      "learning_rate": 3.035355753894242e-06,
      "loss": 0.0716,
      "num_input_tokens_seen": 34562400,
      "step": 39920
    },
    {
      "epoch": 18.823668081093825,
      "grad_norm": 0.008589406497776508,
      "learning_rate": 2.6721874575752477e-06,
      "loss": 0.0641,
      "num_input_tokens_seen": 34567024,
      "step": 39925
    },
    {
      "epoch": 18.826025459688825,
      "grad_norm": 0.010782629251480103,
      "learning_rate": 2.3321506341933418e-06,
      "loss": 0.0532,
      "num_input_tokens_seen": 34571696,
      "step": 39930
    },
    {
      "epoch": 18.82838283828383,
      "grad_norm": 0.02195165306329727,
      "learning_rate": 2.0152453361732546e-06,
      "loss": 0.1307,
      "num_input_tokens_seen": 34575904,
      "step": 39935
    },
    {
      "epoch": 18.83074021687883,
      "grad_norm": 0.02232843078672886,
      "learning_rate": 1.7214716123925554e-06,
      "loss": 0.1079,
      "num_input_tokens_seen": 34579440,
      "step": 39940
    },
    {
      "epoch": 18.833097595473834,
      "grad_norm": 0.019134260714054108,
      "learning_rate": 1.4508295081649968e-06,
      "loss": 0.0735,
      "num_input_tokens_seen": 34583632,
      "step": 39945
    },
    {
      "epoch": 18.835454974068835,
      "grad_norm": 0.024946492165327072,
      "learning_rate": 1.2033190652238623e-06,
      "loss": 0.1048,
      "num_input_tokens_seen": 34588304,
      "step": 39950
    },
    {
      "epoch": 18.83781235266384,
      "grad_norm": 0.01863667368888855,
      "learning_rate": 9.78940321721966e-07,
      "loss": 0.084,
      "num_input_tokens_seen": 34592464,
      "step": 39955
    },
    {
      "epoch": 18.84016973125884,
      "grad_norm": 0.005115286447107792,
      "learning_rate": 7.776933122816132e-07,
      "loss": 0.1006,
      "num_input_tokens_seen": 34597024,
      "step": 39960
    },
    {
      "epoch": 18.842527109853844,
      "grad_norm": 0.004415691830217838,
      "learning_rate": 5.99578067927986e-07,
      "loss": 0.0979,
      "num_input_tokens_seen": 34601488,
      "step": 39965
    },
    {
      "epoch": 18.844884488448844,
      "grad_norm": 0.007164729293435812,
      "learning_rate": 4.445946161224512e-07,
      "loss": 0.035,
      "num_input_tokens_seen": 34606048,
      "step": 39970
    },
    {
      "epoch": 18.84724186704385,
      "grad_norm": 0.010828171856701374,
      "learning_rate": 3.127429807792126e-07,
      "loss": 0.1639,
      "num_input_tokens_seen": 34611376,
      "step": 39975
    },
    {
      "epoch": 18.84959924563885,
      "grad_norm": 0.006289685145020485,
      "learning_rate": 2.040231822320049e-07,
      "loss": 0.0383,
      "num_input_tokens_seen": 34616288,
      "step": 39980
    },
    {
      "epoch": 18.851956624233853,
      "grad_norm": 0.010136693716049194,
      "learning_rate": 1.1843523723409354e-07,
      "loss": 0.0553,
      "num_input_tokens_seen": 34620368,
      "step": 39985
    },
    {
      "epoch": 18.854314002828854,
      "grad_norm": 0.008962450549006462,
      "learning_rate": 5.597915897492811e-08,
      "loss": 0.0717,
      "num_input_tokens_seen": 34624976,
      "step": 39990
    },
    {
      "epoch": 18.85667138142386,
      "grad_norm": 0.020498232915997505,
      "learning_rate": 1.6654957113448885e-08,
      "loss": 0.0731,
      "num_input_tokens_seen": 34629024,
      "step": 39995
    },
    {
      "epoch": 18.85902876001886,
      "grad_norm": 0.016933605074882507,
      "learning_rate": 4.626377114735902e-10,
      "loss": 0.0763,
      "num_input_tokens_seen": 34633072,
      "step": 40000
    },
    {
      "epoch": 18.85902876001886,
      "eval_loss": 0.288237065076828,
      "eval_runtime": 21.9439,
      "eval_samples_per_second": 42.973,
      "eval_steps_per_second": 21.509,
      "num_input_tokens_seen": 34633072,
      "step": 40000
    },
    {
      "epoch": 18.85902876001886,
      "num_input_tokens_seen": 34633072,
      "step": 40000,
      "total_flos": 1.4502170779817165e+17,
      "train_loss": 0.20583191509507595,
      "train_runtime": 12813.2622,
      "train_samples_per_second": 12.487,
      "train_steps_per_second": 3.122
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 34633072,
  "num_train_epochs": 19,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.4502170779817165e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}