{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998919736415686,
  "eval_steps": 500,
  "global_step": 3471,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0014403514457527638,
      "grad_norm": 9.498580060596645,
      "learning_rate": 2.2988505747126437e-07,
      "loss": 1.9905,
      "step": 5
    },
    {
      "epoch": 0.0028807028915055276,
      "grad_norm": 9.29730840350754,
      "learning_rate": 5.172413793103449e-07,
      "loss": 1.9858,
      "step": 10
    },
    {
      "epoch": 0.004321054337258291,
      "grad_norm": 8.510677726863225,
      "learning_rate": 8.045977011494253e-07,
      "loss": 1.9545,
      "step": 15
    },
    {
      "epoch": 0.005761405783011055,
      "grad_norm": 7.429305787241992,
      "learning_rate": 1.0919540229885058e-06,
      "loss": 1.838,
      "step": 20
    },
    {
      "epoch": 0.007201757228763818,
      "grad_norm": 6.639158892472215,
      "learning_rate": 1.3793103448275862e-06,
      "loss": 1.658,
      "step": 25
    },
    {
      "epoch": 0.008642108674516582,
      "grad_norm": 7.2050289141669035,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.3294,
      "step": 30
    },
    {
      "epoch": 0.010082460120269346,
      "grad_norm": 7.64848090123899,
      "learning_rate": 1.9540229885057475e-06,
      "loss": 1.0103,
      "step": 35
    },
    {
      "epoch": 0.01152281156602211,
      "grad_norm": 6.412853419273422,
      "learning_rate": 2.241379310344828e-06,
      "loss": 0.5327,
      "step": 40
    },
    {
      "epoch": 0.012963163011774873,
      "grad_norm": 1.776900584940695,
      "learning_rate": 2.5287356321839083e-06,
      "loss": 0.2908,
      "step": 45
    },
    {
      "epoch": 0.014403514457527637,
      "grad_norm": 1.985660623026991,
      "learning_rate": 2.8160919540229887e-06,
      "loss": 0.1841,
      "step": 50
    },
    {
      "epoch": 0.0158438659032804,
      "grad_norm": 0.7462636137858207,
      "learning_rate": 3.103448275862069e-06,
      "loss": 0.1707,
      "step": 55
    },
    {
      "epoch": 0.017284217349033165,
      "grad_norm": 0.7526761920440953,
      "learning_rate": 3.3908045977011496e-06,
      "loss": 0.1525,
      "step": 60
    },
    {
      "epoch": 0.01872456879478593,
      "grad_norm": 0.7764098707095316,
      "learning_rate": 3.67816091954023e-06,
      "loss": 0.1407,
      "step": 65
    },
    {
      "epoch": 0.020164920240538693,
      "grad_norm": 0.7414932231148509,
      "learning_rate": 3.96551724137931e-06,
      "loss": 0.1339,
      "step": 70
    },
    {
      "epoch": 0.021605271686291457,
      "grad_norm": 0.7014166187976457,
      "learning_rate": 4.252873563218391e-06,
      "loss": 0.1294,
      "step": 75
    },
    {
      "epoch": 0.02304562313204422,
      "grad_norm": 0.4877663944572015,
      "learning_rate": 4.540229885057471e-06,
      "loss": 0.1255,
      "step": 80
    },
    {
      "epoch": 0.02448597457779698,
      "grad_norm": 0.5776767722589294,
      "learning_rate": 4.8275862068965525e-06,
      "loss": 0.1214,
      "step": 85
    },
    {
      "epoch": 0.025926326023549745,
      "grad_norm": 0.6728066183845628,
      "learning_rate": 5.114942528735632e-06,
      "loss": 0.1171,
      "step": 90
    },
    {
      "epoch": 0.02736667746930251,
      "grad_norm": 0.45388962144154116,
      "learning_rate": 5.402298850574713e-06,
      "loss": 0.1189,
      "step": 95
    },
    {
      "epoch": 0.028807028915055273,
      "grad_norm": 0.4586145950809401,
      "learning_rate": 5.689655172413794e-06,
      "loss": 0.1083,
      "step": 100
    },
    {
      "epoch": 0.030247380360808037,
      "grad_norm": 0.5500926742810166,
      "learning_rate": 5.977011494252874e-06,
      "loss": 0.1041,
      "step": 105
    },
    {
      "epoch": 0.0316877318065608,
      "grad_norm": 0.45289151449461224,
      "learning_rate": 6.264367816091954e-06,
      "loss": 0.1005,
      "step": 110
    },
    {
      "epoch": 0.03312808325231356,
      "grad_norm": 1.8638824672613103,
      "learning_rate": 6.551724137931035e-06,
      "loss": 0.1004,
      "step": 115
    },
    {
      "epoch": 0.03456843469806633,
      "grad_norm": 0.4933743801153303,
      "learning_rate": 6.839080459770115e-06,
      "loss": 0.0994,
      "step": 120
    },
    {
      "epoch": 0.03600878614381909,
      "grad_norm": 0.44067596151303207,
      "learning_rate": 7.126436781609196e-06,
      "loss": 0.1001,
      "step": 125
    },
    {
      "epoch": 0.03744913758957186,
      "grad_norm": 0.6952160335299229,
      "learning_rate": 7.413793103448277e-06,
      "loss": 0.1015,
      "step": 130
    },
    {
      "epoch": 0.03888948903532462,
      "grad_norm": 0.48612969095240394,
      "learning_rate": 7.701149425287356e-06,
      "loss": 0.0957,
      "step": 135
    },
    {
      "epoch": 0.040329840481077385,
      "grad_norm": 0.4362186062208644,
      "learning_rate": 7.988505747126438e-06,
      "loss": 0.0949,
      "step": 140
    },
    {
      "epoch": 0.041770191926830146,
      "grad_norm": 0.5207325618446389,
      "learning_rate": 8.275862068965518e-06,
      "loss": 0.0912,
      "step": 145
    },
    {
      "epoch": 0.04321054337258291,
      "grad_norm": 0.3173162407799849,
      "learning_rate": 8.563218390804599e-06,
      "loss": 0.0897,
      "step": 150
    },
    {
      "epoch": 0.044650894818335674,
      "grad_norm": 0.40145266365610965,
      "learning_rate": 8.85057471264368e-06,
      "loss": 0.0904,
      "step": 155
    },
    {
      "epoch": 0.04609124626408844,
      "grad_norm": 0.3558474419241096,
      "learning_rate": 9.13793103448276e-06,
      "loss": 0.083,
      "step": 160
    },
    {
      "epoch": 0.0475315977098412,
      "grad_norm": 0.34959789012602455,
      "learning_rate": 9.42528735632184e-06,
      "loss": 0.0907,
      "step": 165
    },
    {
      "epoch": 0.04897194915559396,
      "grad_norm": 0.4329214768393735,
      "learning_rate": 9.71264367816092e-06,
      "loss": 0.0881,
      "step": 170
    },
    {
      "epoch": 0.05041230060134673,
      "grad_norm": 0.40181207129208235,
      "learning_rate": 1e-05,
      "loss": 0.0827,
      "step": 175
    },
    {
      "epoch": 0.05185265204709949,
      "grad_norm": 0.33262831961207967,
      "learning_rate": 1.0287356321839081e-05,
      "loss": 0.0856,
      "step": 180
    },
    {
      "epoch": 0.05329300349285226,
      "grad_norm": 0.4074048572040497,
      "learning_rate": 1.0574712643678162e-05,
      "loss": 0.0872,
      "step": 185
    },
    {
      "epoch": 0.05473335493860502,
      "grad_norm": 0.6269775026390093,
      "learning_rate": 1.0862068965517242e-05,
      "loss": 0.0872,
      "step": 190
    },
    {
      "epoch": 0.056173706384357786,
      "grad_norm": 0.4503563815568868,
      "learning_rate": 1.1149425287356324e-05,
      "loss": 0.0829,
      "step": 195
    },
    {
      "epoch": 0.057614057830110546,
      "grad_norm": 0.31295787939134934,
      "learning_rate": 1.1436781609195405e-05,
      "loss": 0.0882,
      "step": 200
    },
    {
      "epoch": 0.059054409275863314,
      "grad_norm": 0.4252598579032741,
      "learning_rate": 1.1724137931034483e-05,
      "loss": 0.0806,
      "step": 205
    },
    {
      "epoch": 0.060494760721616074,
      "grad_norm": 2.172621514603248,
      "learning_rate": 1.2011494252873564e-05,
      "loss": 0.0847,
      "step": 210
    },
    {
      "epoch": 0.061935112167368835,
      "grad_norm": 0.29858098908068476,
      "learning_rate": 1.2298850574712644e-05,
      "loss": 0.0814,
      "step": 215
    },
    {
      "epoch": 0.0633754636131216,
      "grad_norm": 0.30454141006022717,
      "learning_rate": 1.2586206896551725e-05,
      "loss": 0.0841,
      "step": 220
    },
    {
      "epoch": 0.06481581505887436,
      "grad_norm": 0.42812641240805016,
      "learning_rate": 1.2873563218390805e-05,
      "loss": 0.0822,
      "step": 225
    },
    {
      "epoch": 0.06625616650462712,
      "grad_norm": 0.4542902663928825,
      "learning_rate": 1.3160919540229885e-05,
      "loss": 0.0807,
      "step": 230
    },
    {
      "epoch": 0.0676965179503799,
      "grad_norm": 0.32671450237608535,
      "learning_rate": 1.3448275862068967e-05,
      "loss": 0.0842,
      "step": 235
    },
    {
      "epoch": 0.06913686939613266,
      "grad_norm": 0.3116704997352597,
      "learning_rate": 1.3735632183908048e-05,
      "loss": 0.0835,
      "step": 240
    },
    {
      "epoch": 0.07057722084188542,
      "grad_norm": 0.3849475487986608,
      "learning_rate": 1.4022988505747128e-05,
      "loss": 0.0807,
      "step": 245
    },
    {
      "epoch": 0.07201757228763818,
      "grad_norm": 0.28484088684959274,
      "learning_rate": 1.4310344827586209e-05,
      "loss": 0.0816,
      "step": 250
    },
    {
      "epoch": 0.07345792373339095,
      "grad_norm": 0.3337157396250414,
      "learning_rate": 1.459770114942529e-05,
      "loss": 0.0839,
      "step": 255
    },
    {
      "epoch": 0.07489827517914371,
      "grad_norm": 0.2664400550606567,
      "learning_rate": 1.4885057471264368e-05,
      "loss": 0.0817,
      "step": 260
    },
    {
      "epoch": 0.07633862662489647,
      "grad_norm": 0.2505155478711773,
      "learning_rate": 1.5172413793103448e-05,
      "loss": 0.0783,
      "step": 265
    },
    {
      "epoch": 0.07777897807064924,
      "grad_norm": 0.3335733996941115,
      "learning_rate": 1.545977011494253e-05,
      "loss": 0.0755,
      "step": 270
    },
    {
      "epoch": 0.079219329516402,
      "grad_norm": 0.26719459163424675,
      "learning_rate": 1.574712643678161e-05,
      "loss": 0.0789,
      "step": 275
    },
    {
      "epoch": 0.08065968096215477,
      "grad_norm": 0.5975638138236752,
      "learning_rate": 1.603448275862069e-05,
      "loss": 0.0811,
      "step": 280
    },
    {
      "epoch": 0.08210003240790753,
      "grad_norm": 0.3120443711700076,
      "learning_rate": 1.632183908045977e-05,
      "loss": 0.0783,
      "step": 285
    },
    {
      "epoch": 0.08354038385366029,
      "grad_norm": 0.5532978578404874,
      "learning_rate": 1.6609195402298854e-05,
      "loss": 0.0774,
      "step": 290
    },
    {
      "epoch": 0.08498073529941305,
      "grad_norm": 0.3019847180020645,
      "learning_rate": 1.6896551724137932e-05,
      "loss": 0.0795,
      "step": 295
    },
    {
      "epoch": 0.08642108674516583,
      "grad_norm": 0.4488280416931883,
      "learning_rate": 1.7183908045977015e-05,
      "loss": 0.0802,
      "step": 300
    },
    {
      "epoch": 0.08786143819091859,
      "grad_norm": 0.2713355446318922,
      "learning_rate": 1.7471264367816093e-05,
      "loss": 0.0825,
      "step": 305
    },
    {
      "epoch": 0.08930178963667135,
      "grad_norm": 0.2670783396564851,
      "learning_rate": 1.7758620689655175e-05,
      "loss": 0.0835,
      "step": 310
    },
    {
      "epoch": 0.09074214108242411,
      "grad_norm": 0.2708166986235282,
      "learning_rate": 1.8045977011494254e-05,
      "loss": 0.0743,
      "step": 315
    },
    {
      "epoch": 0.09218249252817688,
      "grad_norm": 0.23178489300951058,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.0812,
      "step": 320
    },
    {
      "epoch": 0.09362284397392964,
      "grad_norm": 0.2448448154224435,
      "learning_rate": 1.8620689655172415e-05,
      "loss": 0.078,
      "step": 325
    },
    {
      "epoch": 0.0950631954196824,
      "grad_norm": 0.2184469288572962,
      "learning_rate": 1.8908045977011497e-05,
      "loss": 0.076,
      "step": 330
    },
    {
      "epoch": 0.09650354686543516,
      "grad_norm": 0.19322313487222872,
      "learning_rate": 1.9195402298850576e-05,
      "loss": 0.0759,
      "step": 335
    },
    {
      "epoch": 0.09794389831118792,
      "grad_norm": 0.19526384677265263,
      "learning_rate": 1.9482758620689658e-05,
      "loss": 0.0751,
      "step": 340
    },
    {
      "epoch": 0.0993842497569407,
      "grad_norm": 0.4273996002750209,
      "learning_rate": 1.9770114942528737e-05,
      "loss": 0.0791,
      "step": 345
    },
    {
      "epoch": 0.10082460120269346,
      "grad_norm": 0.36910225249575607,
      "learning_rate": 1.9999994940288617e-05,
      "loss": 0.0819,
      "step": 350
    },
    {
      "epoch": 0.10226495264844622,
      "grad_norm": 0.22664069834453254,
      "learning_rate": 1.999981785092774e-05,
      "loss": 0.0715,
      "step": 355
    },
    {
      "epoch": 0.10370530409419898,
      "grad_norm": 0.22035293449009333,
      "learning_rate": 1.9999387781117715e-05,
      "loss": 0.0768,
      "step": 360
    },
    {
      "epoch": 0.10514565553995175,
      "grad_norm": 0.2409697392761486,
      "learning_rate": 1.9998704741738657e-05,
      "loss": 0.0785,
      "step": 365
    },
    {
      "epoch": 0.10658600698570452,
      "grad_norm": 0.19849583830438824,
      "learning_rate": 1.9997768750070442e-05,
      "loss": 0.0789,
      "step": 370
    },
    {
      "epoch": 0.10802635843145728,
      "grad_norm": 0.21065073891564184,
      "learning_rate": 1.9996579829792263e-05,
      "loss": 0.0741,
      "step": 375
    },
    {
      "epoch": 0.10946670987721004,
      "grad_norm": 0.2322217176695769,
      "learning_rate": 1.9995138010982028e-05,
      "loss": 0.0756,
      "step": 380
    },
    {
      "epoch": 0.1109070613229628,
      "grad_norm": 0.18082957951524414,
      "learning_rate": 1.9993443330115592e-05,
      "loss": 0.0755,
      "step": 385
    },
    {
      "epoch": 0.11234741276871557,
      "grad_norm": 0.2057441714111917,
      "learning_rate": 1.9991495830065857e-05,
      "loss": 0.077,
      "step": 390
    },
    {
      "epoch": 0.11378776421446833,
      "grad_norm": 0.1989077999547036,
      "learning_rate": 1.9989295560101656e-05,
      "loss": 0.0773,
      "step": 395
    },
    {
      "epoch": 0.11522811566022109,
      "grad_norm": 0.19081663464856954,
      "learning_rate": 1.998684257588654e-05,
      "loss": 0.0741,
      "step": 400
    },
    {
      "epoch": 0.11666846710597385,
      "grad_norm": 0.19247855517384488,
      "learning_rate": 1.9984136939477333e-05,
      "loss": 0.0757,
      "step": 405
    },
    {
      "epoch": 0.11810881855172663,
      "grad_norm": 0.19993453168502634,
      "learning_rate": 1.9981178719322606e-05,
      "loss": 0.0706,
      "step": 410
    },
    {
      "epoch": 0.11954916999747939,
      "grad_norm": 0.22409961064189973,
      "learning_rate": 1.99779679902609e-05,
      "loss": 0.0802,
      "step": 415
    },
    {
      "epoch": 0.12098952144323215,
      "grad_norm": 0.1850526119008749,
      "learning_rate": 1.9974504833518863e-05,
      "loss": 0.0778,
      "step": 420
    },
    {
      "epoch": 0.12242987288898491,
      "grad_norm": 0.20831229741416568,
      "learning_rate": 1.9970789336709185e-05,
      "loss": 0.0787,
      "step": 425
    },
    {
      "epoch": 0.12387022433473767,
      "grad_norm": 0.17198184608681197,
      "learning_rate": 1.9966821593828393e-05,
      "loss": 0.0758,
      "step": 430
    },
    {
      "epoch": 0.12531057578049043,
      "grad_norm": 0.18655586258401988,
      "learning_rate": 1.9962601705254442e-05,
      "loss": 0.0713,
      "step": 435
    },
    {
      "epoch": 0.1267509272262432,
      "grad_norm": 0.20225353346040226,
      "learning_rate": 1.995812977774421e-05,
      "loss": 0.076,
      "step": 440
    },
    {
      "epoch": 0.12819127867199598,
      "grad_norm": 0.2104521490426979,
      "learning_rate": 1.995340592443078e-05,
      "loss": 0.075,
      "step": 445
    },
    {
      "epoch": 0.12963163011774873,
      "grad_norm": 0.3236324710240072,
      "learning_rate": 1.9948430264820588e-05,
      "loss": 0.0752,
      "step": 450
    },
    {
      "epoch": 0.1310719815635015,
      "grad_norm": 0.1817007734256695,
      "learning_rate": 1.994320292479038e-05,
      "loss": 0.0714,
      "step": 455
    },
    {
      "epoch": 0.13251233300925425,
      "grad_norm": 0.21258018748179575,
      "learning_rate": 1.993772403658405e-05,
      "loss": 0.0741,
      "step": 460
    },
    {
      "epoch": 0.13395268445500702,
      "grad_norm": 0.1908524637842489,
      "learning_rate": 1.9931993738809288e-05,
      "loss": 0.0706,
      "step": 465
    },
    {
      "epoch": 0.1353930359007598,
      "grad_norm": 0.20920739812381056,
      "learning_rate": 1.9926012176434054e-05,
      "loss": 0.0677,
      "step": 470
    },
    {
      "epoch": 0.13683338734651254,
      "grad_norm": 0.1748700492849436,
      "learning_rate": 1.991977950078295e-05,
      "loss": 0.07,
      "step": 475
    },
    {
      "epoch": 0.13827373879226532,
      "grad_norm": 0.20542711837724692,
      "learning_rate": 1.9913295869533345e-05,
      "loss": 0.0717,
      "step": 480
    },
    {
      "epoch": 0.13971409023801806,
      "grad_norm": 0.1976538158422665,
      "learning_rate": 1.990656144671143e-05,
      "loss": 0.0755,
      "step": 485
    },
    {
      "epoch": 0.14115444168377084,
      "grad_norm": 0.18639103707023377,
      "learning_rate": 1.9899576402688038e-05,
      "loss": 0.0712,
      "step": 490
    },
    {
      "epoch": 0.1425947931295236,
      "grad_norm": 0.19411614042491165,
      "learning_rate": 1.9892340914174344e-05,
      "loss": 0.079,
      "step": 495
    },
    {
      "epoch": 0.14403514457527636,
      "grad_norm": 0.17878733422071091,
      "learning_rate": 1.988485516421739e-05,
      "loss": 0.0724,
      "step": 500
    },
    {
      "epoch": 0.14403514457527636,
      "eval_loss": 0.0738767758011818,
      "eval_runtime": 203.7076,
      "eval_samples_per_second": 8.856,
      "eval_steps_per_second": 2.214,
      "step": 500
    },
    {
      "epoch": 0.14547549602102913,
      "grad_norm": 0.20095627980982742,
      "learning_rate": 1.9877119342195478e-05,
      "loss": 0.0751,
      "step": 505
    },
    {
      "epoch": 0.1469158474667819,
      "grad_norm": 0.19127062420032984,
      "learning_rate": 1.986913364381333e-05,
      "loss": 0.0768,
      "step": 510
    },
    {
      "epoch": 0.14835619891253465,
      "grad_norm": 0.17411671796743106,
      "learning_rate": 1.9860898271097194e-05,
      "loss": 0.0721,
      "step": 515
    },
    {
      "epoch": 0.14979655035828743,
      "grad_norm": 0.20042970248260944,
      "learning_rate": 1.9852413432389685e-05,
      "loss": 0.0712,
      "step": 520
    },
    {
      "epoch": 0.15123690180404017,
      "grad_norm": 0.15880253535731764,
      "learning_rate": 1.984367934234455e-05,
      "loss": 0.0705,
      "step": 525
    },
    {
      "epoch": 0.15267725324979295,
      "grad_norm": 0.18651323384428312,
      "learning_rate": 1.9834696221921213e-05,
      "loss": 0.07,
      "step": 530
    },
    {
      "epoch": 0.15411760469554572,
      "grad_norm": 0.17659413165830717,
      "learning_rate": 1.98254642983792e-05,
      "loss": 0.0706,
      "step": 535
    },
    {
      "epoch": 0.15555795614129847,
      "grad_norm": 0.18160644848664637,
      "learning_rate": 1.9815983805272378e-05,
      "loss": 0.0683,
      "step": 540
    },
    {
      "epoch": 0.15699830758705124,
      "grad_norm": 0.19445538759505177,
      "learning_rate": 1.980625498244306e-05,
      "loss": 0.0764,
      "step": 545
    },
    {
      "epoch": 0.158438659032804,
      "grad_norm": 0.18455595535196012,
      "learning_rate": 1.9796278076015924e-05,
      "loss": 0.0697,
      "step": 550
    },
    {
      "epoch": 0.15987901047855677,
      "grad_norm": 0.18184833856061972,
      "learning_rate": 1.9786053338391792e-05,
      "loss": 0.0731,
      "step": 555
    },
    {
      "epoch": 0.16131936192430954,
      "grad_norm": 0.19338517000535457,
      "learning_rate": 1.9775581028241253e-05,
      "loss": 0.0663,
      "step": 560
    },
    {
      "epoch": 0.1627597133700623,
      "grad_norm": 0.19538801377440532,
      "learning_rate": 1.97648614104981e-05,
      "loss": 0.0675,
      "step": 565
    },
    {
      "epoch": 0.16420006481581506,
      "grad_norm": 0.19062085672060544,
      "learning_rate": 1.9753894756352643e-05,
      "loss": 0.0747,
      "step": 570
    },
    {
      "epoch": 0.16564041626156784,
      "grad_norm": 0.16777348688294874,
      "learning_rate": 1.9742681343244853e-05,
      "loss": 0.0667,
      "step": 575
    },
    {
      "epoch": 0.16708076770732058,
      "grad_norm": 0.170537203370038,
      "learning_rate": 1.9731221454857322e-05,
      "loss": 0.0716,
      "step": 580
    },
    {
      "epoch": 0.16852111915307336,
      "grad_norm": 0.18822505175377233,
      "learning_rate": 1.9719515381108093e-05,
      "loss": 0.0737,
      "step": 585
    },
    {
      "epoch": 0.1699614705988261,
      "grad_norm": 0.17674366415250964,
      "learning_rate": 1.970756341814335e-05,
      "loss": 0.0717,
      "step": 590
    },
    {
      "epoch": 0.17140182204457888,
      "grad_norm": 0.15690653895690873,
      "learning_rate": 1.9695365868329895e-05,
      "loss": 0.0721,
      "step": 595
    },
    {
      "epoch": 0.17284217349033165,
      "grad_norm": 0.18017195479881226,
      "learning_rate": 1.9682923040247513e-05,
      "loss": 0.071,
      "step": 600
    },
    {
      "epoch": 0.1742825249360844,
      "grad_norm": 0.18105083556119095,
      "learning_rate": 1.9670235248681154e-05,
      "loss": 0.069,
      "step": 605
    },
    {
      "epoch": 0.17572287638183717,
      "grad_norm": 0.21301510006977412,
      "learning_rate": 1.965730281461299e-05,
      "loss": 0.0728,
      "step": 610
    },
    {
      "epoch": 0.17716322782758992,
      "grad_norm": 0.20100310691523807,
      "learning_rate": 1.964412606521428e-05,
      "loss": 0.0706,
      "step": 615
    },
    {
      "epoch": 0.1786035792733427,
      "grad_norm": 0.19291740459291176,
      "learning_rate": 1.9630705333837096e-05,
      "loss": 0.0716,
      "step": 620
    },
    {
      "epoch": 0.18004393071909547,
      "grad_norm": 0.18688635532483766,
      "learning_rate": 1.9617040960005883e-05,
      "loss": 0.0707,
      "step": 625
    },
    {
      "epoch": 0.18148428216484822,
      "grad_norm": 0.1763269342109086,
      "learning_rate": 1.9603133289408883e-05,
      "loss": 0.0703,
      "step": 630
    },
    {
      "epoch": 0.182924633610601,
      "grad_norm": 0.1814521955212879,
      "learning_rate": 1.9588982673889373e-05,
      "loss": 0.0718,
      "step": 635
    },
    {
      "epoch": 0.18436498505635376,
      "grad_norm": 0.16733655331154948,
      "learning_rate": 1.9574589471436794e-05,
      "loss": 0.0686,
      "step": 640
    },
    {
      "epoch": 0.1858053365021065,
      "grad_norm": 0.17403974789227739,
      "learning_rate": 1.955995404617765e-05,
      "loss": 0.0723,
      "step": 645
    },
    {
      "epoch": 0.18724568794785929,
      "grad_norm": 0.18173738403202522,
      "learning_rate": 1.9545076768366336e-05,
      "loss": 0.0649,
      "step": 650
    },
    {
      "epoch": 0.18868603939361203,
      "grad_norm": 0.19621468099442035,
      "learning_rate": 1.9529958014375748e-05,
      "loss": 0.0723,
      "step": 655
    },
    {
      "epoch": 0.1901263908393648,
      "grad_norm": 0.167135269028582,
      "learning_rate": 1.9514598166687772e-05,
      "loss": 0.0747,
      "step": 660
    },
    {
      "epoch": 0.19156674228511758,
      "grad_norm": 0.17552296206320767,
      "learning_rate": 1.9498997613883597e-05,
      "loss": 0.0767,
      "step": 665
    },
    {
      "epoch": 0.19300709373087033,
      "grad_norm": 0.18484213292382873,
      "learning_rate": 1.9483156750633906e-05,
      "loss": 0.0754,
      "step": 670
    },
    {
      "epoch": 0.1944474451766231,
      "grad_norm": 0.22352663539188908,
      "learning_rate": 1.946707597768886e-05,
      "loss": 0.0678,
      "step": 675
    },
    {
      "epoch": 0.19588779662237585,
      "grad_norm": 0.19471837734529257,
      "learning_rate": 1.9450755701867994e-05,
      "loss": 0.0717,
      "step": 680
    },
    {
      "epoch": 0.19732814806812862,
      "grad_norm": 0.1638183568509793,
      "learning_rate": 1.9434196336049897e-05,
      "loss": 0.0655,
      "step": 685
    },
    {
      "epoch": 0.1987684995138814,
      "grad_norm": 0.20754326959948957,
      "learning_rate": 1.941739829916177e-05,
      "loss": 0.0692,
      "step": 690
    },
    {
      "epoch": 0.20020885095963414,
      "grad_norm": 0.18228455004715827,
      "learning_rate": 1.940036201616886e-05,
      "loss": 0.0706,
      "step": 695
    },
    {
      "epoch": 0.20164920240538692,
      "grad_norm": 0.19405238266885175,
      "learning_rate": 1.9383087918063662e-05,
      "loss": 0.073,
      "step": 700
    },
    {
      "epoch": 0.20308955385113966,
      "grad_norm": 0.17969147951103687,
      "learning_rate": 1.9365576441855046e-05,
      "loss": 0.0715,
      "step": 705
    },
    {
      "epoch": 0.20452990529689244,
      "grad_norm": 0.17634958619875035,
      "learning_rate": 1.9347828030557196e-05,
      "loss": 0.0695,
      "step": 710
    },
    {
      "epoch": 0.2059702567426452,
      "grad_norm": 0.17555207523689614,
      "learning_rate": 1.932984313317839e-05,
      "loss": 0.0726,
      "step": 715
    },
    {
      "epoch": 0.20741060818839796,
      "grad_norm": 0.18762117292247804,
      "learning_rate": 1.931162220470967e-05,
      "loss": 0.0697,
      "step": 720
    },
    {
      "epoch": 0.20885095963415073,
      "grad_norm": 0.18276137401754816,
      "learning_rate": 1.9293165706113287e-05,
      "loss": 0.0703,
      "step": 725
    },
    {
      "epoch": 0.2102913110799035,
      "grad_norm": 0.1773268424439162,
      "learning_rate": 1.9274474104311083e-05,
      "loss": 0.0689,
      "step": 730
    },
    {
      "epoch": 0.21173166252565626,
      "grad_norm": 0.25970999430399005,
      "learning_rate": 1.925554787217265e-05,
      "loss": 0.0706,
      "step": 735
    },
    {
      "epoch": 0.21317201397140903,
      "grad_norm": 0.19568404818114638,
      "learning_rate": 1.9236387488503378e-05,
      "loss": 0.0685,
      "step": 740
    },
    {
      "epoch": 0.21461236541716178,
      "grad_norm": 0.17278153132178845,
      "learning_rate": 1.921699343803235e-05,
      "loss": 0.0627,
      "step": 745
    },
    {
      "epoch": 0.21605271686291455,
      "grad_norm": 0.16307291332850227,
      "learning_rate": 1.9197366211400058e-05,
      "loss": 0.0716,
      "step": 750
    },
    {
      "epoch": 0.21749306830866733,
      "grad_norm": 0.16739005016913056,
      "learning_rate": 1.9177506305146008e-05,
      "loss": 0.0735,
      "step": 755
    },
    {
      "epoch": 0.21893341975442007,
      "grad_norm": 0.1652055715511055,
      "learning_rate": 1.9157414221696155e-05,
      "loss": 0.0695,
      "step": 760
    },
    {
      "epoch": 0.22037377120017285,
      "grad_norm": 0.1536676585476364,
      "learning_rate": 1.9137090469350185e-05,
      "loss": 0.0705,
      "step": 765
    },
    {
      "epoch": 0.2218141226459256,
      "grad_norm": 0.16180623315358525,
      "learning_rate": 1.9116535562268658e-05,
      "loss": 0.0686,
      "step": 770
    },
    {
      "epoch": 0.22325447409167837,
      "grad_norm": 0.16794475897997235,
      "learning_rate": 1.9095750020460017e-05,
      "loss": 0.0642,
      "step": 775
    },
    {
      "epoch": 0.22469482553743114,
      "grad_norm": 0.19565446028768624,
      "learning_rate": 1.9074734369767407e-05,
      "loss": 0.0679,
      "step": 780
    },
    {
      "epoch": 0.2261351769831839,
      "grad_norm": 0.20795674415023988,
      "learning_rate": 1.9053489141855386e-05,
      "loss": 0.0708,
      "step": 785
    },
    {
      "epoch": 0.22757552842893666,
      "grad_norm": 0.19615918609220573,
      "learning_rate": 1.9032014874196476e-05,
      "loss": 0.0748,
      "step": 790
    },
    {
      "epoch": 0.22901587987468944,
      "grad_norm": 0.19145581797389669,
      "learning_rate": 1.901031211005756e-05,
      "loss": 0.0693,
      "step": 795
    },
    {
      "epoch": 0.23045623132044218,
      "grad_norm": 0.18628621622706582,
      "learning_rate": 1.898838139848614e-05,
      "loss": 0.0691,
      "step": 800
    },
    {
      "epoch": 0.23189658276619496,
      "grad_norm": 0.1694305297665475,
      "learning_rate": 1.8966223294296445e-05,
      "loss": 0.0697,
      "step": 805
    },
    {
      "epoch": 0.2333369342119477,
      "grad_norm": 0.16668102270136734,
      "learning_rate": 1.8943838358055403e-05,
      "loss": 0.0756,
      "step": 810
    },
    {
      "epoch": 0.23477728565770048,
      "grad_norm": 0.16726519179537463,
      "learning_rate": 1.892122715606846e-05,
      "loss": 0.0703,
      "step": 815
    },
    {
      "epoch": 0.23621763710345325,
      "grad_norm": 0.15693138947405286,
      "learning_rate": 1.8898390260365227e-05,
      "loss": 0.0666,
      "step": 820
    },
    {
      "epoch": 0.237657988549206,
      "grad_norm": 0.15236512989059178,
      "learning_rate": 1.8875328248685047e-05,
      "loss": 0.0729,
      "step": 825
    },
    {
      "epoch": 0.23909833999495878,
      "grad_norm": 0.16512723377993288,
      "learning_rate": 1.885204170446235e-05,
      "loss": 0.073,
      "step": 830
    },
    {
      "epoch": 0.24053869144071152,
      "grad_norm": 0.8508290681028176,
      "learning_rate": 1.8828531216811912e-05,
      "loss": 0.0664,
      "step": 835
    },
    {
      "epoch": 0.2419790428864643,
      "grad_norm": 0.15837795426732162,
      "learning_rate": 1.8804797380513944e-05,
      "loss": 0.065,
      "step": 840
    },
    {
      "epoch": 0.24341939433221707,
      "grad_norm": 0.17566377373799896,
      "learning_rate": 1.878084079599903e-05,
      "loss": 0.0692,
      "step": 845
    },
    {
      "epoch": 0.24485974577796982,
      "grad_norm": 0.16777496534468336,
      "learning_rate": 1.8756662069332966e-05,
      "loss": 0.0697,
      "step": 850
    },
    {
      "epoch": 0.2463000972237226,
      "grad_norm": 0.15440366173279146,
      "learning_rate": 1.8732261812201408e-05,
      "loss": 0.0662,
      "step": 855
    },
    {
      "epoch": 0.24774044866947534,
      "grad_norm": 0.2354593017395783,
      "learning_rate": 1.8707640641894395e-05,
      "loss": 0.0677,
      "step": 860
    },
    {
      "epoch": 0.2491808001152281,
      "grad_norm": 0.18206050600161866,
      "learning_rate": 1.8682799181290747e-05,
      "loss": 0.0682,
      "step": 865
    },
    {
      "epoch": 0.25062115156098086,
      "grad_norm": 0.1878723763548132,
      "learning_rate": 1.86577380588423e-05,
      "loss": 0.0672,
      "step": 870
    },
    {
      "epoch": 0.25206150300673363,
      "grad_norm": 0.17948035203695148,
      "learning_rate": 1.8632457908558006e-05,
      "loss": 0.0656,
      "step": 875
    },
    {
      "epoch": 0.2535018544524864,
      "grad_norm": 0.19008061284514347,
      "learning_rate": 1.8606959369987885e-05,
      "loss": 0.075,
      "step": 880
    },
    {
      "epoch": 0.2549422058982392,
      "grad_norm": 0.1698902353596563,
      "learning_rate": 1.8581243088206865e-05,
      "loss": 0.0731,
      "step": 885
    },
    {
      "epoch": 0.25638255734399196,
      "grad_norm": 0.1851448703226216,
      "learning_rate": 1.8555309713798445e-05,
      "loss": 0.0678,
      "step": 890
    },
    {
      "epoch": 0.2578229087897447,
      "grad_norm": 0.1722062227581134,
      "learning_rate": 1.8529159902838253e-05,
      "loss": 0.066,
      "step": 895
    },
    {
      "epoch": 0.25926326023549745,
      "grad_norm": 0.15672701800283906,
      "learning_rate": 1.8502794316877423e-05,
      "loss": 0.0663,
      "step": 900
    },
    {
      "epoch": 0.2607036116812502,
      "grad_norm": 0.2613574893910426,
      "learning_rate": 1.8476213622925885e-05,
      "loss": 0.0663,
      "step": 905
    },
    {
      "epoch": 0.262143963127003,
      "grad_norm": 0.5103888927452046,
      "learning_rate": 1.844941849343548e-05,
      "loss": 0.0675,
      "step": 910
    },
    {
      "epoch": 0.2635843145727558,
      "grad_norm": 0.2024265148274089,
      "learning_rate": 1.842240960628294e-05,
      "loss": 0.0691,
      "step": 915
    },
    {
      "epoch": 0.2650246660185085,
      "grad_norm": 0.1938252123776783,
      "learning_rate": 1.8395187644752756e-05,
      "loss": 0.0645,
      "step": 920
    },
    {
      "epoch": 0.26646501746426127,
      "grad_norm": 0.17332839893721164,
      "learning_rate": 1.8367753297519874e-05,
      "loss": 0.0656,
      "step": 925
    },
    {
      "epoch": 0.26790536891001404,
      "grad_norm": 0.16271679417725463,
      "learning_rate": 1.8340107258632288e-05,
      "loss": 0.0698,
      "step": 930
    },
    {
      "epoch": 0.2693457203557668,
      "grad_norm": 0.1948182360731314,
      "learning_rate": 1.831225022749347e-05,
      "loss": 0.0647,
      "step": 935
    },
    {
      "epoch": 0.2707860718015196,
      "grad_norm": 0.18497254720455072,
      "learning_rate": 1.828418290884468e-05,
      "loss": 0.0669,
      "step": 940
    },
    {
      "epoch": 0.2722264232472723,
      "grad_norm": 0.17841388729576735,
      "learning_rate": 1.8255906012747137e-05,
      "loss": 0.0634,
      "step": 945
    },
    {
      "epoch": 0.2736667746930251,
      "grad_norm": 0.17987207372942032,
      "learning_rate": 1.8227420254564066e-05,
      "loss": 0.062,
      "step": 950
    },
    {
      "epoch": 0.27510712613877786,
      "grad_norm": 0.15594487078267172,
      "learning_rate": 1.819872635494258e-05,
      "loss": 0.0662,
      "step": 955
    },
    {
      "epoch": 0.27654747758453063,
      "grad_norm": 0.16476402091196268,
      "learning_rate": 1.816982503979546e-05,
      "loss": 0.0698,
      "step": 960
    },
    {
      "epoch": 0.2779878290302834,
      "grad_norm": 0.17955556471513223,
      "learning_rate": 1.8140717040282797e-05,
      "loss": 0.0653,
      "step": 965
    },
    {
      "epoch": 0.2794281804760361,
      "grad_norm": 0.23723396239220879,
      "learning_rate": 1.811140309279348e-05,
      "loss": 0.0686,
      "step": 970
    },
    {
      "epoch": 0.2808685319217889,
      "grad_norm": 0.18799932816101794,
      "learning_rate": 1.808188393892658e-05,
      "loss": 0.0689,
      "step": 975
    },
    {
      "epoch": 0.2823088833675417,
      "grad_norm": 0.1730477583895621,
      "learning_rate": 1.805216032547258e-05,
      "loss": 0.0689,
      "step": 980
    },
    {
      "epoch": 0.28374923481329445,
      "grad_norm": 0.17077954636100237,
      "learning_rate": 1.8022233004394487e-05,
      "loss": 0.0705,
      "step": 985
    },
    {
      "epoch": 0.2851895862590472,
      "grad_norm": 0.2007911822516418,
      "learning_rate": 1.7992102732808798e-05,
      "loss": 0.0689,
      "step": 990
    },
    {
      "epoch": 0.2866299377048,
      "grad_norm": 0.19728704564899405,
      "learning_rate": 1.796177027296637e-05,
      "loss": 0.0706,
      "step": 995
    },
    {
      "epoch": 0.2880702891505527,
      "grad_norm": 0.17858125265730437,
      "learning_rate": 1.79312363922331e-05,
      "loss": 0.07,
      "step": 1000
    },
    {
      "epoch": 0.2880702891505527,
      "eval_loss": 0.0674939900636673,
      "eval_runtime": 203.1313,
      "eval_samples_per_second": 8.881,
      "eval_steps_per_second": 2.22,
      "step": 1000
    },
    {
      "epoch": 0.2895106405963055,
      "grad_norm": 0.16535683477861135,
      "learning_rate": 1.7900501863070552e-05,
      "loss": 0.0635,
      "step": 1005
    },
    {
      "epoch": 0.29095099204205827,
      "grad_norm": 0.17493397556340032,
      "learning_rate": 1.7869567463016394e-05,
      "loss": 0.0626,
      "step": 1010
    },
    {
      "epoch": 0.29239134348781104,
      "grad_norm": 0.18719256822064576,
      "learning_rate": 1.7838433974664714e-05,
      "loss": 0.0711,
      "step": 1015
    },
    {
      "epoch": 0.2938316949335638,
      "grad_norm": 0.17416319375618142,
      "learning_rate": 1.7807102185646247e-05,
      "loss": 0.0686,
      "step": 1020
    },
    {
      "epoch": 0.29527204637931653,
      "grad_norm": 0.18401267882699013,
      "learning_rate": 1.7775572888608438e-05,
      "loss": 0.0669,
      "step": 1025
    },
    {
      "epoch": 0.2967123978250693,
      "grad_norm": 0.18105211685459255,
      "learning_rate": 1.774384688119539e-05,
      "loss": 0.0714,
      "step": 1030
    },
    {
      "epoch": 0.2981527492708221,
      "grad_norm": 0.15402578107011958,
      "learning_rate": 1.7711924966027678e-05,
      "loss": 0.0662,
      "step": 1035
    },
    {
      "epoch": 0.29959310071657486,
      "grad_norm": 0.18966226185659374,
      "learning_rate": 1.767980795068206e-05,
      "loss": 0.0715,
      "step": 1040
    },
    {
      "epoch": 0.30103345216232763,
      "grad_norm": 0.22533074590736946,
      "learning_rate": 1.7647496647671033e-05,
      "loss": 0.0678,
      "step": 1045
    },
    {
      "epoch": 0.30247380360808035,
      "grad_norm": 0.14978667904322185,
      "learning_rate": 1.761499187442228e-05,
      "loss": 0.0602,
      "step": 1050
    },
    {
      "epoch": 0.3039141550538331,
      "grad_norm": 0.16009119628660812,
      "learning_rate": 1.7582294453257996e-05,
      "loss": 0.0737,
      "step": 1055
    },
    {
      "epoch": 0.3053545064995859,
      "grad_norm": 0.1547831176643596,
      "learning_rate": 1.7549405211374072e-05,
      "loss": 0.0686,
      "step": 1060
    },
    {
      "epoch": 0.3067948579453387,
      "grad_norm": 0.17560213727576401,
      "learning_rate": 1.7516324980819185e-05,
      "loss": 0.0651,
      "step": 1065
    },
    {
      "epoch": 0.30823520939109145,
      "grad_norm": 0.1667942680628308,
      "learning_rate": 1.7483054598473734e-05,
      "loss": 0.0731,
      "step": 1070
    },
    {
      "epoch": 0.30967556083684417,
      "grad_norm": 0.18465697565580333,
      "learning_rate": 1.7449594906028684e-05,
      "loss": 0.067,
      "step": 1075
    },
    {
      "epoch": 0.31111591228259694,
      "grad_norm": 0.16066322224426005,
      "learning_rate": 1.7415946749964252e-05,
      "loss": 0.0667,
      "step": 1080
    },
    {
      "epoch": 0.3125562637283497,
      "grad_norm": 0.158008871885263,
      "learning_rate": 1.7382110981528506e-05,
      "loss": 0.0643,
      "step": 1085
    },
    {
      "epoch": 0.3139966151741025,
      "grad_norm": 0.18023373354288208,
      "learning_rate": 1.734808845671583e-05,
      "loss": 0.0603,
      "step": 1090
    },
    {
      "epoch": 0.31543696661985526,
      "grad_norm": 0.16286673700727694,
      "learning_rate": 1.7313880036245257e-05,
      "loss": 0.0649,
      "step": 1095
    },
    {
      "epoch": 0.316877318065608,
      "grad_norm": 0.1673107111388145,
      "learning_rate": 1.7279486585538712e-05,
      "loss": 0.067,
      "step": 1100
    },
    {
      "epoch": 0.31831766951136076,
      "grad_norm": 0.180435927792874,
      "learning_rate": 1.7244908974699112e-05,
      "loss": 0.0683,
      "step": 1105
    },
    {
      "epoch": 0.31975802095711353,
      "grad_norm": 0.15646264704364563,
      "learning_rate": 1.721014807848833e-05,
      "loss": 0.0698,
      "step": 1110
    },
    {
      "epoch": 0.3211983724028663,
      "grad_norm": 0.19141062322094995,
      "learning_rate": 1.7175204776305102e-05,
      "loss": 0.0634,
      "step": 1115
    },
    {
      "epoch": 0.3226387238486191,
      "grad_norm": 0.1539348164012501,
      "learning_rate": 1.7140079952162765e-05,
      "loss": 0.0649,
      "step": 1120
    },
    {
      "epoch": 0.32407907529437185,
      "grad_norm": 0.16127336339051726,
      "learning_rate": 1.7104774494666877e-05,
      "loss": 0.0631,
      "step": 1125
    },
    {
      "epoch": 0.3255194267401246,
      "grad_norm": 0.1541424306237969,
      "learning_rate": 1.7069289296992756e-05,
      "loss": 0.0667,
      "step": 1130
    },
    {
      "epoch": 0.32695977818587735,
      "grad_norm": 0.15026114726792159,
      "learning_rate": 1.703362525686288e-05,
      "loss": 0.0648,
      "step": 1135
    },
    {
      "epoch": 0.3284001296316301,
      "grad_norm": 0.15367694481808417,
      "learning_rate": 1.6997783276524177e-05,
      "loss": 0.0678,
      "step": 1140
    },
    {
      "epoch": 0.3298404810773829,
      "grad_norm": 0.14648710292293746,
      "learning_rate": 1.6961764262725187e-05,
      "loss": 0.0632,
      "step": 1145
    },
    {
      "epoch": 0.33128083252313567,
      "grad_norm": 0.15611228592369236,
      "learning_rate": 1.6925569126693135e-05,
      "loss": 0.0604,
      "step": 1150
    },
    {
      "epoch": 0.3327211839688884,
      "grad_norm": 0.16896274297951375,
      "learning_rate": 1.6889198784110883e-05,
      "loss": 0.0617,
      "step": 1155
    },
    {
      "epoch": 0.33416153541464116,
      "grad_norm": 0.14046862243355399,
      "learning_rate": 1.6852654155093745e-05,
      "loss": 0.0667,
      "step": 1160
    },
    {
      "epoch": 0.33560188686039394,
      "grad_norm": 0.16033306567178615,
      "learning_rate": 1.681593616416623e-05,
      "loss": 0.0667,
      "step": 1165
    },
    {
      "epoch": 0.3370422383061467,
      "grad_norm": 0.1464089660272697,
      "learning_rate": 1.6779045740238643e-05,
      "loss": 0.0689,
      "step": 1170
    },
    {
      "epoch": 0.3384825897518995,
      "grad_norm": 0.14637802553541343,
      "learning_rate": 1.6741983816583583e-05,
      "loss": 0.0644,
      "step": 1175
    },
    {
      "epoch": 0.3399229411976522,
      "grad_norm": 0.14849896211371577,
      "learning_rate": 1.6704751330812342e-05,
      "loss": 0.0593,
      "step": 1180
    },
    {
      "epoch": 0.341363292643405,
      "grad_norm": 0.16869632868201492,
      "learning_rate": 1.666734922485117e-05,
      "loss": 0.0651,
      "step": 1185
    },
    {
      "epoch": 0.34280364408915776,
      "grad_norm": 0.16979103264214,
      "learning_rate": 1.662977844491746e-05,
      "loss": 0.0681,
      "step": 1190
    },
    {
      "epoch": 0.34424399553491053,
      "grad_norm": 0.1485954334355779,
      "learning_rate": 1.6592039941495803e-05,
      "loss": 0.0608,
      "step": 1195
    },
    {
      "epoch": 0.3456843469806633,
      "grad_norm": 0.14777980222615944,
      "learning_rate": 1.6554134669313943e-05,
      "loss": 0.0676,
      "step": 1200
    },
    {
      "epoch": 0.347124698426416,
      "grad_norm": 0.1509175328811579,
      "learning_rate": 1.6516063587318627e-05,
      "loss": 0.0629,
      "step": 1205
    },
    {
      "epoch": 0.3485650498721688,
      "grad_norm": 0.16676709581052593,
      "learning_rate": 1.647782765865134e-05,
      "loss": 0.0619,
      "step": 1210
    },
    {
      "epoch": 0.35000540131792157,
      "grad_norm": 0.15838480213664755,
      "learning_rate": 1.6439427850623944e-05,
      "loss": 0.067,
      "step": 1215
    },
    {
      "epoch": 0.35144575276367435,
      "grad_norm": 0.14839184568071354,
      "learning_rate": 1.64008651346942e-05,
      "loss": 0.0642,
      "step": 1220
    },
    {
      "epoch": 0.3528861042094271,
      "grad_norm": 0.1648207628445432,
      "learning_rate": 1.63621404864412e-05,
      "loss": 0.0658,
      "step": 1225
    },
    {
      "epoch": 0.35432645565517984,
      "grad_norm": 0.1591196777066707,
      "learning_rate": 1.6323254885540672e-05,
      "loss": 0.068,
      "step": 1230
    },
    {
      "epoch": 0.3557668071009326,
      "grad_norm": 0.14084965766747376,
      "learning_rate": 1.6284209315740225e-05,
      "loss": 0.0617,
      "step": 1235
    },
    {
      "epoch": 0.3572071585466854,
      "grad_norm": 0.15809745947459444,
      "learning_rate": 1.6245004764834423e-05,
      "loss": 0.0657,
      "step": 1240
    },
    {
      "epoch": 0.35864750999243816,
      "grad_norm": 0.15253088492537956,
      "learning_rate": 1.620564222463982e-05,
      "loss": 0.065,
      "step": 1245
    },
    {
      "epoch": 0.36008786143819094,
      "grad_norm": 0.14590704135450994,
      "learning_rate": 1.6166122690969872e-05,
      "loss": 0.0618,
      "step": 1250
    },
    {
      "epoch": 0.36152821288394366,
      "grad_norm": 0.13085111936326246,
      "learning_rate": 1.612644716360972e-05,
      "loss": 0.0568,
      "step": 1255
    },
    {
      "epoch": 0.36296856432969643,
      "grad_norm": 0.15489937634073334,
      "learning_rate": 1.6086616646290926e-05,
      "loss": 0.0704,
      "step": 1260
    },
    {
      "epoch": 0.3644089157754492,
      "grad_norm": 0.15995869543145932,
      "learning_rate": 1.6046632146666056e-05,
      "loss": 0.0639,
      "step": 1265
    },
    {
      "epoch": 0.365849267221202,
      "grad_norm": 0.14803784398219583,
      "learning_rate": 1.60064946762832e-05,
      "loss": 0.063,
      "step": 1270
    },
    {
      "epoch": 0.36728961866695475,
      "grad_norm": 0.1482802413474464,
      "learning_rate": 1.5966205250560393e-05,
      "loss": 0.0667,
      "step": 1275
    },
    {
      "epoch": 0.36872997011270753,
      "grad_norm": 0.14080015648710145,
      "learning_rate": 1.592576488875989e-05,
      "loss": 0.0657,
      "step": 1280
    },
    {
      "epoch": 0.37017032155846025,
      "grad_norm": 0.14023456289860875,
      "learning_rate": 1.5885174613962427e-05,
      "loss": 0.0661,
      "step": 1285
    },
    {
      "epoch": 0.371610673004213,
      "grad_norm": 0.1609532074245431,
      "learning_rate": 1.5844435453041294e-05,
      "loss": 0.0675,
      "step": 1290
    },
    {
      "epoch": 0.3730510244499658,
      "grad_norm": 0.13842067239979322,
      "learning_rate": 1.5803548436636394e-05,
      "loss": 0.0655,
      "step": 1295
    },
    {
      "epoch": 0.37449137589571857,
      "grad_norm": 0.16642819320304691,
      "learning_rate": 1.576251459912814e-05,
      "loss": 0.069,
      "step": 1300
    },
    {
      "epoch": 0.37593172734147134,
      "grad_norm": 0.15113047288834136,
      "learning_rate": 1.5721334978611307e-05,
      "loss": 0.0673,
      "step": 1305
    },
    {
      "epoch": 0.37737207878722406,
      "grad_norm": 0.13337077100229278,
      "learning_rate": 1.5680010616868762e-05,
      "loss": 0.0635,
      "step": 1310
    },
    {
      "epoch": 0.37881243023297684,
      "grad_norm": 0.14914710569165873,
      "learning_rate": 1.5638542559345106e-05,
      "loss": 0.0619,
      "step": 1315
    },
    {
      "epoch": 0.3802527816787296,
      "grad_norm": 0.1716264208900402,
      "learning_rate": 1.559693185512023e-05,
      "loss": 0.0671,
      "step": 1320
    },
    {
      "epoch": 0.3816931331244824,
      "grad_norm": 0.14840365175842551,
      "learning_rate": 1.555517955688277e-05,
      "loss": 0.0633,
      "step": 1325
    },
    {
      "epoch": 0.38313348457023516,
      "grad_norm": 0.16640813297411597,
      "learning_rate": 1.5513286720903488e-05,
      "loss": 0.0707,
      "step": 1330
    },
    {
      "epoch": 0.3845738360159879,
      "grad_norm": 0.14134033460133827,
      "learning_rate": 1.5471254407008526e-05,
      "loss": 0.0611,
      "step": 1335
    },
    {
      "epoch": 0.38601418746174065,
      "grad_norm": 0.1959632020773705,
      "learning_rate": 1.542908367855263e-05,
      "loss": 0.066,
      "step": 1340
    },
    {
      "epoch": 0.38745453890749343,
      "grad_norm": 0.15059928506128456,
      "learning_rate": 1.53867756023922e-05,
      "loss": 0.0626,
      "step": 1345
    },
    {
      "epoch": 0.3888948903532462,
      "grad_norm": 0.14705027197162424,
      "learning_rate": 1.534433124885836e-05,
      "loss": 0.0612,
      "step": 1350
    },
    {
      "epoch": 0.390335241798999,
      "grad_norm": 0.1712504949998861,
      "learning_rate": 1.530175169172982e-05,
      "loss": 0.0652,
      "step": 1355
    },
    {
      "epoch": 0.3917755932447517,
      "grad_norm": 0.16428240182950052,
      "learning_rate": 1.525903800820575e-05,
      "loss": 0.0638,
      "step": 1360
    },
    {
      "epoch": 0.39321594469050447,
      "grad_norm": 0.16329645975133253,
      "learning_rate": 1.5216191278878522e-05,
      "loss": 0.0636,
      "step": 1365
    },
    {
      "epoch": 0.39465629613625725,
      "grad_norm": 0.13480884471675694,
      "learning_rate": 1.517321258770636e-05,
      "loss": 0.0613,
      "step": 1370
    },
    {
      "epoch": 0.39609664758201,
      "grad_norm": 0.1669515249018309,
      "learning_rate": 1.5130103021985929e-05,
      "loss": 0.0678,
      "step": 1375
    },
    {
      "epoch": 0.3975369990277628,
      "grad_norm": 0.15183706654976248,
      "learning_rate": 1.5086863672324826e-05,
      "loss": 0.0614,
      "step": 1380
    },
    {
      "epoch": 0.3989773504735155,
      "grad_norm": 0.16444568652959443,
      "learning_rate": 1.5043495632613982e-05,
      "loss": 0.0639,
      "step": 1385
    },
    {
      "epoch": 0.4004177019192683,
      "grad_norm": 1.6562960432751872,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.0665,
      "step": 1390
    },
    {
      "epoch": 0.40185805336502106,
      "grad_norm": 0.14191330098363916,
      "learning_rate": 1.4956377874857395e-05,
      "loss": 0.0595,
      "step": 1395
    },
    {
      "epoch": 0.40329840481077384,
      "grad_norm": 0.14989109086985877,
      "learning_rate": 1.4912630360760743e-05,
      "loss": 0.0653,
      "step": 1400
    },
    {
      "epoch": 0.4047387562565266,
      "grad_norm": 0.15488669647541722,
      "learning_rate": 1.4868758564456785e-05,
      "loss": 0.0674,
      "step": 1405
    },
    {
      "epoch": 0.40617910770227933,
      "grad_norm": 0.13853389542189282,
      "learning_rate": 1.4824763595836404e-05,
      "loss": 0.0637,
      "step": 1410
    },
    {
      "epoch": 0.4076194591480321,
      "grad_norm": 0.157569240556149,
      "learning_rate": 1.4780646567906571e-05,
      "loss": 0.0605,
      "step": 1415
    },
    {
      "epoch": 0.4090598105937849,
      "grad_norm": 0.16007008642510542,
      "learning_rate": 1.473640859676217e-05,
      "loss": 0.0632,
      "step": 1420
    },
    {
      "epoch": 0.41050016203953765,
      "grad_norm": 0.15116913790513345,
      "learning_rate": 1.4692050801557769e-05,
      "loss": 0.0632,
      "step": 1425
    },
    {
      "epoch": 0.4119405134852904,
      "grad_norm": 0.14587678441383706,
      "learning_rate": 1.4647574304479295e-05,
      "loss": 0.0675,
      "step": 1430
    },
    {
      "epoch": 0.4133808649310432,
      "grad_norm": 0.1518975910563791,
      "learning_rate": 1.4602980230715674e-05,
      "loss": 0.0655,
      "step": 1435
    },
    {
      "epoch": 0.4148212163767959,
      "grad_norm": 0.14736642886138512,
      "learning_rate": 1.4558269708430333e-05,
      "loss": 0.0628,
      "step": 1440
    },
    {
      "epoch": 0.4162615678225487,
      "grad_norm": 0.16686599891471496,
      "learning_rate": 1.4513443868732674e-05,
      "loss": 0.0667,
      "step": 1445
    },
    {
      "epoch": 0.41770191926830147,
      "grad_norm": 0.1409423422519682,
      "learning_rate": 1.4468503845649462e-05,
      "loss": 0.0644,
      "step": 1450
    },
    {
      "epoch": 0.41914227071405424,
      "grad_norm": 0.14764702695382043,
      "learning_rate": 1.4423450776096122e-05,
      "loss": 0.0612,
      "step": 1455
    },
    {
      "epoch": 0.420582622159807,
      "grad_norm": 0.14452959429650908,
      "learning_rate": 1.4378285799848004e-05,
      "loss": 0.0677,
      "step": 1460
    },
    {
      "epoch": 0.42202297360555974,
      "grad_norm": 0.1573610389293893,
      "learning_rate": 1.4333010059511505e-05,
      "loss": 0.062,
      "step": 1465
    },
    {
      "epoch": 0.4234633250513125,
      "grad_norm": 0.1480919971980055,
      "learning_rate": 1.4287624700495211e-05,
      "loss": 0.0619,
      "step": 1470
    },
    {
      "epoch": 0.4249036764970653,
      "grad_norm": 0.15894026434712646,
      "learning_rate": 1.4242130870980882e-05,
      "loss": 0.0611,
      "step": 1475
    },
    {
      "epoch": 0.42634402794281806,
      "grad_norm": 0.15069243728870851,
      "learning_rate": 1.4196529721894427e-05,
      "loss": 0.064,
      "step": 1480
    },
    {
      "epoch": 0.42778437938857083,
      "grad_norm": 0.1682063987256438,
      "learning_rate": 1.4150822406876774e-05,
      "loss": 0.0639,
      "step": 1485
    },
    {
      "epoch": 0.42922473083432355,
      "grad_norm": 0.15006120941638998,
      "learning_rate": 1.4105010082254697e-05,
      "loss": 0.0665,
      "step": 1490
    },
    {
      "epoch": 0.43066508228007633,
      "grad_norm": 0.13479491661214907,
      "learning_rate": 1.4059093907011552e-05,
      "loss": 0.0617,
      "step": 1495
    },
    {
      "epoch": 0.4321054337258291,
      "grad_norm": 0.1715681331105154,
      "learning_rate": 1.401307504275796e-05,
      "loss": 0.0638,
      "step": 1500
    },
    {
      "epoch": 0.4321054337258291,
      "eval_loss": 0.06420264393091202,
      "eval_runtime": 204.1795,
      "eval_samples_per_second": 8.835,
      "eval_steps_per_second": 2.209,
      "step": 1500
    },
    {
      "epoch": 0.4335457851715819,
      "grad_norm": 0.1519666716391639,
      "learning_rate": 1.3966954653702423e-05,
      "loss": 0.0648,
      "step": 1505
    },
    {
      "epoch": 0.43498613661733465,
      "grad_norm": 0.14120235921333318,
      "learning_rate": 1.3920733906621861e-05,
      "loss": 0.0625,
      "step": 1510
    },
    {
      "epoch": 0.43642648806308737,
      "grad_norm": 0.15467790310945892,
      "learning_rate": 1.3874413970832123e-05,
      "loss": 0.064,
      "step": 1515
    },
    {
      "epoch": 0.43786683950884014,
      "grad_norm": 0.13984061839469614,
      "learning_rate": 1.3827996018158356e-05,
      "loss": 0.0605,
      "step": 1520
    },
    {
      "epoch": 0.4393071909545929,
      "grad_norm": 0.145852830960354,
      "learning_rate": 1.378148122290541e-05,
      "loss": 0.0655,
      "step": 1525
    },
    {
      "epoch": 0.4407475424003457,
      "grad_norm": 0.15286178112634702,
      "learning_rate": 1.3734870761828095e-05,
      "loss": 0.063,
      "step": 1530
    },
    {
      "epoch": 0.44218789384609847,
      "grad_norm": 0.1259636985152776,
      "learning_rate": 1.368816581410143e-05,
      "loss": 0.0615,
      "step": 1535
    },
    {
      "epoch": 0.4436282452918512,
      "grad_norm": 0.15476898913684597,
      "learning_rate": 1.3641367561290795e-05,
      "loss": 0.0644,
      "step": 1540
    },
    {
      "epoch": 0.44506859673760396,
      "grad_norm": 0.1617280661787022,
      "learning_rate": 1.3594477187322065e-05,
      "loss": 0.0614,
      "step": 1545
    },
    {
      "epoch": 0.44650894818335674,
      "grad_norm": 0.14998584782237406,
      "learning_rate": 1.3547495878451635e-05,
      "loss": 0.0635,
      "step": 1550
    },
    {
      "epoch": 0.4479492996291095,
      "grad_norm": 0.15207416735601922,
      "learning_rate": 1.3500424823236413e-05,
      "loss": 0.0669,
      "step": 1555
    },
    {
      "epoch": 0.4493896510748623,
      "grad_norm": 0.13476067626565744,
      "learning_rate": 1.3453265212503756e-05,
      "loss": 0.0606,
      "step": 1560
    },
    {
      "epoch": 0.450830002520615,
      "grad_norm": 0.15284821669337595,
      "learning_rate": 1.340601823932135e-05,
      "loss": 0.0624,
      "step": 1565
    },
    {
      "epoch": 0.4522703539663678,
      "grad_norm": 0.15463620126735736,
      "learning_rate": 1.335868509896702e-05,
      "loss": 0.0651,
      "step": 1570
    },
    {
      "epoch": 0.45371070541212055,
      "grad_norm": 0.15250686045319206,
      "learning_rate": 1.3311266988898477e-05,
      "loss": 0.0646,
      "step": 1575
    },
    {
      "epoch": 0.4551510568578733,
      "grad_norm": 0.13859935456186184,
      "learning_rate": 1.3263765108723061e-05,
      "loss": 0.0623,
      "step": 1580
    },
    {
      "epoch": 0.4565914083036261,
      "grad_norm": 0.1614173179165357,
      "learning_rate": 1.3216180660167355e-05,
      "loss": 0.0611,
      "step": 1585
    },
    {
      "epoch": 0.4580317597493789,
      "grad_norm": 0.16964451095340738,
      "learning_rate": 1.3168514847046802e-05,
      "loss": 0.0612,
      "step": 1590
    },
    {
      "epoch": 0.4594721111951316,
      "grad_norm": 0.1497788074170894,
      "learning_rate": 1.3120768875235252e-05,
      "loss": 0.0593,
      "step": 1595
    },
    {
      "epoch": 0.46091246264088437,
      "grad_norm": 0.12383289401240005,
      "learning_rate": 1.3072943952634446e-05,
      "loss": 0.0592,
      "step": 1600
    },
    {
      "epoch": 0.46235281408663714,
      "grad_norm": 0.146818726530369,
      "learning_rate": 1.3025041289143459e-05,
      "loss": 0.0605,
      "step": 1605
    },
    {
      "epoch": 0.4637931655323899,
      "grad_norm": 0.14306567032337267,
      "learning_rate": 1.2977062096628096e-05,
      "loss": 0.0605,
      "step": 1610
    },
    {
      "epoch": 0.4652335169781427,
      "grad_norm": 0.14605676143356022,
      "learning_rate": 1.2929007588890241e-05,
      "loss": 0.0612,
      "step": 1615
    },
    {
      "epoch": 0.4666738684238954,
      "grad_norm": 0.15086264411486566,
      "learning_rate": 1.2880878981637129e-05,
      "loss": 0.063,
      "step": 1620
    },
    {
      "epoch": 0.4681142198696482,
      "grad_norm": 0.15203796445266168,
      "learning_rate": 1.2832677492450602e-05,
      "loss": 0.0635,
      "step": 1625
    },
    {
      "epoch": 0.46955457131540096,
      "grad_norm": 0.1644232831015843,
      "learning_rate": 1.2784404340756315e-05,
      "loss": 0.0656,
      "step": 1630
    },
    {
      "epoch": 0.47099492276115373,
      "grad_norm": 0.1417073515720032,
      "learning_rate": 1.2736060747792877e-05,
      "loss": 0.0639,
      "step": 1635
    },
    {
      "epoch": 0.4724352742069065,
      "grad_norm": 0.14017747593627589,
      "learning_rate": 1.268764793658094e-05,
      "loss": 0.0677,
      "step": 1640
    },
    {
      "epoch": 0.4738756256526592,
      "grad_norm": 0.14349670833488368,
      "learning_rate": 1.2639167131892294e-05,
      "loss": 0.0575,
      "step": 1645
    },
    {
      "epoch": 0.475315977098412,
      "grad_norm": 0.1399709068160109,
      "learning_rate": 1.2590619560218851e-05,
      "loss": 0.0667,
      "step": 1650
    },
    {
      "epoch": 0.4767563285441648,
      "grad_norm": 0.16424535413090544,
      "learning_rate": 1.2542006449741631e-05,
      "loss": 0.0638,
      "step": 1655
    },
    {
      "epoch": 0.47819667998991755,
      "grad_norm": 0.13951898085012698,
      "learning_rate": 1.249332903029969e-05,
      "loss": 0.0636,
      "step": 1660
    },
    {
      "epoch": 0.4796370314356703,
      "grad_norm": 0.14482227044886414,
      "learning_rate": 1.2444588533358996e-05,
      "loss": 0.0623,
      "step": 1665
    },
    {
      "epoch": 0.48107738288142304,
      "grad_norm": 0.149830114351691,
      "learning_rate": 1.23957861919813e-05,
      "loss": 0.0631,
      "step": 1670
    },
    {
      "epoch": 0.4825177343271758,
      "grad_norm": 0.15277720600057818,
      "learning_rate": 1.2346923240792907e-05,
      "loss": 0.0656,
      "step": 1675
    },
    {
      "epoch": 0.4839580857729286,
      "grad_norm": 0.14145469398357127,
      "learning_rate": 1.229800091595347e-05,
      "loss": 0.0602,
      "step": 1680
    },
    {
      "epoch": 0.48539843721868137,
      "grad_norm": 0.145030274869654,
      "learning_rate": 1.2249020455124703e-05,
      "loss": 0.0605,
      "step": 1685
    },
    {
      "epoch": 0.48683878866443414,
      "grad_norm": 0.13292226014635264,
      "learning_rate": 1.2199983097439079e-05,
      "loss": 0.0576,
      "step": 1690
    },
    {
      "epoch": 0.48827914011018686,
      "grad_norm": 0.14980471445078763,
      "learning_rate": 1.2150890083468465e-05,
      "loss": 0.0613,
      "step": 1695
    },
    {
      "epoch": 0.48971949155593963,
      "grad_norm": 0.15237623136710496,
      "learning_rate": 1.2101742655192761e-05,
      "loss": 0.0641,
      "step": 1700
    },
    {
      "epoch": 0.4911598430016924,
      "grad_norm": 0.14853782098493126,
      "learning_rate": 1.2052542055968461e-05,
      "loss": 0.0635,
      "step": 1705
    },
    {
      "epoch": 0.4926001944474452,
      "grad_norm": 0.16064984443815533,
      "learning_rate": 1.2003289530497206e-05,
      "loss": 0.0664,
      "step": 1710
    },
    {
      "epoch": 0.49404054589319796,
      "grad_norm": 0.14592019591217842,
      "learning_rate": 1.1953986324794295e-05,
      "loss": 0.0621,
      "step": 1715
    },
    {
      "epoch": 0.4954808973389507,
      "grad_norm": 0.14899551585302004,
      "learning_rate": 1.1904633686157158e-05,
      "loss": 0.0661,
      "step": 1720
    },
    {
      "epoch": 0.49692124878470345,
      "grad_norm": 0.1662839926677418,
      "learning_rate": 1.1855232863133809e-05,
      "loss": 0.06,
      "step": 1725
    },
    {
      "epoch": 0.4983616002304562,
      "grad_norm": 0.14234586755951342,
      "learning_rate": 1.1805785105491247e-05,
      "loss": 0.0621,
      "step": 1730
    },
    {
      "epoch": 0.499801951676209,
      "grad_norm": 0.14261214561399946,
      "learning_rate": 1.1756291664183858e-05,
      "loss": 0.0649,
      "step": 1735
    },
    {
      "epoch": 0.5012423031219617,
      "grad_norm": 0.15004211923520408,
      "learning_rate": 1.1706753791321748e-05,
      "loss": 0.0619,
      "step": 1740
    },
    {
      "epoch": 0.5026826545677145,
      "grad_norm": 0.1605343977528051,
      "learning_rate": 1.1657172740139074e-05,
      "loss": 0.0612,
      "step": 1745
    },
    {
      "epoch": 0.5041230060134673,
      "grad_norm": 0.1447253493029267,
      "learning_rate": 1.1607549764962342e-05,
      "loss": 0.0634,
      "step": 1750
    },
    {
      "epoch": 0.5055633574592201,
      "grad_norm": 0.1346139878314749,
      "learning_rate": 1.1557886121178683e-05,
      "loss": 0.0664,
      "step": 1755
    },
    {
      "epoch": 0.5070037089049728,
      "grad_norm": 0.14646185824109836,
      "learning_rate": 1.1508183065204066e-05,
      "loss": 0.0621,
      "step": 1760
    },
    {
      "epoch": 0.5084440603507255,
      "grad_norm": 0.147637884179789,
      "learning_rate": 1.1458441854451539e-05,
      "loss": 0.0598,
      "step": 1765
    },
    {
      "epoch": 0.5098844117964784,
      "grad_norm": 0.1379823925555806,
      "learning_rate": 1.1408663747299409e-05,
      "loss": 0.0609,
      "step": 1770
    },
    {
      "epoch": 0.5113247632422311,
      "grad_norm": 0.1367073920260954,
      "learning_rate": 1.13588500030594e-05,
      "loss": 0.0625,
      "step": 1775
    },
    {
      "epoch": 0.5127651146879839,
      "grad_norm": 0.1358599716163257,
      "learning_rate": 1.130900188194481e-05,
      "loss": 0.0609,
      "step": 1780
    },
    {
      "epoch": 0.5142054661337366,
      "grad_norm": 0.1490120576953094,
      "learning_rate": 1.1259120645038612e-05,
      "loss": 0.0603,
      "step": 1785
    },
    {
      "epoch": 0.5156458175794894,
      "grad_norm": 0.1387405614374518,
      "learning_rate": 1.1209207554261573e-05,
      "loss": 0.0584,
      "step": 1790
    },
    {
      "epoch": 0.5170861690252422,
      "grad_norm": 0.1279767885162475,
      "learning_rate": 1.1159263872340293e-05,
      "loss": 0.06,
      "step": 1795
    },
    {
      "epoch": 0.5185265204709949,
      "grad_norm": 0.14033610983416972,
      "learning_rate": 1.1109290862775307e-05,
      "loss": 0.0559,
      "step": 1800
    },
    {
      "epoch": 0.5199668719167477,
      "grad_norm": 0.14071210053007446,
      "learning_rate": 1.1059289789809071e-05,
      "loss": 0.0591,
      "step": 1805
    },
    {
      "epoch": 0.5214072233625004,
      "grad_norm": 0.14899424536618833,
      "learning_rate": 1.1009261918394028e-05,
      "loss": 0.0573,
      "step": 1810
    },
    {
      "epoch": 0.5228475748082532,
      "grad_norm": 0.13803763706679575,
      "learning_rate": 1.0959208514160561e-05,
      "loss": 0.0643,
      "step": 1815
    },
    {
      "epoch": 0.524287926254006,
      "grad_norm": 0.14811878913823373,
      "learning_rate": 1.0909130843385009e-05,
      "loss": 0.0581,
      "step": 1820
    },
    {
      "epoch": 0.5257282776997587,
      "grad_norm": 0.14341490101278728,
      "learning_rate": 1.085903017295761e-05,
      "loss": 0.0575,
      "step": 1825
    },
    {
      "epoch": 0.5271686291455115,
      "grad_norm": 0.13792111506735585,
      "learning_rate": 1.0808907770350463e-05,
      "loss": 0.0593,
      "step": 1830
    },
    {
      "epoch": 0.5286089805912643,
      "grad_norm": 0.14395673234387163,
      "learning_rate": 1.0758764903585457e-05,
      "loss": 0.0609,
      "step": 1835
    },
    {
      "epoch": 0.530049332037017,
      "grad_norm": 0.1405053430609704,
      "learning_rate": 1.070860284120219e-05,
      "loss": 0.06,
      "step": 1840
    },
    {
      "epoch": 0.5314896834827698,
      "grad_norm": 0.15982572109675375,
      "learning_rate": 1.0658422852225889e-05,
      "loss": 0.0702,
      "step": 1845
    },
    {
      "epoch": 0.5329300349285225,
      "grad_norm": 0.14599264911552556,
      "learning_rate": 1.0608226206135292e-05,
      "loss": 0.0604,
      "step": 1850
    },
    {
      "epoch": 0.5343703863742754,
      "grad_norm": 0.14298260555858402,
      "learning_rate": 1.0558014172830537e-05,
      "loss": 0.0619,
      "step": 1855
    },
    {
      "epoch": 0.5358107378200281,
      "grad_norm": 0.13092410126397575,
      "learning_rate": 1.0507788022601033e-05,
      "loss": 0.057,
      "step": 1860
    },
    {
      "epoch": 0.5372510892657808,
      "grad_norm": 0.1453972372171612,
      "learning_rate": 1.0457549026093338e-05,
      "loss": 0.0585,
      "step": 1865
    },
    {
      "epoch": 0.5386914407115336,
      "grad_norm": 0.12317966498614596,
      "learning_rate": 1.0407298454278983e-05,
      "loss": 0.0564,
      "step": 1870
    },
    {
      "epoch": 0.5401317921572864,
      "grad_norm": 0.14436498244704113,
      "learning_rate": 1.0357037578422349e-05,
      "loss": 0.0645,
      "step": 1875
    },
    {
      "epoch": 0.5415721436030392,
      "grad_norm": 0.12970657416072792,
      "learning_rate": 1.0306767670048497e-05,
      "loss": 0.0589,
      "step": 1880
    },
    {
      "epoch": 0.5430124950487919,
      "grad_norm": 0.1367592762940604,
      "learning_rate": 1.0256490000910986e-05,
      "loss": 0.0621,
      "step": 1885
    },
    {
      "epoch": 0.5444528464945446,
      "grad_norm": 0.13370324714287835,
      "learning_rate": 1.0206205842959718e-05,
      "loss": 0.0626,
      "step": 1890
    },
    {
      "epoch": 0.5458931979402974,
      "grad_norm": 0.13187409015415524,
      "learning_rate": 1.0155916468308749e-05,
      "loss": 0.0614,
      "step": 1895
    },
    {
      "epoch": 0.5473335493860502,
      "grad_norm": 0.16314950628212277,
      "learning_rate": 1.0105623149204118e-05,
      "loss": 0.0642,
      "step": 1900
    },
    {
      "epoch": 0.548773900831803,
      "grad_norm": 0.15059666297472246,
      "learning_rate": 1.0055327157991639e-05,
      "loss": 0.0613,
      "step": 1905
    },
    {
      "epoch": 0.5502142522775557,
      "grad_norm": 0.1479385664820797,
      "learning_rate": 1.0005029767084739e-05,
      "loss": 0.0608,
      "step": 1910
    },
    {
      "epoch": 0.5516546037233084,
      "grad_norm": 0.14109693631194206,
      "learning_rate": 9.954732248932243e-06,
      "loss": 0.064,
      "step": 1915
    },
    {
      "epoch": 0.5530949551690613,
      "grad_norm": 0.13106823837594228,
      "learning_rate": 9.904435875986213e-06,
      "loss": 0.0584,
      "step": 1920
    },
    {
      "epoch": 0.554535306614814,
      "grad_norm": 0.14125376801616402,
      "learning_rate": 9.85414192066972e-06,
      "loss": 0.0619,
      "step": 1925
    },
    {
      "epoch": 0.5559756580605668,
      "grad_norm": 0.13715317817212974,
      "learning_rate": 9.803851655344682e-06,
      "loss": 0.0557,
      "step": 1930
    },
    {
      "epoch": 0.5574160095063195,
      "grad_norm": 0.12415591514509057,
      "learning_rate": 9.75356635227966e-06,
      "loss": 0.0605,
      "step": 1935
    },
    {
      "epoch": 0.5588563609520723,
      "grad_norm": 0.1471168680809387,
      "learning_rate": 9.70328728361769e-06,
      "loss": 0.0624,
      "step": 1940
    },
    {
      "epoch": 0.5602967123978251,
      "grad_norm": 0.13972784391125845,
      "learning_rate": 9.653015721344073e-06,
      "loss": 0.0625,
      "step": 1945
    },
    {
      "epoch": 0.5617370638435778,
      "grad_norm": 0.14230983317081475,
      "learning_rate": 9.602752937254215e-06,
      "loss": 0.0595,
      "step": 1950
    },
    {
      "epoch": 0.5631774152893306,
      "grad_norm": 0.17270902776994063,
      "learning_rate": 9.552500202921449e-06,
      "loss": 0.0633,
      "step": 1955
    },
    {
      "epoch": 0.5646177667350833,
      "grad_norm": 0.16874434258709992,
      "learning_rate": 9.502258789664865e-06,
      "loss": 0.0647,
      "step": 1960
    },
    {
      "epoch": 0.5660581181808362,
      "grad_norm": 0.14698351108372149,
      "learning_rate": 9.45202996851714e-06,
      "loss": 0.0652,
      "step": 1965
    },
    {
      "epoch": 0.5674984696265889,
      "grad_norm": 0.1583270327898255,
      "learning_rate": 9.401815010192388e-06,
      "loss": 0.0646,
      "step": 1970
    },
    {
      "epoch": 0.5689388210723416,
      "grad_norm": 0.1449159304203145,
      "learning_rate": 9.351615185054029e-06,
      "loss": 0.0613,
      "step": 1975
    },
    {
      "epoch": 0.5703791725180944,
      "grad_norm": 0.1365799756032449,
      "learning_rate": 9.301431763082623e-06,
      "loss": 0.059,
      "step": 1980
    },
    {
      "epoch": 0.5718195239638472,
      "grad_norm": 0.14707584276953978,
      "learning_rate": 9.251266013843757e-06,
      "loss": 0.0588,
      "step": 1985
    },
    {
      "epoch": 0.5732598754096,
      "grad_norm": 0.16188309788542712,
      "learning_rate": 9.201119206455927e-06,
      "loss": 0.0655,
      "step": 1990
    },
    {
      "epoch": 0.5747002268553527,
      "grad_norm": 0.1524865885491688,
      "learning_rate": 9.150992609558425e-06,
      "loss": 0.0608,
      "step": 1995
    },
    {
      "epoch": 0.5761405783011054,
      "grad_norm": 0.1439363997300933,
      "learning_rate": 9.10088749127926e-06,
      "loss": 0.0602,
      "step": 2000
    },
    {
      "epoch": 0.5761405783011054,
      "eval_loss": 0.06202974170446396,
      "eval_runtime": 197.6148,
      "eval_samples_per_second": 9.129,
      "eval_steps_per_second": 2.282,
      "step": 2000
    },
    {
      "epoch": 0.5775809297468583,
      "grad_norm": 0.13774654558009133,
      "learning_rate": 9.050805119203035e-06,
      "loss": 0.0574,
      "step": 2005
    },
    {
      "epoch": 0.579021281192611,
      "grad_norm": 0.1598545346397363,
      "learning_rate": 9.000746760338938e-06,
      "loss": 0.0644,
      "step": 2010
    },
    {
      "epoch": 0.5804616326383638,
      "grad_norm": 0.13573728248991526,
      "learning_rate": 8.950713681088647e-06,
      "loss": 0.0587,
      "step": 2015
    },
    {
      "epoch": 0.5819019840841165,
      "grad_norm": 0.14813992842491036,
      "learning_rate": 8.900707147214301e-06,
      "loss": 0.0627,
      "step": 2020
    },
    {
      "epoch": 0.5833423355298692,
      "grad_norm": 0.1440587796894665,
      "learning_rate": 8.850728423806487e-06,
      "loss": 0.061,
      "step": 2025
    },
    {
      "epoch": 0.5847826869756221,
      "grad_norm": 0.1358163633361261,
      "learning_rate": 8.800778775252221e-06,
      "loss": 0.0582,
      "step": 2030
    },
    {
      "epoch": 0.5862230384213748,
      "grad_norm": 0.14949783609406306,
      "learning_rate": 8.75085946520298e-06,
      "loss": 0.0673,
      "step": 2035
    },
    {
      "epoch": 0.5876633898671276,
      "grad_norm": 0.14548649497172944,
      "learning_rate": 8.700971756542707e-06,
      "loss": 0.06,
      "step": 2040
    },
    {
      "epoch": 0.5891037413128803,
      "grad_norm": 0.1482887794432985,
      "learning_rate": 8.65111691135589e-06,
      "loss": 0.0605,
      "step": 2045
    },
    {
      "epoch": 0.5905440927586331,
      "grad_norm": 0.14245894718321123,
      "learning_rate": 8.601296190895611e-06,
      "loss": 0.0592,
      "step": 2050
    },
    {
      "epoch": 0.5919844442043859,
      "grad_norm": 0.1299014295335545,
      "learning_rate": 8.551510855551658e-06,
      "loss": 0.0625,
      "step": 2055
    },
    {
      "epoch": 0.5934247956501386,
      "grad_norm": 0.14990167404663077,
      "learning_rate": 8.501762164818615e-06,
      "loss": 0.0615,
      "step": 2060
    },
    {
      "epoch": 0.5948651470958914,
      "grad_norm": 0.15687154771193254,
      "learning_rate": 8.452051377264025e-06,
      "loss": 0.0676,
      "step": 2065
    },
    {
      "epoch": 0.5963054985416442,
      "grad_norm": 0.14285391504986938,
      "learning_rate": 8.402379750496535e-06,
      "loss": 0.063,
      "step": 2070
    },
    {
      "epoch": 0.5977458499873969,
      "grad_norm": 0.16481562129681976,
      "learning_rate": 8.35274854113407e-06,
      "loss": 0.062,
      "step": 2075
    },
    {
      "epoch": 0.5991862014331497,
      "grad_norm": 0.13142064433487616,
      "learning_rate": 8.303159004772065e-06,
      "loss": 0.0603,
      "step": 2080
    },
    {
      "epoch": 0.6006265528789024,
      "grad_norm": 0.1515273341700813,
      "learning_rate": 8.253612395951697e-06,
      "loss": 0.0656,
      "step": 2085
    },
    {
      "epoch": 0.6020669043246553,
      "grad_norm": 0.17295436986543897,
      "learning_rate": 8.204109968128128e-06,
      "loss": 0.0655,
      "step": 2090
    },
    {
      "epoch": 0.603507255770408,
      "grad_norm": 0.14037313951418073,
      "learning_rate": 8.15465297363881e-06,
      "loss": 0.0662,
      "step": 2095
    },
    {
      "epoch": 0.6049476072161607,
      "grad_norm": 0.13747939963648245,
      "learning_rate": 8.105242663671807e-06,
      "loss": 0.0615,
      "step": 2100
    },
    {
      "epoch": 0.6063879586619135,
      "grad_norm": 0.12848974547228537,
      "learning_rate": 8.055880288234127e-06,
      "loss": 0.0615,
      "step": 2105
    },
    {
      "epoch": 0.6078283101076662,
      "grad_norm": 0.1429108532569868,
      "learning_rate": 8.006567096120103e-06,
      "loss": 0.0606,
      "step": 2110
    },
    {
      "epoch": 0.6092686615534191,
      "grad_norm": 0.13212576027139222,
      "learning_rate": 7.957304334879809e-06,
      "loss": 0.0644,
      "step": 2115
    },
    {
      "epoch": 0.6107090129991718,
      "grad_norm": 0.14626257804070625,
      "learning_rate": 7.908093250787496e-06,
      "loss": 0.0616,
      "step": 2120
    },
    {
      "epoch": 0.6121493644449245,
      "grad_norm": 0.1369876288040746,
      "learning_rate": 7.858935088810054e-06,
      "loss": 0.0643,
      "step": 2125
    },
    {
      "epoch": 0.6135897158906773,
      "grad_norm": 0.1562204322111819,
      "learning_rate": 7.809831092575528e-06,
      "loss": 0.0626,
      "step": 2130
    },
    {
      "epoch": 0.6150300673364301,
      "grad_norm": 0.1478800925724685,
      "learning_rate": 7.760782504341644e-06,
      "loss": 0.0643,
      "step": 2135
    },
    {
      "epoch": 0.6164704187821829,
      "grad_norm": 0.14017951944044135,
      "learning_rate": 7.7117905649644e-06,
      "loss": 0.0582,
      "step": 2140
    },
    {
      "epoch": 0.6179107702279356,
      "grad_norm": 0.18625663278817955,
      "learning_rate": 7.662856513866643e-06,
      "loss": 0.0567,
      "step": 2145
    },
    {
      "epoch": 0.6193511216736883,
      "grad_norm": 0.13723738608866848,
      "learning_rate": 7.613981589006754e-06,
      "loss": 0.0595,
      "step": 2150
    },
    {
      "epoch": 0.6207914731194412,
      "grad_norm": 0.17391455144126491,
      "learning_rate": 7.565167026847294e-06,
      "loss": 0.0577,
      "step": 2155
    },
    {
      "epoch": 0.6222318245651939,
      "grad_norm": 0.13637323163266152,
      "learning_rate": 7.5164140623237454e-06,
      "loss": 0.0583,
      "step": 2160
    },
    {
      "epoch": 0.6236721760109467,
      "grad_norm": 0.13490860613290204,
      "learning_rate": 7.467723928813262e-06,
      "loss": 0.0616,
      "step": 2165
    },
    {
      "epoch": 0.6251125274566994,
      "grad_norm": 0.13694189106475327,
      "learning_rate": 7.419097858103464e-06,
      "loss": 0.0584,
      "step": 2170
    },
    {
      "epoch": 0.6265528789024521,
      "grad_norm": 0.14264218273335677,
      "learning_rate": 7.370537080361282e-06,
      "loss": 0.0584,
      "step": 2175
    },
    {
      "epoch": 0.627993230348205,
      "grad_norm": 0.14133949369553425,
      "learning_rate": 7.3220428241018225e-06,
      "loss": 0.0565,
      "step": 2180
    },
    {
      "epoch": 0.6294335817939577,
      "grad_norm": 0.13811563164931434,
      "learning_rate": 7.273616316157312e-06,
      "loss": 0.0593,
      "step": 2185
    },
    {
      "epoch": 0.6308739332397105,
      "grad_norm": 0.14172526952252731,
      "learning_rate": 7.225258781646036e-06,
      "loss": 0.0563,
      "step": 2190
    },
    {
      "epoch": 0.6323142846854632,
      "grad_norm": 0.14556762141841273,
      "learning_rate": 7.176971443941359e-06,
      "loss": 0.0581,
      "step": 2195
    },
    {
      "epoch": 0.633754636131216,
      "grad_norm": 0.14455168465042534,
      "learning_rate": 7.128755524640771e-06,
      "loss": 0.0616,
      "step": 2200
    },
    {
      "epoch": 0.6351949875769688,
      "grad_norm": 0.14914771737576238,
      "learning_rate": 7.080612243534981e-06,
      "loss": 0.0632,
      "step": 2205
    },
    {
      "epoch": 0.6366353390227215,
      "grad_norm": 0.15500897160871308,
      "learning_rate": 7.032542818577066e-06,
      "loss": 0.0574,
      "step": 2210
    },
    {
      "epoch": 0.6380756904684743,
      "grad_norm": 0.12621800276033995,
      "learning_rate": 6.984548465851641e-06,
      "loss": 0.0578,
      "step": 2215
    },
    {
      "epoch": 0.6395160419142271,
      "grad_norm": 0.14453230217275803,
      "learning_rate": 6.93663039954412e-06,
      "loss": 0.0598,
      "step": 2220
    },
    {
      "epoch": 0.6409563933599798,
      "grad_norm": 0.12835711140154188,
      "learning_rate": 6.888789831909972e-06,
      "loss": 0.0609,
      "step": 2225
    },
    {
      "epoch": 0.6423967448057326,
      "grad_norm": 0.14406333134449356,
      "learning_rate": 6.841027973244077e-06,
      "loss": 0.0573,
      "step": 2230
    },
    {
      "epoch": 0.6438370962514853,
      "grad_norm": 0.14298113650078734,
      "learning_rate": 6.793346031850085e-06,
      "loss": 0.0596,
      "step": 2235
    },
    {
      "epoch": 0.6452774476972382,
      "grad_norm": 0.1351787448129904,
      "learning_rate": 6.745745214009866e-06,
      "loss": 0.0579,
      "step": 2240
    },
    {
      "epoch": 0.6467177991429909,
      "grad_norm": 0.1416662783049815,
      "learning_rate": 6.698226723952985e-06,
      "loss": 0.0601,
      "step": 2245
    },
    {
      "epoch": 0.6481581505887437,
      "grad_norm": 0.13123613630770536,
      "learning_rate": 6.65079176382623e-06,
      "loss": 0.0622,
      "step": 2250
    },
    {
      "epoch": 0.6495985020344964,
      "grad_norm": 0.13862477589129096,
      "learning_rate": 6.603441533663214e-06,
      "loss": 0.0641,
      "step": 2255
    },
    {
      "epoch": 0.6510388534802491,
      "grad_norm": 0.13893589833870676,
      "learning_rate": 6.556177231354003e-06,
      "loss": 0.0623,
      "step": 2260
    },
    {
      "epoch": 0.652479204926002,
      "grad_norm": 0.13047767245398073,
      "learning_rate": 6.509000052614824e-06,
      "loss": 0.0625,
      "step": 2265
    },
    {
      "epoch": 0.6539195563717547,
      "grad_norm": 0.14948071139040017,
      "learning_rate": 6.4619111909577995e-06,
      "loss": 0.0568,
      "step": 2270
    },
    {
      "epoch": 0.6553599078175075,
      "grad_norm": 0.16740289210607096,
      "learning_rate": 6.414911837660768e-06,
      "loss": 0.0593,
      "step": 2275
    },
    {
      "epoch": 0.6568002592632602,
      "grad_norm": 0.1392403260751934,
      "learning_rate": 6.36800318173714e-06,
      "loss": 0.0614,
      "step": 2280
    },
    {
      "epoch": 0.658240610709013,
      "grad_norm": 0.14713068471331234,
      "learning_rate": 6.321186409905812e-06,
      "loss": 0.0591,
      "step": 2285
    },
    {
      "epoch": 0.6596809621547658,
      "grad_norm": 0.15702296926939516,
      "learning_rate": 6.274462706561153e-06,
      "loss": 0.06,
      "step": 2290
    },
    {
      "epoch": 0.6611213136005185,
      "grad_norm": 0.14577218240146522,
      "learning_rate": 6.227833253743045e-06,
      "loss": 0.0577,
      "step": 2295
    },
    {
      "epoch": 0.6625616650462713,
      "grad_norm": 0.1347810878894777,
      "learning_rate": 6.181299231106963e-06,
      "loss": 0.0595,
      "step": 2300
    },
    {
      "epoch": 0.6640020164920241,
      "grad_norm": 0.13273221655238504,
      "learning_rate": 6.134861815894146e-06,
      "loss": 0.0569,
      "step": 2305
    },
    {
      "epoch": 0.6654423679377768,
      "grad_norm": 0.13814471094934377,
      "learning_rate": 6.08852218290181e-06,
      "loss": 0.0561,
      "step": 2310
    },
    {
      "epoch": 0.6668827193835296,
      "grad_norm": 0.13189040653435405,
      "learning_rate": 6.0422815044534265e-06,
      "loss": 0.0632,
      "step": 2315
    },
    {
      "epoch": 0.6683230708292823,
      "grad_norm": 0.15479186111317447,
      "learning_rate": 5.9961409503690605e-06,
      "loss": 0.0566,
      "step": 2320
    },
    {
      "epoch": 0.6697634222750352,
      "grad_norm": 0.15426426306388247,
      "learning_rate": 5.950101687935783e-06,
      "loss": 0.0615,
      "step": 2325
    },
    {
      "epoch": 0.6712037737207879,
      "grad_norm": 0.12977972650554861,
      "learning_rate": 5.904164881878143e-06,
      "loss": 0.064,
      "step": 2330
    },
    {
      "epoch": 0.6726441251665406,
      "grad_norm": 0.12489906355375564,
      "learning_rate": 5.858331694328686e-06,
      "loss": 0.0588,
      "step": 2335
    },
    {
      "epoch": 0.6740844766122934,
      "grad_norm": 0.14177743912393773,
      "learning_rate": 5.812603284798575e-06,
      "loss": 0.0559,
      "step": 2340
    },
    {
      "epoch": 0.6755248280580461,
      "grad_norm": 0.127402859937035,
      "learning_rate": 5.766980810148238e-06,
      "loss": 0.0585,
      "step": 2345
    },
    {
      "epoch": 0.676965179503799,
      "grad_norm": 0.14721828485950575,
      "learning_rate": 5.721465424558111e-06,
      "loss": 0.0569,
      "step": 2350
    },
    {
      "epoch": 0.6784055309495517,
      "grad_norm": 0.12933895802216824,
      "learning_rate": 5.676058279499438e-06,
      "loss": 0.0583,
      "step": 2355
    },
    {
      "epoch": 0.6798458823953044,
      "grad_norm": 0.2334988683393806,
      "learning_rate": 5.630760523705137e-06,
      "loss": 0.0577,
      "step": 2360
    },
    {
      "epoch": 0.6812862338410572,
      "grad_norm": 0.13307693046592253,
      "learning_rate": 5.585573303140741e-06,
      "loss": 0.0607,
      "step": 2365
    },
    {
      "epoch": 0.68272658528681,
      "grad_norm": 0.14681808383769718,
      "learning_rate": 5.540497760975412e-06,
      "loss": 0.0596,
      "step": 2370
    },
    {
      "epoch": 0.6841669367325628,
      "grad_norm": 0.14526949538273093,
      "learning_rate": 5.495535037553003e-06,
      "loss": 0.0575,
      "step": 2375
    },
    {
      "epoch": 0.6856072881783155,
      "grad_norm": 0.15463516823806045,
      "learning_rate": 5.450686270363244e-06,
      "loss": 0.063,
      "step": 2380
    },
    {
      "epoch": 0.6870476396240682,
      "grad_norm": 0.1370433070265185,
      "learning_rate": 5.405952594012921e-06,
      "loss": 0.0574,
      "step": 2385
    },
    {
      "epoch": 0.6884879910698211,
      "grad_norm": 0.13459031723410086,
      "learning_rate": 5.361335140197199e-06,
      "loss": 0.0616,
      "step": 2390
    },
    {
      "epoch": 0.6899283425155738,
      "grad_norm": 0.15099977742892481,
      "learning_rate": 5.316835037670987e-06,
      "loss": 0.0585,
      "step": 2395
    },
    {
      "epoch": 0.6913686939613266,
      "grad_norm": 0.1475383831706844,
      "learning_rate": 5.272453412220389e-06,
      "loss": 0.0606,
      "step": 2400
    },
    {
      "epoch": 0.6928090454070793,
      "grad_norm": 0.12116703992402802,
      "learning_rate": 5.228191386634212e-06,
      "loss": 0.0552,
      "step": 2405
    },
    {
      "epoch": 0.694249396852832,
      "grad_norm": 0.18353789252461628,
      "learning_rate": 5.184050080675558e-06,
      "loss": 0.0581,
      "step": 2410
    },
    {
      "epoch": 0.6956897482985849,
      "grad_norm": 0.15256328875682043,
      "learning_rate": 5.140030611053512e-06,
      "loss": 0.0583,
      "step": 2415
    },
    {
      "epoch": 0.6971300997443376,
      "grad_norm": 0.14298118193344345,
      "learning_rate": 5.096134091394879e-06,
      "loss": 0.0625,
      "step": 2420
    },
    {
      "epoch": 0.6985704511900904,
      "grad_norm": 0.15308814352700958,
      "learning_rate": 5.052361632216009e-06,
      "loss": 0.0563,
      "step": 2425
    },
    {
      "epoch": 0.7000108026358431,
      "grad_norm": 0.12806804658727294,
      "learning_rate": 5.008714340894716e-06,
      "loss": 0.0621,
      "step": 2430
    },
    {
      "epoch": 0.7014511540815959,
      "grad_norm": 0.1283153150885506,
      "learning_rate": 4.965193321642245e-06,
      "loss": 0.0639,
      "step": 2435
    },
    {
      "epoch": 0.7028915055273487,
      "grad_norm": 0.16204924222767175,
      "learning_rate": 4.9217996754753505e-06,
      "loss": 0.0632,
      "step": 2440
    },
    {
      "epoch": 0.7043318569731014,
      "grad_norm": 0.13720075582090696,
      "learning_rate": 4.878534500188443e-06,
      "loss": 0.0579,
      "step": 2445
    },
    {
      "epoch": 0.7057722084188542,
      "grad_norm": 0.1336786402162477,
      "learning_rate": 4.835398890325806e-06,
      "loss": 0.0602,
      "step": 2450
    },
    {
      "epoch": 0.707212559864607,
      "grad_norm": 0.15250075242936512,
      "learning_rate": 4.792393937153914e-06,
      "loss": 0.0587,
      "step": 2455
    },
    {
      "epoch": 0.7086529113103597,
      "grad_norm": 0.14586936476718837,
      "learning_rate": 4.749520728633812e-06,
      "loss": 0.0599,
      "step": 2460
    },
    {
      "epoch": 0.7100932627561125,
      "grad_norm": 0.13737451021327757,
      "learning_rate": 4.706780349393621e-06,
      "loss": 0.059,
      "step": 2465
    },
    {
      "epoch": 0.7115336142018652,
      "grad_norm": 0.14534033618919,
      "learning_rate": 4.664173880701065e-06,
      "loss": 0.06,
      "step": 2470
    },
    {
      "epoch": 0.7129739656476181,
      "grad_norm": 0.14705542529286947,
      "learning_rate": 4.62170240043614e-06,
      "loss": 0.0576,
      "step": 2475
    },
    {
      "epoch": 0.7144143170933708,
      "grad_norm": 0.1599807488079511,
      "learning_rate": 4.579366983063829e-06,
      "loss": 0.0597,
      "step": 2480
    },
    {
      "epoch": 0.7158546685391235,
      "grad_norm": 0.1345238151706872,
      "learning_rate": 4.537168699606928e-06,
      "loss": 0.0618,
      "step": 2485
    },
    {
      "epoch": 0.7172950199848763,
      "grad_norm": 0.1469249032470179,
      "learning_rate": 4.4951086176189666e-06,
      "loss": 0.0633,
      "step": 2490
    },
    {
      "epoch": 0.718735371430629,
      "grad_norm": 0.13426918611168998,
      "learning_rate": 4.453187801157155e-06,
      "loss": 0.0585,
      "step": 2495
    },
    {
      "epoch": 0.7201757228763819,
      "grad_norm": 0.14672886262551482,
      "learning_rate": 4.411407310755513e-06,
      "loss": 0.0599,
      "step": 2500
    },
    {
      "epoch": 0.7201757228763819,
      "eval_loss": 0.06022631376981735,
      "eval_runtime": 199.5971,
      "eval_samples_per_second": 9.038,
      "eval_steps_per_second": 2.26,
      "step": 2500
    },
    {
      "epoch": 0.7216160743221346,
      "grad_norm": 0.1432101633635202,
      "learning_rate": 4.369768203398014e-06,
      "loss": 0.0554,
      "step": 2505
    },
    {
      "epoch": 0.7230564257678873,
      "grad_norm": 0.14150540798087066,
      "learning_rate": 4.328271532491859e-06,
      "loss": 0.0606,
      "step": 2510
    },
    {
      "epoch": 0.7244967772136401,
      "grad_norm": 0.14496753728614564,
      "learning_rate": 4.2869183478408125e-06,
      "loss": 0.0601,
      "step": 2515
    },
    {
      "epoch": 0.7259371286593929,
      "grad_norm": 0.13903106714425764,
      "learning_rate": 4.2457096956186525e-06,
      "loss": 0.0608,
      "step": 2520
    },
    {
      "epoch": 0.7273774801051457,
      "grad_norm": 0.15285152111447173,
      "learning_rate": 4.2046466183427035e-06,
      "loss": 0.0653,
      "step": 2525
    },
    {
      "epoch": 0.7288178315508984,
      "grad_norm": 0.1266370772861076,
      "learning_rate": 4.163730154847462e-06,
      "loss": 0.0527,
      "step": 2530
    },
    {
      "epoch": 0.7302581829966511,
      "grad_norm": 0.13117482319657733,
      "learning_rate": 4.122961340258312e-06,
      "loss": 0.0559,
      "step": 2535
    },
    {
      "epoch": 0.731698534442404,
      "grad_norm": 0.14194653004592997,
      "learning_rate": 4.082341205965344e-06,
      "loss": 0.059,
      "step": 2540
    },
    {
      "epoch": 0.7331388858881567,
      "grad_norm": 0.15066840894513114,
      "learning_rate": 4.0418707795972575e-06,
      "loss": 0.0592,
      "step": 2545
    },
    {
      "epoch": 0.7345792373339095,
      "grad_norm": 0.14684421420293026,
      "learning_rate": 4.001551084995363e-06,
      "loss": 0.0627,
      "step": 2550
    },
    {
      "epoch": 0.7360195887796622,
      "grad_norm": 0.13640481951408895,
      "learning_rate": 3.961383142187691e-06,
      "loss": 0.0608,
      "step": 2555
    },
    {
      "epoch": 0.7374599402254151,
      "grad_norm": 0.14295407920295589,
      "learning_rate": 3.9213679673631745e-06,
      "loss": 0.0614,
      "step": 2560
    },
    {
      "epoch": 0.7389002916711678,
      "grad_norm": 0.1407341748583866,
      "learning_rate": 3.881506572845933e-06,
      "loss": 0.063,
      "step": 2565
    },
    {
      "epoch": 0.7403406431169205,
      "grad_norm": 0.14190189878723156,
      "learning_rate": 3.841799967069686e-06,
      "loss": 0.0657,
      "step": 2570
    },
    {
      "epoch": 0.7417809945626733,
      "grad_norm": 0.13086316062234218,
      "learning_rate": 3.8022491545522346e-06,
      "loss": 0.0625,
      "step": 2575
    },
    {
      "epoch": 0.743221346008426,
      "grad_norm": 0.15392265951922993,
      "learning_rate": 3.7628551358700303e-06,
      "loss": 0.0594,
      "step": 2580
    },
    {
      "epoch": 0.7446616974541789,
      "grad_norm": 0.1417233830083159,
      "learning_rate": 3.723618907632882e-06,
      "loss": 0.0613,
      "step": 2585
    },
    {
      "epoch": 0.7461020488999316,
      "grad_norm": 0.1454111678080567,
      "learning_rate": 3.6845414624587326e-06,
      "loss": 0.063,
      "step": 2590
    },
    {
      "epoch": 0.7475424003456843,
      "grad_norm": 0.13443095013823156,
      "learning_rate": 3.6456237889485547e-06,
      "loss": 0.0623,
      "step": 2595
    },
    {
      "epoch": 0.7489827517914371,
      "grad_norm": 0.12036202234264778,
      "learning_rate": 3.606866871661333e-06,
      "loss": 0.0567,
      "step": 2600
    },
    {
      "epoch": 0.7504231032371899,
      "grad_norm": 0.13114907916546412,
      "learning_rate": 3.5682716910891613e-06,
      "loss": 0.0586,
      "step": 2605
    },
    {
      "epoch": 0.7518634546829427,
      "grad_norm": 0.15054577661915033,
      "learning_rate": 3.5298392236324365e-06,
      "loss": 0.0608,
      "step": 2610
    },
    {
      "epoch": 0.7533038061286954,
      "grad_norm": 0.13218839559511986,
      "learning_rate": 3.491570441575154e-06,
      "loss": 0.0587,
      "step": 2615
    },
    {
      "epoch": 0.7547441575744481,
      "grad_norm": 0.1268003731336924,
      "learning_rate": 3.453466313060322e-06,
      "loss": 0.0601,
      "step": 2620
    },
    {
      "epoch": 0.756184509020201,
      "grad_norm": 0.14035220802374737,
      "learning_rate": 3.4155278020654502e-06,
      "loss": 0.0605,
      "step": 2625
    },
    {
      "epoch": 0.7576248604659537,
      "grad_norm": 0.1426120779032921,
      "learning_rate": 3.3777558683781795e-06,
      "loss": 0.0592,
      "step": 2630
    },
    {
      "epoch": 0.7590652119117065,
      "grad_norm": 0.12614500048877936,
      "learning_rate": 3.3401514675719815e-06,
      "loss": 0.0577,
      "step": 2635
    },
    {
      "epoch": 0.7605055633574592,
      "grad_norm": 0.13720240974898895,
      "learning_rate": 3.302715550982014e-06,
      "loss": 0.0593,
      "step": 2640
    },
    {
      "epoch": 0.7619459148032119,
      "grad_norm": 0.3087564092748528,
      "learning_rate": 3.2654490656810256e-06,
      "loss": 0.0566,
      "step": 2645
    },
    {
      "epoch": 0.7633862662489648,
      "grad_norm": 0.16717331588926376,
      "learning_rate": 3.228352954455406e-06,
      "loss": 0.0564,
      "step": 2650
    },
    {
      "epoch": 0.7648266176947175,
      "grad_norm": 0.14323201872042673,
      "learning_rate": 3.1914281557813386e-06,
      "loss": 0.0567,
      "step": 2655
    },
    {
      "epoch": 0.7662669691404703,
      "grad_norm": 0.1497485160314171,
      "learning_rate": 3.1546756038010507e-06,
      "loss": 0.0588,
      "step": 2660
    },
    {
      "epoch": 0.767707320586223,
      "grad_norm": 0.13401880002666341,
      "learning_rate": 3.1180962282991976e-06,
      "loss": 0.0606,
      "step": 2665
    },
    {
      "epoch": 0.7691476720319758,
      "grad_norm": 0.13208176864571497,
      "learning_rate": 3.081690954679313e-06,
      "loss": 0.0566,
      "step": 2670
    },
    {
      "epoch": 0.7705880234777286,
      "grad_norm": 0.12464358137882478,
      "learning_rate": 3.0454607039404206e-06,
      "loss": 0.0572,
      "step": 2675
    },
    {
      "epoch": 0.7720283749234813,
      "grad_norm": 0.13502277775724228,
      "learning_rate": 3.0094063926537233e-06,
      "loss": 0.0585,
      "step": 2680
    },
    {
      "epoch": 0.7734687263692341,
      "grad_norm": 0.13081561307511916,
      "learning_rate": 2.973528932939429e-06,
      "loss": 0.0564,
      "step": 2685
    },
    {
      "epoch": 0.7749090778149869,
      "grad_norm": 0.1268910010135427,
      "learning_rate": 2.937829232443654e-06,
      "loss": 0.0555,
      "step": 2690
    },
    {
      "epoch": 0.7763494292607396,
      "grad_norm": 0.13623501931518958,
      "learning_rate": 2.9023081943154753e-06,
      "loss": 0.0601,
      "step": 2695
    },
    {
      "epoch": 0.7777897807064924,
      "grad_norm": 0.14656323058804094,
      "learning_rate": 2.86696671718408e-06,
      "loss": 0.0564,
      "step": 2700
    },
    {
      "epoch": 0.7792301321522451,
      "grad_norm": 0.15181738159360775,
      "learning_rate": 2.8318056951360294e-06,
      "loss": 0.0566,
      "step": 2705
    },
    {
      "epoch": 0.780670483597998,
      "grad_norm": 0.16379931893275942,
      "learning_rate": 2.7968260176926407e-06,
      "loss": 0.0611,
      "step": 2710
    },
    {
      "epoch": 0.7821108350437507,
      "grad_norm": 0.13279024083856536,
      "learning_rate": 2.762028569787485e-06,
      "loss": 0.0632,
      "step": 2715
    },
    {
      "epoch": 0.7835511864895034,
      "grad_norm": 0.1514977510458698,
      "learning_rate": 2.7274142317439956e-06,
      "loss": 0.0579,
      "step": 2720
    },
    {
      "epoch": 0.7849915379352562,
      "grad_norm": 0.13057048011714373,
      "learning_rate": 2.6929838792532035e-06,
      "loss": 0.062,
      "step": 2725
    },
    {
      "epoch": 0.7864318893810089,
      "grad_norm": 0.13540271426738182,
      "learning_rate": 2.6587383833515746e-06,
      "loss": 0.0561,
      "step": 2730
    },
    {
      "epoch": 0.7878722408267618,
      "grad_norm": 0.1571274616935904,
      "learning_rate": 2.6246786103989887e-06,
      "loss": 0.0565,
      "step": 2735
    },
    {
      "epoch": 0.7893125922725145,
      "grad_norm": 0.13901055115200536,
      "learning_rate": 2.590805422056807e-06,
      "loss": 0.0581,
      "step": 2740
    },
    {
      "epoch": 0.7907529437182672,
      "grad_norm": 0.1292805259799327,
      "learning_rate": 2.5571196752660733e-06,
      "loss": 0.0582,
      "step": 2745
    },
    {
      "epoch": 0.79219329516402,
      "grad_norm": 0.14166775661867112,
      "learning_rate": 2.5236222222258455e-06,
      "loss": 0.0592,
      "step": 2750
    },
    {
      "epoch": 0.7936336466097728,
      "grad_norm": 0.14804665844069495,
      "learning_rate": 2.4903139103716365e-06,
      "loss": 0.0574,
      "step": 2755
    },
    {
      "epoch": 0.7950739980555256,
      "grad_norm": 0.13663548348085658,
      "learning_rate": 2.4571955823539617e-06,
      "loss": 0.0546,
      "step": 2760
    },
    {
      "epoch": 0.7965143495012783,
      "grad_norm": 0.16386374204432605,
      "learning_rate": 2.424268076017032e-06,
      "loss": 0.0582,
      "step": 2765
    },
    {
      "epoch": 0.797954700947031,
      "grad_norm": 0.13788750638406272,
      "learning_rate": 2.3915322243775564e-06,
      "loss": 0.054,
      "step": 2770
    },
    {
      "epoch": 0.7993950523927839,
      "grad_norm": 0.13024548818008383,
      "learning_rate": 2.3589888556036623e-06,
      "loss": 0.0596,
      "step": 2775
    },
    {
      "epoch": 0.8008354038385366,
      "grad_norm": 0.1434591335226828,
      "learning_rate": 2.3266387929939525e-06,
      "loss": 0.0595,
      "step": 2780
    },
    {
      "epoch": 0.8022757552842894,
      "grad_norm": 0.1323094772826723,
      "learning_rate": 2.294482854956672e-06,
      "loss": 0.0577,
      "step": 2785
    },
    {
      "epoch": 0.8037161067300421,
      "grad_norm": 0.13191375863763416,
      "learning_rate": 2.2625218549890014e-06,
      "loss": 0.0586,
      "step": 2790
    },
    {
      "epoch": 0.8051564581757948,
      "grad_norm": 0.1436295049255119,
      "learning_rate": 2.230756601656481e-06,
      "loss": 0.0614,
      "step": 2795
    },
    {
      "epoch": 0.8065968096215477,
      "grad_norm": 0.14620897997093418,
      "learning_rate": 2.1991878985725566e-06,
      "loss": 0.0549,
      "step": 2800
    },
    {
      "epoch": 0.8080371610673004,
      "grad_norm": 0.13487749668910298,
      "learning_rate": 2.167816544378244e-06,
      "loss": 0.0604,
      "step": 2805
    },
    {
      "epoch": 0.8094775125130532,
      "grad_norm": 0.14233905408675931,
      "learning_rate": 2.1366433327219284e-06,
      "loss": 0.0609,
      "step": 2810
    },
    {
      "epoch": 0.8109178639588059,
      "grad_norm": 0.13538036372546858,
      "learning_rate": 2.105669052239274e-06,
      "loss": 0.0604,
      "step": 2815
    },
    {
      "epoch": 0.8123582154045587,
      "grad_norm": 0.14197120797805796,
      "learning_rate": 2.0748944865333033e-06,
      "loss": 0.0584,
      "step": 2820
    },
    {
      "epoch": 0.8137985668503115,
      "grad_norm": 0.14183713480867427,
      "learning_rate": 2.0443204141545393e-06,
      "loss": 0.0581,
      "step": 2825
    },
    {
      "epoch": 0.8152389182960642,
      "grad_norm": 0.15177244791728164,
      "learning_rate": 2.013947608581327e-06,
      "loss": 0.0571,
      "step": 2830
    },
    {
      "epoch": 0.816679269741817,
      "grad_norm": 0.13725362054115173,
      "learning_rate": 1.983776838200262e-06,
      "loss": 0.0555,
      "step": 2835
    },
    {
      "epoch": 0.8181196211875698,
      "grad_norm": 0.13907159884961337,
      "learning_rate": 1.9538088662867495e-06,
      "loss": 0.0584,
      "step": 2840
    },
    {
      "epoch": 0.8195599726333225,
      "grad_norm": 0.13469702777956086,
      "learning_rate": 1.924044450985706e-06,
      "loss": 0.0608,
      "step": 2845
    },
    {
      "epoch": 0.8210003240790753,
      "grad_norm": 0.13554271157894637,
      "learning_rate": 1.8944843452923546e-06,
      "loss": 0.061,
      "step": 2850
    },
    {
      "epoch": 0.822440675524828,
      "grad_norm": 0.12070016414097579,
      "learning_rate": 1.8651292970332003e-06,
      "loss": 0.057,
      "step": 2855
    },
    {
      "epoch": 0.8238810269705809,
      "grad_norm": 0.14543851459148718,
      "learning_rate": 1.835980048847098e-06,
      "loss": 0.0609,
      "step": 2860
    },
    {
      "epoch": 0.8253213784163336,
      "grad_norm": 0.12894063460571178,
      "learning_rate": 1.8070373381664752e-06,
      "loss": 0.0558,
      "step": 2865
    },
    {
      "epoch": 0.8267617298620864,
      "grad_norm": 0.13829694305046422,
      "learning_rate": 1.77830189719866e-06,
      "loss": 0.0577,
      "step": 2870
    },
    {
      "epoch": 0.8282020813078391,
      "grad_norm": 0.23328526390899323,
      "learning_rate": 1.7497744529073712e-06,
      "loss": 0.0557,
      "step": 2875
    },
    {
      "epoch": 0.8296424327535918,
      "grad_norm": 0.14339826319501542,
      "learning_rate": 1.721455726994321e-06,
      "loss": 0.0617,
      "step": 2880
    },
    {
      "epoch": 0.8310827841993447,
      "grad_norm": 0.13288242940677378,
      "learning_rate": 1.6933464358809593e-06,
      "loss": 0.0531,
      "step": 2885
    },
    {
      "epoch": 0.8325231356450974,
      "grad_norm": 0.16090987532307297,
      "learning_rate": 1.6654472906903486e-06,
      "loss": 0.0557,
      "step": 2890
    },
    {
      "epoch": 0.8339634870908502,
      "grad_norm": 0.13565187898118353,
      "learning_rate": 1.637758997229173e-06,
      "loss": 0.0551,
      "step": 2895
    },
    {
      "epoch": 0.8354038385366029,
      "grad_norm": 0.13398237328307583,
      "learning_rate": 1.6102822559698828e-06,
      "loss": 0.0562,
      "step": 2900
    },
    {
      "epoch": 0.8368441899823557,
      "grad_norm": 0.12547576724680876,
      "learning_rate": 1.5830177620329712e-06,
      "loss": 0.058,
      "step": 2905
    },
    {
      "epoch": 0.8382845414281085,
      "grad_norm": 0.21201898166629826,
      "learning_rate": 1.5559662051694002e-06,
      "loss": 0.0595,
      "step": 2910
    },
    {
      "epoch": 0.8397248928738612,
      "grad_norm": 0.13854983530846096,
      "learning_rate": 1.5291282697431353e-06,
      "loss": 0.059,
      "step": 2915
    },
    {
      "epoch": 0.841165244319614,
      "grad_norm": 0.1412910572198349,
      "learning_rate": 1.502504634713835e-06,
      "loss": 0.057,
      "step": 2920
    },
    {
      "epoch": 0.8426055957653668,
      "grad_norm": 0.13150667676384337,
      "learning_rate": 1.4760959736196834e-06,
      "loss": 0.0566,
      "step": 2925
    },
    {
      "epoch": 0.8440459472111195,
      "grad_norm": 0.12609782628169966,
      "learning_rate": 1.4499029545603472e-06,
      "loss": 0.0571,
      "step": 2930
    },
    {
      "epoch": 0.8454862986568723,
      "grad_norm": 0.1556487705171841,
      "learning_rate": 1.423926240180068e-06,
      "loss": 0.0609,
      "step": 2935
    },
    {
      "epoch": 0.846926650102625,
      "grad_norm": 0.1388357105513542,
      "learning_rate": 1.3981664876509028e-06,
      "loss": 0.0592,
      "step": 2940
    },
    {
      "epoch": 0.8483670015483779,
      "grad_norm": 0.14513822447031188,
      "learning_rate": 1.3726243486560975e-06,
      "loss": 0.062,
      "step": 2945
    },
    {
      "epoch": 0.8498073529941306,
      "grad_norm": 0.14134009678939588,
      "learning_rate": 1.3473004693736037e-06,
      "loss": 0.0605,
      "step": 2950
    },
    {
      "epoch": 0.8512477044398833,
      "grad_norm": 0.13877865523926497,
      "learning_rate": 1.3221954904597256e-06,
      "loss": 0.0567,
      "step": 2955
    },
    {
      "epoch": 0.8526880558856361,
      "grad_norm": 0.1338985016688266,
      "learning_rate": 1.2973100470329159e-06,
      "loss": 0.0592,
      "step": 2960
    },
    {
      "epoch": 0.8541284073313888,
      "grad_norm": 0.13691724800191765,
      "learning_rate": 1.272644768657707e-06,
      "loss": 0.0552,
      "step": 2965
    },
    {
      "epoch": 0.8555687587771417,
      "grad_norm": 0.12943506117621348,
      "learning_rate": 1.248200279328784e-06,
      "loss": 0.0581,
      "step": 2970
    },
    {
      "epoch": 0.8570091102228944,
      "grad_norm": 0.1350768678776799,
      "learning_rate": 1.223977197455204e-06,
      "loss": 0.0596,
      "step": 2975
    },
    {
      "epoch": 0.8584494616686471,
      "grad_norm": 0.14454784322868913,
      "learning_rate": 1.1999761358447403e-06,
      "loss": 0.0595,
      "step": 2980
    },
    {
      "epoch": 0.8598898131143999,
      "grad_norm": 0.1480760428605418,
      "learning_rate": 1.1761977016883897e-06,
      "loss": 0.0559,
      "step": 2985
    },
    {
      "epoch": 0.8613301645601527,
      "grad_norm": 0.1298300829120424,
      "learning_rate": 1.152642496544998e-06,
      "loss": 0.0556,
      "step": 2990
    },
    {
      "epoch": 0.8627705160059055,
      "grad_norm": 0.14190562141399815,
      "learning_rate": 1.1293111163260639e-06,
      "loss": 0.0592,
      "step": 2995
    },
    {
      "epoch": 0.8642108674516582,
      "grad_norm": 0.1340146209373762,
      "learning_rate": 1.1062041512806409e-06,
      "loss": 0.0624,
      "step": 3000
    },
    {
      "epoch": 0.8642108674516582,
      "eval_loss": 0.05916040763258934,
      "eval_runtime": 198.9831,
      "eval_samples_per_second": 9.066,
      "eval_steps_per_second": 2.267,
      "step": 3000
    },
    {
      "epoch": 0.8656512188974109,
      "grad_norm": 0.1359785326923773,
      "learning_rate": 1.0833221859804188e-06,
      "loss": 0.0582,
      "step": 3005
    },
    {
      "epoch": 0.8670915703431638,
      "grad_norm": 0.1273647539914811,
      "learning_rate": 1.0606657993049253e-06,
      "loss": 0.0574,
      "step": 3010
    },
    {
      "epoch": 0.8685319217889165,
      "grad_norm": 0.14728516217233614,
      "learning_rate": 1.0382355644268871e-06,
      "loss": 0.062,
      "step": 3015
    },
    {
      "epoch": 0.8699722732346693,
      "grad_norm": 0.19933620610664268,
      "learning_rate": 1.0160320487977349e-06,
      "loss": 0.0569,
      "step": 3020
    },
    {
      "epoch": 0.871412624680422,
      "grad_norm": 0.15209153121626354,
      "learning_rate": 9.940558141332323e-07,
      "loss": 0.0619,
      "step": 3025
    },
    {
      "epoch": 0.8728529761261747,
      "grad_norm": 0.1310096696794578,
      "learning_rate": 9.723074163992774e-07,
      "loss": 0.0582,
      "step": 3030
    },
    {
      "epoch": 0.8742933275719276,
      "grad_norm": 0.13609449965932524,
      "learning_rate": 9.507874057978339e-07,
      "loss": 0.0569,
      "step": 3035
    },
    {
      "epoch": 0.8757336790176803,
      "grad_norm": 0.13290384955473775,
      "learning_rate": 9.294963267530177e-07,
      "loss": 0.0571,
      "step": 3040
    },
    {
      "epoch": 0.8771740304634331,
      "grad_norm": 0.1343134556624474,
      "learning_rate": 9.084347178973107e-07,
      "loss": 0.0571,
      "step": 3045
    },
    {
      "epoch": 0.8786143819091858,
      "grad_norm": 0.14323917922303897,
      "learning_rate": 8.876031120579454e-07,
      "loss": 0.0568,
      "step": 3050
    },
    {
      "epoch": 0.8800547333549386,
      "grad_norm": 0.13867733291744297,
      "learning_rate": 8.670020362434229e-07,
      "loss": 0.0575,
      "step": 3055
    },
    {
      "epoch": 0.8814950848006914,
      "grad_norm": 0.13288663406354173,
      "learning_rate": 8.466320116301752e-07,
      "loss": 0.0561,
      "step": 3060
    },
    {
      "epoch": 0.8829354362464441,
      "grad_norm": 0.13259296941859702,
      "learning_rate": 8.264935535493879e-07,
      "loss": 0.0555,
      "step": 3065
    },
    {
      "epoch": 0.8843757876921969,
      "grad_norm": 0.13951091711976848,
      "learning_rate": 8.065871714739581e-07,
      "loss": 0.0607,
      "step": 3070
    },
    {
      "epoch": 0.8858161391379497,
      "grad_norm": 0.12832449227606774,
      "learning_rate": 7.869133690056063e-07,
      "loss": 0.058,
      "step": 3075
    },
    {
      "epoch": 0.8872564905837024,
      "grad_norm": 0.1321296955840568,
      "learning_rate": 7.67472643862136e-07,
      "loss": 0.0557,
      "step": 3080
    },
    {
      "epoch": 0.8886968420294552,
      "grad_norm": 0.14306664554533388,
      "learning_rate": 7.482654878648465e-07,
      "loss": 0.0642,
      "step": 3085
    },
    {
      "epoch": 0.8901371934752079,
      "grad_norm": 0.12616475566967347,
      "learning_rate": 7.292923869260837e-07,
      "loss": 0.0609,
      "step": 3090
    },
    {
      "epoch": 0.8915775449209608,
      "grad_norm": 0.13027354471908487,
      "learning_rate": 7.105538210369467e-07,
      "loss": 0.0551,
      "step": 3095
    },
    {
      "epoch": 0.8930178963667135,
      "grad_norm": 0.12728728659733196,
      "learning_rate": 6.920502642551519e-07,
      "loss": 0.0577,
      "step": 3100
    },
    {
      "epoch": 0.8944582478124662,
      "grad_norm": 0.14539496144756697,
      "learning_rate": 6.737821846930403e-07,
      "loss": 0.0594,
      "step": 3105
    },
    {
      "epoch": 0.895898599258219,
      "grad_norm": 0.142416710805929,
      "learning_rate": 6.557500445057252e-07,
      "loss": 0.0618,
      "step": 3110
    },
    {
      "epoch": 0.8973389507039717,
      "grad_norm": 0.1372415423626945,
      "learning_rate": 6.379542998794086e-07,
      "loss": 0.0547,
      "step": 3115
    },
    {
      "epoch": 0.8987793021497246,
      "grad_norm": 0.14646165458637372,
      "learning_rate": 6.203954010198387e-07,
      "loss": 0.0598,
      "step": 3120
    },
    {
      "epoch": 0.9002196535954773,
      "grad_norm": 0.12761881782441253,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.0581,
      "step": 3125
    },
    {
      "epoch": 0.90166000504123,
      "grad_norm": 0.13614699389750304,
      "learning_rate": 5.859899114534662e-07,
      "loss": 0.0556,
      "step": 3130
    },
    {
      "epoch": 0.9031003564869828,
      "grad_norm": 0.1386025331106829,
      "learning_rate": 5.691441911541385e-07,
      "loss": 0.0582,
      "step": 3135
    },
    {
      "epoch": 0.9045407079327356,
      "grad_norm": 0.13232032514165884,
      "learning_rate": 5.525370574144873e-07,
      "loss": 0.0545,
      "step": 3140
    },
    {
      "epoch": 0.9059810593784884,
      "grad_norm": 0.12527580811393457,
      "learning_rate": 5.361689303701767e-07,
      "loss": 0.0533,
      "step": 3145
    },
    {
      "epoch": 0.9074214108242411,
      "grad_norm": 0.13604485662549864,
      "learning_rate": 5.200402241103674e-07,
      "loss": 0.0563,
      "step": 3150
    },
    {
      "epoch": 0.9088617622699939,
      "grad_norm": 0.12602730701179232,
      "learning_rate": 5.041513466672254e-07,
      "loss": 0.0528,
      "step": 3155
    },
    {
      "epoch": 0.9103021137157467,
      "grad_norm": 0.1331744306571229,
      "learning_rate": 4.885027000056075e-07,
      "loss": 0.0596,
      "step": 3160
    },
    {
      "epoch": 0.9117424651614994,
      "grad_norm": 0.13237107440276355,
      "learning_rate": 4.730946800128888e-07,
      "loss": 0.0586,
      "step": 3165
    },
    {
      "epoch": 0.9131828166072522,
      "grad_norm": 0.12937474830598486,
      "learning_rate": 4.5792767648895396e-07,
      "loss": 0.061,
      "step": 3170
    },
    {
      "epoch": 0.9146231680530049,
      "grad_norm": 0.1398098738438078,
      "learning_rate": 4.4300207313632713e-07,
      "loss": 0.0569,
      "step": 3175
    },
    {
      "epoch": 0.9160635194987578,
      "grad_norm": 0.1302735721399194,
      "learning_rate": 4.2831824755046994e-07,
      "loss": 0.0559,
      "step": 3180
    },
    {
      "epoch": 0.9175038709445105,
      "grad_norm": 0.1246210367618306,
      "learning_rate": 4.138765712102299e-07,
      "loss": 0.0564,
      "step": 3185
    },
    {
      "epoch": 0.9189442223902632,
      "grad_norm": 0.13332208166529505,
      "learning_rate": 3.9967740946843523e-07,
      "loss": 0.0554,
      "step": 3190
    },
    {
      "epoch": 0.920384573836016,
      "grad_norm": 0.14610411993837782,
      "learning_rate": 3.8572112154266593e-07,
      "loss": 0.0578,
      "step": 3195
    },
    {
      "epoch": 0.9218249252817687,
      "grad_norm": 0.15226138047700943,
      "learning_rate": 3.7200806050614714e-07,
      "loss": 0.0567,
      "step": 3200
    },
    {
      "epoch": 0.9232652767275216,
      "grad_norm": 0.1308225111781015,
      "learning_rate": 3.585385732788327e-07,
      "loss": 0.0563,
      "step": 3205
    },
    {
      "epoch": 0.9247056281732743,
      "grad_norm": 0.14684465929558974,
      "learning_rate": 3.453130006186234e-07,
      "loss": 0.0581,
      "step": 3210
    },
    {
      "epoch": 0.926145979619027,
      "grad_norm": 0.13266776739181627,
      "learning_rate": 3.3233167711274496e-07,
      "loss": 0.0601,
      "step": 3215
    },
    {
      "epoch": 0.9275863310647798,
      "grad_norm": 0.1358476730566447,
      "learning_rate": 3.1959493116928473e-07,
      "loss": 0.0576,
      "step": 3220
    },
    {
      "epoch": 0.9290266825105326,
      "grad_norm": 0.17850604077128118,
      "learning_rate": 3.0710308500888184e-07,
      "loss": 0.0603,
      "step": 3225
    },
    {
      "epoch": 0.9304670339562854,
      "grad_norm": 0.14331524298007126,
      "learning_rate": 2.948564546565791e-07,
      "loss": 0.0569,
      "step": 3230
    },
    {
      "epoch": 0.9319073854020381,
      "grad_norm": 0.13970764920849238,
      "learning_rate": 2.828553499338227e-07,
      "loss": 0.0555,
      "step": 3235
    },
    {
      "epoch": 0.9333477368477908,
      "grad_norm": 0.14419134050474353,
      "learning_rate": 2.71100074450632e-07,
      "loss": 0.0555,
      "step": 3240
    },
    {
      "epoch": 0.9347880882935437,
      "grad_norm": 0.1461736889168219,
      "learning_rate": 2.595909255979079e-07,
      "loss": 0.0572,
      "step": 3245
    },
    {
      "epoch": 0.9362284397392964,
      "grad_norm": 0.13124715065443018,
      "learning_rate": 2.4832819453992073e-07,
      "loss": 0.0591,
      "step": 3250
    },
    {
      "epoch": 0.9376687911850492,
      "grad_norm": 0.1306820541079532,
      "learning_rate": 2.3731216620693554e-07,
      "loss": 0.0605,
      "step": 3255
    },
    {
      "epoch": 0.9391091426308019,
      "grad_norm": 0.13542742296333235,
      "learning_rate": 2.2654311928800965e-07,
      "loss": 0.0536,
      "step": 3260
    },
    {
      "epoch": 0.9405494940765546,
      "grad_norm": 0.14758285988575107,
      "learning_rate": 2.1602132622393745e-07,
      "loss": 0.0606,
      "step": 3265
    },
    {
      "epoch": 0.9419898455223075,
      "grad_norm": 0.14532495800717846,
      "learning_rate": 2.0574705320036025e-07,
      "loss": 0.0639,
      "step": 3270
    },
    {
      "epoch": 0.9434301969680602,
      "grad_norm": 0.12954948070786576,
      "learning_rate": 1.9572056014103281e-07,
      "loss": 0.0579,
      "step": 3275
    },
    {
      "epoch": 0.944870548413813,
      "grad_norm": 0.14435133264907077,
      "learning_rate": 1.8594210070124852e-07,
      "loss": 0.0623,
      "step": 3280
    },
    {
      "epoch": 0.9463108998595657,
      "grad_norm": 0.13182234725830524,
      "learning_rate": 1.7641192226141913e-07,
      "loss": 0.0541,
      "step": 3285
    },
    {
      "epoch": 0.9477512513053185,
      "grad_norm": 1.4633024268191004,
      "learning_rate": 1.671302659208185e-07,
      "loss": 0.0572,
      "step": 3290
    },
    {
      "epoch": 0.9491916027510713,
      "grad_norm": 0.1503610098957111,
      "learning_rate": 1.58097366491482e-07,
      "loss": 0.061,
      "step": 3295
    },
    {
      "epoch": 0.950631954196824,
      "grad_norm": 0.1443388849896849,
      "learning_rate": 1.4931345249226792e-07,
      "loss": 0.0555,
      "step": 3300
    },
    {
      "epoch": 0.9520723056425768,
      "grad_norm": 0.12701776281849797,
      "learning_rate": 1.407787461430743e-07,
      "loss": 0.0628,
      "step": 3305
    },
    {
      "epoch": 0.9535126570883296,
      "grad_norm": 0.13129256088024213,
      "learning_rate": 1.324934633592201e-07,
      "loss": 0.0541,
      "step": 3310
    },
    {
      "epoch": 0.9549530085340823,
      "grad_norm": 0.1327785363196553,
      "learning_rate": 1.2445781374597842e-07,
      "loss": 0.0588,
      "step": 3315
    },
    {
      "epoch": 0.9563933599798351,
      "grad_norm": 0.13315365290665568,
      "learning_rate": 1.1667200059327644e-07,
      "loss": 0.0558,
      "step": 3320
    },
    {
      "epoch": 0.9578337114255878,
      "grad_norm": 0.14439945965305848,
      "learning_rate": 1.0913622087055264e-07,
      "loss": 0.0597,
      "step": 3325
    },
    {
      "epoch": 0.9592740628713406,
      "grad_norm": 0.13166148371196648,
      "learning_rate": 1.0185066522177545e-07,
      "loss": 0.0533,
      "step": 3330
    },
    {
      "epoch": 0.9607144143170934,
      "grad_norm": 0.1400270351664016,
      "learning_rate": 9.481551796061472e-08,
      "loss": 0.0591,
      "step": 3335
    },
    {
      "epoch": 0.9621547657628461,
      "grad_norm": 0.14119139667530914,
      "learning_rate": 8.803095706578335e-08,
      "loss": 0.0583,
      "step": 3340
    },
    {
      "epoch": 0.9635951172085989,
      "grad_norm": 0.13763855728395058,
      "learning_rate": 8.149715417653414e-08,
      "loss": 0.0567,
      "step": 3345
    },
    {
      "epoch": 0.9650354686543516,
      "grad_norm": 0.13842628044965613,
      "learning_rate": 7.521427458831776e-08,
      "loss": 0.0588,
      "step": 3350
    },
    {
      "epoch": 0.9664758201001045,
      "grad_norm": 0.12724872379714094,
      "learning_rate": 6.918247724859939e-08,
      "loss": 0.0541,
      "step": 3355
    },
    {
      "epoch": 0.9679161715458572,
      "grad_norm": 0.14898075476210107,
      "learning_rate": 6.340191475283753e-08,
      "loss": 0.06,
      "step": 3360
    },
    {
      "epoch": 0.9693565229916099,
      "grad_norm": 0.1347911477145421,
      "learning_rate": 5.787273334062593e-08,
      "loss": 0.0596,
      "step": 3365
    },
    {
      "epoch": 0.9707968744373627,
      "grad_norm": 0.13079385477539324,
      "learning_rate": 5.259507289199328e-08,
      "loss": 0.0619,
      "step": 3370
    },
    {
      "epoch": 0.9722372258831155,
      "grad_norm": 0.14313408319905535,
      "learning_rate": 4.756906692386043e-08,
      "loss": 0.0597,
      "step": 3375
    },
    {
      "epoch": 0.9736775773288683,
      "grad_norm": 0.13044063273707465,
      "learning_rate": 4.2794842586670884e-08,
      "loss": 0.0589,
      "step": 3380
    },
    {
      "epoch": 0.975117928774621,
      "grad_norm": 0.148098221291949,
      "learning_rate": 3.827252066116338e-08,
      "loss": 0.0547,
      "step": 3385
    },
    {
      "epoch": 0.9765582802203737,
      "grad_norm": 0.1344069973111941,
      "learning_rate": 3.400221555532768e-08,
      "loss": 0.0544,
      "step": 3390
    },
    {
      "epoch": 0.9779986316661265,
      "grad_norm": 0.12133249021447175,
      "learning_rate": 2.998403530150018e-08,
      "loss": 0.06,
      "step": 3395
    },
    {
      "epoch": 0.9794389831118793,
      "grad_norm": 0.13732750082544246,
      "learning_rate": 2.6218081553638363e-08,
      "loss": 0.0605,
      "step": 3400
    },
    {
      "epoch": 0.9808793345576321,
      "grad_norm": 0.13176353353269668,
      "learning_rate": 2.2704449584745046e-08,
      "loss": 0.0592,
      "step": 3405
    },
    {
      "epoch": 0.9823196860033848,
      "grad_norm": 0.13323188939537126,
      "learning_rate": 1.9443228284455882e-08,
      "loss": 0.0568,
      "step": 3410
    },
    {
      "epoch": 0.9837600374491375,
      "grad_norm": 0.1207130237315737,
      "learning_rate": 1.6434500156800037e-08,
      "loss": 0.0566,
      "step": 3415
    },
    {
      "epoch": 0.9852003888948904,
      "grad_norm": 0.12840927154274398,
      "learning_rate": 1.3678341318100751e-08,
      "loss": 0.0547,
      "step": 3420
    },
    {
      "epoch": 0.9866407403406431,
      "grad_norm": 0.15035865147658953,
      "learning_rate": 1.1174821495059106e-08,
      "loss": 0.0553,
      "step": 3425
    },
    {
      "epoch": 0.9880810917863959,
      "grad_norm": 0.1435557916571953,
      "learning_rate": 8.924004022986543e-09,
      "loss": 0.0555,
      "step": 3430
    },
    {
      "epoch": 0.9895214432321486,
      "grad_norm": 0.12838967785145278,
      "learning_rate": 6.9259458442005875e-09,
      "loss": 0.0614,
      "step": 3435
    },
    {
      "epoch": 0.9909617946779014,
      "grad_norm": 0.14083462190226395,
      "learning_rate": 5.180697506587118e-09,
      "loss": 0.0543,
      "step": 3440
    },
    {
      "epoch": 0.9924021461236542,
      "grad_norm": 0.12037599402495351,
      "learning_rate": 3.688303162322493e-09,
      "loss": 0.0568,
      "step": 3445
    },
    {
      "epoch": 0.9938424975694069,
      "grad_norm": 0.12739736699073684,
      "learning_rate": 2.44880056675334e-09,
      "loss": 0.0538,
      "step": 3450
    },
    {
      "epoch": 0.9952828490151597,
      "grad_norm": 0.1356848847097788,
      "learning_rate": 1.4622210774428714e-09,
      "loss": 0.0581,
      "step": 3455
    },
    {
      "epoch": 0.9967232004609125,
      "grad_norm": 0.1406371102406219,
      "learning_rate": 7.285896533770765e-10,
      "loss": 0.0577,
      "step": 3460
    },
    {
      "epoch": 0.9981635519066653,
      "grad_norm": 0.13450694809907887,
      "learning_rate": 2.479248543363344e-10,
      "loss": 0.0566,
      "step": 3465
    },
    {
      "epoch": 0.999603903352418,
      "grad_norm": 0.14028964451278583,
      "learning_rate": 2.0238840421349382e-11,
      "loss": 0.0579,
      "step": 3470
    },
    {
      "epoch": 0.9998919736415686,
      "step": 3471,
      "total_flos": 2.0789981999005696e+16,
      "train_loss": 0.0,
      "train_runtime": 2.6266,
      "train_samples_per_second": 15511.09,
      "train_steps_per_second": 242.135
    }
  ],
  "logging_steps": 5,
  "max_steps": 636,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.0789981999005696e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}