diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,31636 @@
+{
+    "epoch": 0.9999831931831551,
+    "global_step": 22312,
+    "max_steps": 22312,
+    "logging_steps": 5,
+    "eval_steps": 500,
+    "save_steps": 500,
+    "train_batch_size": 32,
+    "num_train_epochs": 1,
+    "num_input_tokens_seen": 0,
+    "total_flos": 1.364299850107899e+19,
+    "log_history": [
+        {
+            "loss": 14.6143,
+            "grad_norm": 3.2611544132232666,
+            "learning_rate": 0.0005,
+            "epoch": 0.00022409089126549728,
+            "step": 5
+        },
+        {
+            "loss": 14.4676,
+            "grad_norm": 2.4582467079162598,
+            "learning_rate": 0.0005,
+            "epoch": 0.00044818178253099456,
+            "step": 10
+        },
+        {
+            "loss": 14.4128,
+            "grad_norm": 2.327120065689087,
+            "learning_rate": 0.0005,
+            "epoch": 0.0006722726737964918,
+            "step": 15
+        },
+        {
+            "loss": 14.3947,
+            "grad_norm": 2.182023048400879,
+            "learning_rate": 0.0005,
+            "epoch": 0.0008963635650619891,
+            "step": 20
+        },
+        {
+            "loss": 14.3541,
+            "grad_norm": 2.0157508850097656,
+            "learning_rate": 0.0005,
+            "epoch": 0.0011204544563274863,
+            "step": 25
+        },
+        {
+            "loss": 14.2979,
+            "grad_norm": 2.0835866928100586,
+            "learning_rate": 0.0005,
+            "epoch": 0.0013445453475929836,
+            "step": 30
+        },
+        {
+            "loss": 14.3136,
+            "grad_norm": 2.124310255050659,
+            "learning_rate": 0.0005,
+            "epoch": 0.001568636238858481,
+            "step": 35
+        },
+        {
+            "loss": 14.3155,
+            "grad_norm": 2.100733757019043,
+            "learning_rate": 0.0005,
+            "epoch": 0.0017927271301239783,
+            "step": 40
+        },
+        {
+            "loss": 14.2023,
+            "grad_norm": 2.044724702835083,
+            "learning_rate": 0.0005,
+            "epoch": 0.0020168180213894758,
+            "step": 45
+        },
+        {
+            "loss": 14.3092,
+            "grad_norm": 2.281487464904785,
+            "learning_rate": 0.0005,
+            "epoch": 0.0022409089126549727,
+            "step": 50
+        },
+        {
+            "loss": 14.1618,
+            "grad_norm": 1.9946085214614868,
+            "learning_rate": 0.0005,
+            "epoch": 0.00246499980392047,
+            "step": 55
+        },
+        {
+            "loss": 14.3028,
+            "grad_norm": 2.176398754119873,
+            "learning_rate": 0.0005,
+            "epoch": 0.0026890906951859673,
+            "step": 60
+        },
+        {
+            "loss": 14.3326,
+            "grad_norm": 1.999497890472412,
+            "learning_rate": 0.0005,
+            "epoch": 0.0029131815864514646,
+            "step": 65
+        },
+        {
+            "loss": 14.4064,
+            "grad_norm": 1.9895137548446655,
+            "learning_rate": 0.0005,
+            "epoch": 0.003137272477716962,
+            "step": 70
+        },
+        {
+            "loss": 14.3487,
+            "grad_norm": 1.9479091167449951,
+            "learning_rate": 0.0005,
+            "epoch": 0.003361363368982459,
+            "step": 75
+        },
+        {
+            "loss": 14.2659,
+            "grad_norm": 2.2668464183807373,
+            "learning_rate": 0.0005,
+            "epoch": 0.0035854542602479565,
+            "step": 80
+        },
+        {
+            "loss": 14.2306,
+            "grad_norm": 2.197277069091797,
+            "learning_rate": 0.0005,
+            "epoch": 0.003809545151513454,
+            "step": 85
+        },
+        {
+            "loss": 14.2925,
+            "grad_norm": 2.349569797515869,
+            "learning_rate": 0.0005,
+            "epoch": 0.0040336360427789516,
+            "step": 90
+        },
+        {
+            "loss": 14.2346,
+            "grad_norm": 2.1121737957000732,
+            "learning_rate": 0.0005,
+            "epoch": 0.004257726934044448,
+            "step": 95
+        },
+        {
+            "loss": 14.2444,
+            "grad_norm": 2.3515448570251465,
+            "learning_rate": 0.0005,
+            "epoch": 0.004481817825309945,
+            "step": 100
+        },
+        {
+            "loss": 14.2322,
+            "grad_norm": 2.0645346641540527,
+            "learning_rate": 0.0005,
+            "epoch": 0.004705908716575443,
+            "step": 105
+        },
+        {
+            "loss": 14.2183,
+            "grad_norm": 2.159444808959961,
+            "learning_rate": 0.0005,
+            "epoch": 0.00492999960784094,
+            "step": 110
+        },
+        {
+            "loss": 14.3575,
+            "grad_norm": 2.2058215141296387,
+            "learning_rate": 0.0005,
+            "epoch": 0.005154090499106438,
+            "step": 115
+        },
+        {
+            "loss": 14.3028,
+            "grad_norm": 2.4209940433502197,
+            "learning_rate": 0.0005,
+            "epoch": 0.0053781813903719345,
+            "step": 120
+        },
+        {
+            "loss": 14.2203,
+            "grad_norm": 2.0729739665985107,
+            "learning_rate": 0.0005,
+            "epoch": 0.005602272281637432,
+            "step": 125
+        },
+        {
+            "loss": 14.2853,
+            "grad_norm": 2.0709404945373535,
+            "learning_rate": 0.0005,
+            "epoch": 0.005826363172902929,
+            "step": 130
+        },
+        {
+            "loss": 14.3667,
+            "grad_norm": 1.994726300239563,
+            "learning_rate": 0.0005,
+            "epoch": 0.006050454064168427,
+            "step": 135
+        },
+        {
+            "loss": 14.2726,
+            "grad_norm": 2.0086257457733154,
+            "learning_rate": 0.0005,
+            "epoch": 0.006274544955433924,
+            "step": 140
+        },
+        {
+            "loss": 14.3021,
+            "grad_norm": 2.103579044342041,
+            "learning_rate": 0.0005,
+            "epoch": 0.0064986358466994215,
+            "step": 145
+        },
+        {
+            "loss": 14.2639,
+            "grad_norm": 2.1623342037200928,
+            "learning_rate": 0.0005,
+            "epoch": 0.006722726737964918,
+            "step": 150
+        },
+        {
+            "loss": 14.3839,
+            "grad_norm": 2.345151662826538,
+            "learning_rate": 0.0005,
+            "epoch": 0.006946817629230416,
+            "step": 155
+        },
+        {
+            "loss": 14.3292,
+            "grad_norm": 2.145923137664795,
+            "learning_rate": 0.0005,
+            "epoch": 0.007170908520495913,
+            "step": 160
+        },
+        {
+            "loss": 14.2732,
+            "grad_norm": 2.088350296020508,
+            "learning_rate": 0.0005,
+            "epoch": 0.007394999411761411,
+            "step": 165
+        },
+        {
+            "loss": 14.25,
+            "grad_norm": 2.0546774864196777,
+            "learning_rate": 0.0005,
+            "epoch": 0.007619090303026908,
+            "step": 170
+        },
+        {
+            "loss": 14.3347,
+            "grad_norm": 1.9553978443145752,
+            "learning_rate": 0.0005,
+            "epoch": 0.007843181194292404,
+            "step": 175
+        },
+        {
+            "loss": 14.3019,
+            "grad_norm": 2.0136640071868896,
+            "learning_rate": 0.0005,
+            "epoch": 0.008067272085557903,
+            "step": 180
+        },
+        {
+            "loss": 14.3584,
+            "grad_norm": 2.148486614227295,
+            "learning_rate": 0.0005,
+            "epoch": 0.0082913629768234,
+            "step": 185
+        },
+        {
+            "loss": 14.2891,
+            "grad_norm": 2.0075933933258057,
+            "learning_rate": 0.0005,
+            "epoch": 0.008515453868088897,
+            "step": 190
+        },
+        {
+            "loss": 14.2923,
+            "grad_norm": 1.9557173252105713,
+            "learning_rate": 0.0005,
+            "epoch": 0.008739544759354394,
+            "step": 195
+        },
+        {
+            "loss": 14.386,
+            "grad_norm": 2.001337766647339,
+            "learning_rate": 0.0005,
+            "epoch": 0.00896363565061989,
+            "step": 200
+        },
+        {
+            "loss": 14.2192,
+            "grad_norm": 2.044628381729126,
+            "learning_rate": 0.0005,
+            "epoch": 0.00918772654188539,
+            "step": 205
+        },
+        {
+            "loss": 14.3734,
+            "grad_norm": 2.002190351486206,
+            "learning_rate": 0.0005,
+            "epoch": 0.009411817433150886,
+            "step": 210
+        },
+        {
+            "loss": 14.3697,
+            "grad_norm": 2.089207172393799,
+            "learning_rate": 0.0005,
+            "epoch": 0.009635908324416383,
+            "step": 215
+        },
+        {
+            "loss": 14.2986,
+            "grad_norm": 1.9658682346343994,
+            "learning_rate": 0.0005,
+            "epoch": 0.00985999921568188,
+            "step": 220
+        },
+        {
+            "loss": 14.3278,
+            "grad_norm": 2.009406566619873,
+            "learning_rate": 0.0005,
+            "epoch": 0.010084090106947378,
+            "step": 225
+        },
+        {
+            "loss": 14.347,
+            "grad_norm": 2.0346028804779053,
+            "learning_rate": 0.0005,
+            "epoch": 0.010308180998212875,
+            "step": 230
+        },
+        {
+            "loss": 14.3171,
+            "grad_norm": 2.057018756866455,
+            "learning_rate": 0.0005,
+            "epoch": 0.010532271889478372,
+            "step": 235
+        },
+        {
+            "loss": 14.3144,
+            "grad_norm": 2.078429937362671,
+            "learning_rate": 0.0005,
+            "epoch": 0.010756362780743869,
+            "step": 240
+        },
+        {
+            "loss": 14.2827,
+            "grad_norm": 2.3576812744140625,
+            "learning_rate": 0.0005,
+            "epoch": 0.010980453672009368,
+            "step": 245
+        },
+        {
+            "loss": 14.2424,
+            "grad_norm": 2.1042275428771973,
+            "learning_rate": 0.0005,
+            "epoch": 0.011204544563274865,
+            "step": 250
+        },
+        {
+            "loss": 14.2774,
+            "grad_norm": 2.023210048675537,
+            "learning_rate": 0.0005,
+            "epoch": 0.011428635454540361,
+            "step": 255
+        },
+        {
+            "loss": 14.2762,
+            "grad_norm": 1.9983354806900024,
+            "learning_rate": 0.0005,
+            "epoch": 0.011652726345805858,
+            "step": 260
+        },
+        {
+            "loss": 14.3338,
+            "grad_norm": 1.9574092626571655,
+            "learning_rate": 0.0005,
+            "epoch": 0.011876817237071357,
+            "step": 265
+        },
+        {
+            "loss": 14.2085,
+            "grad_norm": 1.9682737588882446,
+            "learning_rate": 0.0005,
+            "epoch": 0.012100908128336854,
+            "step": 270
+        },
+        {
+            "loss": 14.2391,
+            "grad_norm": 2.040682077407837,
+            "learning_rate": 0.0005,
+            "epoch": 0.01232499901960235,
+            "step": 275
+        },
+        {
+            "loss": 14.2422,
+            "grad_norm": 2.072660446166992,
+            "learning_rate": 0.0005,
+            "epoch": 0.012549089910867848,
+            "step": 280
+        },
+        {
+            "loss": 14.2232,
+            "grad_norm": 1.9866284132003784,
+            "learning_rate": 0.0005,
+            "epoch": 0.012773180802133344,
+            "step": 285
+        },
+        {
+            "loss": 14.3504,
+            "grad_norm": 2.0564022064208984,
+            "learning_rate": 0.0005,
+            "epoch": 0.012997271693398843,
+            "step": 290
+        },
+        {
+            "loss": 14.3313,
+            "grad_norm": 2.2206408977508545,
+            "learning_rate": 0.0005,
+            "epoch": 0.01322136258466434,
+            "step": 295
+        },
+        {
+            "loss": 14.2808,
+            "grad_norm": 2.1048293113708496,
+            "learning_rate": 0.0005,
+            "epoch": 0.013445453475929837,
+            "step": 300
+        },
+        {
+            "loss": 14.2852,
+            "grad_norm": 1.9940338134765625,
+            "learning_rate": 0.0005,
+            "epoch": 0.013669544367195334,
+            "step": 305
+        },
+        {
+            "loss": 14.2704,
+            "grad_norm": 2.0356545448303223,
+            "learning_rate": 0.0005,
+            "epoch": 0.013893635258460832,
+            "step": 310
+        },
+        {
+            "loss": 14.3152,
+            "grad_norm": 2.060826063156128,
+            "learning_rate": 0.0005,
+            "epoch": 0.01411772614972633,
+            "step": 315
+        },
+        {
+            "loss": 14.158,
+            "grad_norm": 2.0694780349731445,
+            "learning_rate": 0.0005,
+            "epoch": 0.014341817040991826,
+            "step": 320
+        },
+        {
+            "loss": 14.385,
+            "grad_norm": 2.0996909141540527,
+            "learning_rate": 0.0005,
+            "epoch": 0.014565907932257323,
+            "step": 325
+        },
+        {
+            "loss": 14.2717,
+            "grad_norm": 2.0667507648468018,
+            "learning_rate": 0.0005,
+            "epoch": 0.014789998823522821,
+            "step": 330
+        },
+        {
+            "loss": 14.2032,
+            "grad_norm": 2.118894577026367,
+            "learning_rate": 0.0005,
+            "epoch": 0.015014089714788318,
+            "step": 335
+        },
+        {
+            "loss": 14.3441,
+            "grad_norm": 2.37672758102417,
+            "learning_rate": 0.0005,
+            "epoch": 0.015238180606053815,
+            "step": 340
+        },
+        {
+            "loss": 14.2123,
+            "grad_norm": 2.0683722496032715,
+            "learning_rate": 0.0005,
+            "epoch": 0.015462271497319312,
+            "step": 345
+        },
+        {
+            "loss": 14.2399,
+            "grad_norm": 1.9291362762451172,
+            "learning_rate": 0.0005,
+            "epoch": 0.01568636238858481,
+            "step": 350
+        },
+        {
+            "loss": 14.2786,
+            "grad_norm": 2.073303461074829,
+            "learning_rate": 0.0005,
+            "epoch": 0.015910453279850308,
+            "step": 355
+        },
+        {
+            "loss": 14.1956,
+            "grad_norm": 2.015531539916992,
+            "learning_rate": 0.0005,
+            "epoch": 0.016134544171115806,
+            "step": 360
+        },
+        {
+            "loss": 14.2929,
+            "grad_norm": 1.978757381439209,
+            "learning_rate": 0.0005,
+            "epoch": 0.0163586350623813,
+            "step": 365
+        },
+        {
+            "loss": 14.3134,
+            "grad_norm": 2.015641927719116,
+            "learning_rate": 0.0005,
+            "epoch": 0.0165827259536468,
+            "step": 370
+        },
+        {
+            "loss": 14.316,
+            "grad_norm": 2.088158130645752,
+            "learning_rate": 0.0005,
+            "epoch": 0.016806816844912295,
+            "step": 375
+        },
+        {
+            "loss": 14.2379,
+            "grad_norm": 2.0318617820739746,
+            "learning_rate": 0.0005,
+            "epoch": 0.017030907736177794,
+            "step": 380
+        },
+        {
+            "loss": 14.3892,
+            "grad_norm": 1.9806574583053589,
+            "learning_rate": 0.0005,
+            "epoch": 0.017254998627443292,
+            "step": 385
+        },
+        {
+            "loss": 14.2584,
+            "grad_norm": 1.9797101020812988,
+            "learning_rate": 0.0005,
+            "epoch": 0.017479089518708787,
+            "step": 390
+        },
+        {
+            "loss": 14.227,
+            "grad_norm": 1.906246542930603,
+            "learning_rate": 0.0005,
+            "epoch": 0.017703180409974286,
+            "step": 395
+        },
+        {
+            "loss": 14.3051,
+            "grad_norm": 2.0552897453308105,
+            "learning_rate": 0.0005,
+            "epoch": 0.01792727130123978,
+            "step": 400
+        },
+        {
+            "loss": 14.2216,
+            "grad_norm": 2.062342405319214,
+            "learning_rate": 0.0005,
+            "epoch": 0.01815136219250528,
+            "step": 405
+        },
+        {
+            "loss": 14.2193,
+            "grad_norm": 1.9998198747634888,
+            "learning_rate": 0.0005,
+            "epoch": 0.01837545308377078,
+            "step": 410
+        },
+        {
+            "loss": 14.2426,
+            "grad_norm": 2.0867531299591064,
+            "learning_rate": 0.0005,
+            "epoch": 0.018599543975036274,
+            "step": 415
+        },
+        {
+            "loss": 14.3968,
+            "grad_norm": 2.1598596572875977,
+            "learning_rate": 0.0005,
+            "epoch": 0.018823634866301772,
+            "step": 420
+        },
+        {
+            "loss": 14.2109,
+            "grad_norm": 2.1043760776519775,
+            "learning_rate": 0.0005,
+            "epoch": 0.01904772575756727,
+            "step": 425
+        },
+        {
+            "loss": 14.2744,
+            "grad_norm": 2.138683557510376,
+            "learning_rate": 0.0005,
+            "epoch": 0.019271816648832766,
+            "step": 430
+        },
+        {
+            "loss": 14.1959,
+            "grad_norm": 2.0307869911193848,
+            "learning_rate": 0.0005,
+            "epoch": 0.019495907540098265,
+            "step": 435
+        },
+        {
+            "loss": 14.2097,
+            "grad_norm": 1.9247474670410156,
+            "learning_rate": 0.0005,
+            "epoch": 0.01971999843136376,
+            "step": 440
+        },
+        {
+            "loss": 14.2815,
+            "grad_norm": 2.0889439582824707,
+            "learning_rate": 0.0005,
+            "epoch": 0.01994408932262926,
+            "step": 445
+        },
+        {
+            "loss": 14.1621,
+            "grad_norm": 2.0505826473236084,
+            "learning_rate": 0.0005,
+            "epoch": 0.020168180213894757,
+            "step": 450
+        },
+        {
+            "loss": 14.3243,
+            "grad_norm": 2.0084340572357178,
+            "learning_rate": 0.0005,
+            "epoch": 0.020392271105160252,
+            "step": 455
+        },
+        {
+            "loss": 14.2527,
+            "grad_norm": 1.934202790260315,
+            "learning_rate": 0.0005,
+            "epoch": 0.02061636199642575,
+            "step": 460
+        },
+        {
+            "loss": 14.2158,
+            "grad_norm": 2.235403537750244,
+            "learning_rate": 0.0005,
+            "epoch": 0.02084045288769125,
+            "step": 465
+        },
+        {
+            "loss": 14.3807,
+            "grad_norm": 2.1436288356781006,
+            "learning_rate": 0.0005,
+            "epoch": 0.021064543778956744,
+            "step": 470
+        },
+        {
+            "loss": 14.2475,
+            "grad_norm": 2.0434954166412354,
+            "learning_rate": 0.0005,
+            "epoch": 0.021288634670222243,
+            "step": 475
+        },
+        {
+            "loss": 14.3062,
+            "grad_norm": 2.029393434524536,
+            "learning_rate": 0.0005,
+            "epoch": 0.021512725561487738,
+            "step": 480
+        },
+        {
+            "loss": 14.2349,
+            "grad_norm": 2.0223193168640137,
+            "learning_rate": 0.0005,
+            "epoch": 0.021736816452753237,
+            "step": 485
+        },
+        {
+            "loss": 14.2309,
+            "grad_norm": 1.913985013961792,
+            "learning_rate": 0.0005,
+            "epoch": 0.021960907344018735,
+            "step": 490
+        },
+        {
+            "loss": 14.3093,
+            "grad_norm": 2.0677883625030518,
+            "learning_rate": 0.0005,
+            "epoch": 0.02218499823528423,
+            "step": 495
+        },
+        {
+            "loss": 14.3713,
+            "grad_norm": 2.2938449382781982,
+            "learning_rate": 0.0005,
+            "epoch": 0.02240908912654973,
+            "step": 500
+        },
+        {
+            "eval_loss": 1.7718801498413086,
+            "eval_runtime": 18.8403,
+            "eval_samples_per_second": 869.624,
+            "eval_steps_per_second": 7.802,
+            "epoch": 0.02240908912654973,
+            "step": 500
+        },
+        {
+            "loss": 14.2151,
+            "grad_norm": 2.096050500869751,
+            "learning_rate": 0.0005,
+            "epoch": 0.022633180017815224,
+            "step": 505
+        },
+        {
+            "loss": 14.3219,
+            "grad_norm": 2.2422337532043457,
+            "learning_rate": 0.0005,
+            "epoch": 0.022857270909080723,
+            "step": 510
+        },
+        {
+            "loss": 14.2732,
+            "grad_norm": 2.0611579418182373,
+            "learning_rate": 0.0005,
+            "epoch": 0.02308136180034622,
+            "step": 515
+        },
+        {
+            "loss": 14.2907,
+            "grad_norm": 1.9564974308013916,
+            "learning_rate": 0.0005,
+            "epoch": 0.023305452691611717,
+            "step": 520
+        },
+        {
+            "loss": 14.2737,
+            "grad_norm": 2.0616042613983154,
+            "learning_rate": 0.0005,
+            "epoch": 0.023529543582877215,
+            "step": 525
+        },
+        {
+            "loss": 14.32,
+            "grad_norm": 2.1140406131744385,
+            "learning_rate": 0.0005,
+            "epoch": 0.023753634474142714,
+            "step": 530
+        },
+        {
+            "loss": 14.2913,
+            "grad_norm": 1.8935840129852295,
+            "learning_rate": 0.0005,
+            "epoch": 0.02397772536540821,
+            "step": 535
+        },
+        {
+            "loss": 14.3307,
+            "grad_norm": 2.028771162033081,
+            "learning_rate": 0.0005,
+            "epoch": 0.024201816256673708,
+            "step": 540
+        },
+        {
+            "loss": 14.2717,
+            "grad_norm": 1.976778507232666,
+            "learning_rate": 0.0005,
+            "epoch": 0.024425907147939203,
+            "step": 545
+        },
+        {
+            "loss": 14.2348,
+            "grad_norm": 2.082973003387451,
+            "learning_rate": 0.0005,
+            "epoch": 0.0246499980392047,
+            "step": 550
+        },
+        {
+            "loss": 14.2249,
+            "grad_norm": 2.036085367202759,
+            "learning_rate": 0.0005,
+            "epoch": 0.0248740889304702,
+            "step": 555
+        },
+        {
+            "loss": 14.2696,
+            "grad_norm": 1.8578675985336304,
+            "learning_rate": 0.0005,
+            "epoch": 0.025098179821735695,
+            "step": 560
+        },
+        {
+            "loss": 14.2171,
+            "grad_norm": 2.2652697563171387,
+            "learning_rate": 0.0005,
+            "epoch": 0.025322270713001194,
+            "step": 565
+        },
+        {
+            "loss": 14.353,
+            "grad_norm": 2.3761842250823975,
+            "learning_rate": 0.0005,
+            "epoch": 0.02554636160426669,
+            "step": 570
+        },
+        {
+            "loss": 14.3057,
+            "grad_norm": 1.9577360153198242,
+            "learning_rate": 0.0005,
+            "epoch": 0.025770452495532187,
+            "step": 575
+        },
+        {
+            "loss": 14.3683,
+            "grad_norm": 1.9947212934494019,
+            "learning_rate": 0.0005,
+            "epoch": 0.025994543386797686,
+            "step": 580
+        },
+        {
+            "loss": 14.291,
+            "grad_norm": 2.12111234664917,
+            "learning_rate": 0.0005,
+            "epoch": 0.02621863427806318,
+            "step": 585
+        },
+        {
+            "loss": 14.3058,
+            "grad_norm": 2.155707836151123,
+            "learning_rate": 0.0005,
+            "epoch": 0.02644272516932868,
+            "step": 590
+        },
+        {
+            "loss": 14.2909,
+            "grad_norm": 2.0623831748962402,
+            "learning_rate": 0.0005,
+            "epoch": 0.02666681606059418,
+            "step": 595
+        },
+        {
+            "loss": 14.3209,
+            "grad_norm": 2.009176254272461,
+            "learning_rate": 0.0005,
+            "epoch": 0.026890906951859674,
+            "step": 600
+        },
+        {
+            "loss": 14.2734,
+            "grad_norm": 2.1131885051727295,
+            "learning_rate": 0.0005,
+            "epoch": 0.027114997843125172,
+            "step": 605
+        },
+        {
+            "loss": 14.3535,
+            "grad_norm": 2.228571653366089,
+            "learning_rate": 0.0005,
+            "epoch": 0.027339088734390667,
+            "step": 610
+        },
+        {
+            "loss": 14.2595,
+            "grad_norm": 2.2658498287200928,
+            "learning_rate": 0.0005,
+            "epoch": 0.027563179625656166,
+            "step": 615
+        },
+        {
+            "loss": 14.2853,
+            "grad_norm": 2.1453394889831543,
+            "learning_rate": 0.0005,
+            "epoch": 0.027787270516921665,
+            "step": 620
+        },
+        {
+            "loss": 14.383,
+            "grad_norm": 1.982365369796753,
+            "learning_rate": 0.0005,
+            "epoch": 0.02801136140818716,
+            "step": 625
+        },
+        {
+            "loss": 14.3929,
+            "grad_norm": 2.0140199661254883,
+            "learning_rate": 0.0005,
+            "epoch": 0.02823545229945266,
+            "step": 630
+        },
+        {
+            "loss": 14.1989,
+            "grad_norm": 1.998089075088501,
+            "learning_rate": 0.0005,
+            "epoch": 0.028459543190718157,
+            "step": 635
+        },
+        {
+            "loss": 14.2745,
+            "grad_norm": 2.0165891647338867,
+            "learning_rate": 0.0005,
+            "epoch": 0.028683634081983652,
+            "step": 640
+        },
+        {
+            "loss": 14.2618,
+            "grad_norm": 2.0754010677337646,
+            "learning_rate": 0.0005,
+            "epoch": 0.02890772497324915,
+            "step": 645
+        },
+        {
+            "loss": 14.2327,
+            "grad_norm": 2.3930583000183105,
+            "learning_rate": 0.0005,
+            "epoch": 0.029131815864514646,
+            "step": 650
+        },
+        {
+            "loss": 14.2808,
+            "grad_norm": 2.123385190963745,
+            "learning_rate": 0.0005,
+            "epoch": 0.029355906755780144,
+            "step": 655
+        },
+        {
+            "loss": 14.1984,
+            "grad_norm": 1.9736683368682861,
+            "learning_rate": 0.0005,
+            "epoch": 0.029579997647045643,
+            "step": 660
+        },
+        {
+            "loss": 14.3508,
+            "grad_norm": 2.030437707901001,
+            "learning_rate": 0.0005,
+            "epoch": 0.029804088538311138,
+            "step": 665
+        },
+        {
+            "loss": 14.131,
+            "grad_norm": 1.8928433656692505,
+            "learning_rate": 0.0005,
+            "epoch": 0.030028179429576637,
+            "step": 670
+        },
+        {
+            "loss": 14.1608,
+            "grad_norm": 1.977250099182129,
+            "learning_rate": 0.0005,
+            "epoch": 0.030252270320842132,
+            "step": 675
+        },
+        {
+            "loss": 14.4085,
+            "grad_norm": 2.0198206901550293,
+            "learning_rate": 0.0005,
+            "epoch": 0.03047636121210763,
+            "step": 680
+        },
+        {
+            "loss": 14.2706,
+            "grad_norm": 2.0707778930664062,
+            "learning_rate": 0.0005,
+            "epoch": 0.03070045210337313,
+            "step": 685
+        },
+        {
+            "loss": 14.1478,
+            "grad_norm": 2.046593189239502,
+            "learning_rate": 0.0005,
+            "epoch": 0.030924542994638624,
+            "step": 690
+        },
+        {
+            "loss": 14.1779,
+            "grad_norm": 2.0935871601104736,
+            "learning_rate": 0.0005,
+            "epoch": 0.031148633885904123,
+            "step": 695
+        },
+        {
+            "loss": 14.2842,
+            "grad_norm": 2.1891965866088867,
+            "learning_rate": 0.0005,
+            "epoch": 0.03137272477716962,
+            "step": 700
+        },
+        {
+            "loss": 14.2074,
+            "grad_norm": 2.070681095123291,
+            "learning_rate": 0.0005,
+            "epoch": 0.03159681566843512,
+            "step": 705
+        },
+        {
+            "loss": 14.2629,
+            "grad_norm": 2.0730481147766113,
+            "learning_rate": 0.0005,
+            "epoch": 0.031820906559700615,
+            "step": 710
+        },
+        {
+            "loss": 14.252,
+            "grad_norm": 1.9831678867340088,
+            "learning_rate": 0.0005,
+            "epoch": 0.03204499745096611,
+            "step": 715
+        },
+        {
+            "loss": 14.3001,
+            "grad_norm": 1.9226957559585571,
+            "learning_rate": 0.0005,
+            "epoch": 0.03226908834223161,
+            "step": 720
+        },
+        {
+            "loss": 14.3861,
+            "grad_norm": 1.9222818613052368,
+            "learning_rate": 0.0005,
+            "epoch": 0.03249317923349711,
+            "step": 725
+        },
+        {
+            "loss": 14.1986,
+            "grad_norm": 2.1250648498535156,
+            "learning_rate": 0.0005,
+            "epoch": 0.0327172701247626,
+            "step": 730
+        },
+        {
+            "loss": 14.2355,
+            "grad_norm": 2.0338075160980225,
+            "learning_rate": 0.0005,
+            "epoch": 0.0329413610160281,
+            "step": 735
+        },
+        {
+            "loss": 14.2037,
+            "grad_norm": 1.9043233394622803,
+            "learning_rate": 0.0005,
+            "epoch": 0.0331654519072936,
+            "step": 740
+        },
+        {
+            "loss": 14.3108,
+            "grad_norm": 1.9609804153442383,
+            "learning_rate": 0.0005,
+            "epoch": 0.033389542798559095,
+            "step": 745
+        },
+        {
+            "loss": 14.3283,
+            "grad_norm": 1.9094796180725098,
+            "learning_rate": 0.0005,
+            "epoch": 0.03361363368982459,
+            "step": 750
+        },
+        {
+            "loss": 14.2745,
+            "grad_norm": 2.0199835300445557,
+            "learning_rate": 0.0005,
+            "epoch": 0.03383772458109009,
+            "step": 755
+        },
+        {
+            "loss": 14.248,
+            "grad_norm": 1.8590503931045532,
+            "learning_rate": 0.0005,
+            "epoch": 0.03406181547235559,
+            "step": 760
+        },
+        {
+            "loss": 14.2879,
+            "grad_norm": 1.8757070302963257,
+            "learning_rate": 0.0005,
+            "epoch": 0.03428590636362108,
+            "step": 765
+        },
+        {
+            "loss": 14.2926,
+            "grad_norm": 2.2693803310394287,
+            "learning_rate": 0.0005,
+            "epoch": 0.034509997254886585,
+            "step": 770
+        },
+        {
+            "loss": 14.4677,
+            "grad_norm": 2.126627206802368,
+            "learning_rate": 0.0005,
+            "epoch": 0.03473408814615208,
+            "step": 775
+        },
+        {
+            "loss": 14.3263,
+            "grad_norm": 2.0534279346466064,
+            "learning_rate": 0.0005,
+            "epoch": 0.034958179037417575,
+            "step": 780
+        },
+        {
+            "loss": 14.3332,
+            "grad_norm": 2.061572313308716,
+            "learning_rate": 0.0005,
+            "epoch": 0.03518226992868308,
+            "step": 785
+        },
+        {
+            "loss": 14.3802,
+            "grad_norm": 2.1430110931396484,
+            "learning_rate": 0.0005,
+            "epoch": 0.03540636081994857,
+            "step": 790
+        },
+        {
+            "loss": 14.1992,
+            "grad_norm": 2.114748477935791,
+            "learning_rate": 0.0005,
+            "epoch": 0.03563045171121407,
+            "step": 795
+        },
+        {
+            "loss": 14.239,
+            "grad_norm": 1.9971671104431152,
+            "learning_rate": 0.0005,
+            "epoch": 0.03585454260247956,
+            "step": 800
+        },
+        {
+            "loss": 14.2222,
+            "grad_norm": 1.95136559009552,
+            "learning_rate": 0.0005,
+            "epoch": 0.036078633493745065,
+            "step": 805
+        },
+        {
+            "loss": 14.1883,
+            "grad_norm": 2.1996357440948486,
+            "learning_rate": 0.0005,
+            "epoch": 0.03630272438501056,
+            "step": 810
+        },
+        {
+            "loss": 14.2557,
+            "grad_norm": 2.095583200454712,
+            "learning_rate": 0.0005,
+            "epoch": 0.036526815276276055,
+            "step": 815
+        },
+        {
+            "loss": 14.2175,
+            "grad_norm": 2.0955843925476074,
+            "learning_rate": 0.0005,
+            "epoch": 0.03675090616754156,
+            "step": 820
+        },
+        {
+            "loss": 14.2187,
+            "grad_norm": 1.9569734334945679,
+            "learning_rate": 0.0005,
+            "epoch": 0.03697499705880705,
+            "step": 825
+        },
+        {
+            "loss": 14.3528,
+            "grad_norm": 1.9772740602493286,
+            "learning_rate": 0.0005,
+            "epoch": 0.03719908795007255,
+            "step": 830
+        },
+        {
+            "loss": 14.2456,
+            "grad_norm": 2.1345417499542236,
+            "learning_rate": 0.0005,
+            "epoch": 0.03742317884133805,
+            "step": 835
+        },
+        {
+            "loss": 14.2124,
+            "grad_norm": 1.975181221961975,
+            "learning_rate": 0.0005,
+            "epoch": 0.037647269732603544,
+            "step": 840
+        },
+        {
+            "loss": 14.2323,
+            "grad_norm": 2.1009271144866943,
+            "learning_rate": 0.0005,
+            "epoch": 0.03787136062386904,
+            "step": 845
+        },
+        {
+            "loss": 14.251,
+            "grad_norm": 2.0016536712646484,
+            "learning_rate": 0.0005,
+            "epoch": 0.03809545151513454,
+            "step": 850
+        },
+        {
+            "loss": 14.2497,
+            "grad_norm": 1.9874367713928223,
+            "learning_rate": 0.0005,
+            "epoch": 0.03831954240640004,
+            "step": 855
+        },
+        {
+            "loss": 14.1931,
+            "grad_norm": 2.083117961883545,
+            "learning_rate": 0.0005,
+            "epoch": 0.03854363329766553,
+            "step": 860
+        },
+        {
+            "loss": 14.24,
+            "grad_norm": 2.0738794803619385,
+            "learning_rate": 0.0005,
+            "epoch": 0.03876772418893103,
+            "step": 865
+        },
+        {
+            "loss": 14.2042,
+            "grad_norm": 1.9692766666412354,
+            "learning_rate": 0.0005,
+            "epoch": 0.03899181508019653,
+            "step": 870
+        },
+        {
+            "loss": 14.3241,
+            "grad_norm": 2.271322727203369,
+            "learning_rate": 0.0005,
+            "epoch": 0.039215905971462024,
+            "step": 875
+        },
+        {
+            "loss": 14.2399,
+            "grad_norm": 2.0843918323516846,
+            "learning_rate": 0.0005,
+            "epoch": 0.03943999686272752,
+            "step": 880
+        },
+        {
+            "loss": 14.2721,
+            "grad_norm": 1.9291036128997803,
+            "learning_rate": 0.0005,
+            "epoch": 0.03966408775399302,
+            "step": 885
+        },
+        {
+            "loss": 14.2569,
+            "grad_norm": 2.0397393703460693,
+            "learning_rate": 0.0005,
+            "epoch": 0.03988817864525852,
+            "step": 890
+        },
+        {
+            "loss": 14.2288,
+            "grad_norm": 2.144932508468628,
+            "learning_rate": 0.0005,
+            "epoch": 0.04011226953652401,
+            "step": 895
+        },
+        {
+            "loss": 14.191,
+            "grad_norm": 1.9231230020523071,
+            "learning_rate": 0.0005,
+            "epoch": 0.040336360427789514,
+            "step": 900
+        },
+        {
+            "loss": 14.2027,
+            "grad_norm": 1.8474692106246948,
+            "learning_rate": 0.0005,
+            "epoch": 0.04056045131905501,
+            "step": 905
+        },
+        {
+            "loss": 14.2725,
+            "grad_norm": 2.227571725845337,
+            "learning_rate": 0.0005,
+            "epoch": 0.040784542210320504,
+            "step": 910
+        },
+        {
+            "loss": 14.291,
+            "grad_norm": 2.0123486518859863,
+            "learning_rate": 0.0005,
+            "epoch": 0.041008633101586006,
+            "step": 915
+        },
+        {
+            "loss": 14.3685,
+            "grad_norm": 2.12306809425354,
+            "learning_rate": 0.0005,
+            "epoch": 0.0412327239928515,
+            "step": 920
+        },
+        {
+            "loss": 14.3251,
+            "grad_norm": 1.951163411140442,
+            "learning_rate": 0.0005,
+            "epoch": 0.041456814884116996,
+            "step": 925
+        },
+        {
+            "loss": 14.3223,
+            "grad_norm": 2.004356861114502,
+            "learning_rate": 0.0005,
+            "epoch": 0.0416809057753825,
+            "step": 930
+        },
+        {
+            "loss": 14.2391,
+            "grad_norm": 2.049684524536133,
+            "learning_rate": 0.0005,
+            "epoch": 0.041904996666647994,
+            "step": 935
+        },
+        {
+            "loss": 14.3082,
+            "grad_norm": 2.043984889984131,
+            "learning_rate": 0.0005,
+            "epoch": 0.04212908755791349,
+            "step": 940
+        },
+        {
+            "loss": 14.2308,
+            "grad_norm": 2.0580008029937744,
+            "learning_rate": 0.0005,
+            "epoch": 0.042353178449178984,
+            "step": 945
+        },
+        {
+            "loss": 14.2479,
+            "grad_norm": 2.0241260528564453,
+            "learning_rate": 0.0005,
+            "epoch": 0.042577269340444486,
+            "step": 950
+        },
+        {
+            "loss": 14.2361,
+            "grad_norm": 2.0930376052856445,
+            "learning_rate": 0.0005,
+            "epoch": 0.04280136023170998,
+            "step": 955
+        },
+        {
+            "loss": 14.2378,
+            "grad_norm": 2.057032585144043,
+            "learning_rate": 0.0005,
+            "epoch": 0.043025451122975476,
+            "step": 960
+        },
+        {
+            "loss": 14.2969,
+            "grad_norm": 2.033273696899414,
+            "learning_rate": 0.0005,
+            "epoch": 0.04324954201424098,
+            "step": 965
+        },
+        {
+            "loss": 14.1837,
+            "grad_norm": 1.8709105253219604,
+            "learning_rate": 0.0005,
+            "epoch": 0.043473632905506474,
+            "step": 970
+        },
+        {
+            "loss": 14.2079,
+            "grad_norm": 2.157222270965576,
+            "learning_rate": 0.0005,
+            "epoch": 0.04369772379677197,
+            "step": 975
+        },
+        {
+            "loss": 14.4298,
+            "grad_norm": 2.202359437942505,
+            "learning_rate": 0.0005,
+            "epoch": 0.04392181468803747,
+            "step": 980
+        },
+        {
+            "loss": 14.1975,
+            "grad_norm": 2.245966672897339,
+            "learning_rate": 0.0005,
+            "epoch": 0.044145905579302966,
+            "step": 985
+        },
+        {
+            "loss": 14.2658,
+            "grad_norm": 2.1800920963287354,
+            "learning_rate": 0.0005,
+            "epoch": 0.04436999647056846,
+            "step": 990
+        },
+        {
+            "loss": 14.3301,
+            "grad_norm": 2.060558795928955,
+            "learning_rate": 0.0005,
+            "epoch": 0.04459408736183396,
+            "step": 995
+        },
+        {
+            "loss": 14.2039,
+            "grad_norm": 2.113269329071045,
+            "learning_rate": 0.0005,
+            "epoch": 0.04481817825309946,
+            "step": 1000
+        },
+        {
+            "eval_loss": 1.7727333307266235,
+            "eval_runtime": 18.7595,
+            "eval_samples_per_second": 873.37,
+            "eval_steps_per_second": 7.836,
+            "epoch": 0.04481817825309946,
+            "step": 1000
+        },
+        {
+            "loss": 14.2288,
+            "grad_norm": 2.1471664905548096,
+            "learning_rate": 0.0005,
+            "epoch": 0.04504226914436495,
+            "step": 1005
+        },
+        {
+            "loss": 14.3332,
+            "grad_norm": 2.2036454677581787,
+            "learning_rate": 0.0005,
+            "epoch": 0.04526636003563045,
+            "step": 1010
+        },
+        {
+            "loss": 14.1147,
+            "grad_norm": 2.000432252883911,
+            "learning_rate": 0.0005,
+            "epoch": 0.04549045092689595,
+            "step": 1015
+        },
+        {
+            "loss": 14.2888,
+            "grad_norm": 1.8297663927078247,
+            "learning_rate": 0.0005,
+            "epoch": 0.045714541818161446,
+            "step": 1020
+        },
+        {
+            "loss": 14.3685,
+            "grad_norm": 1.954213261604309,
+            "learning_rate": 0.0005,
+            "epoch": 0.04593863270942694,
+            "step": 1025
+        },
+        {
+            "loss": 14.2465,
+            "grad_norm": 2.0441508293151855,
+            "learning_rate": 0.0005,
+            "epoch": 0.04616272360069244,
+            "step": 1030
+        },
+        {
+            "loss": 14.1629,
+            "grad_norm": 1.9861886501312256,
+            "learning_rate": 0.0005,
+            "epoch": 0.04638681449195794,
+            "step": 1035
+        },
+        {
+            "loss": 14.3546,
+            "grad_norm": 1.9641690254211426,
+            "learning_rate": 0.0005,
+            "epoch": 0.04661090538322343,
+            "step": 1040
+        },
+        {
+            "loss": 14.3615,
+            "grad_norm": 1.914635419845581,
+            "learning_rate": 0.0005,
+            "epoch": 0.046834996274488935,
+            "step": 1045
+        },
+        {
+            "loss": 14.1544,
+            "grad_norm": 2.138408660888672,
+            "learning_rate": 0.0005,
+            "epoch": 0.04705908716575443,
+            "step": 1050
+        },
+        {
+            "loss": 14.1677,
+            "grad_norm": 2.0267415046691895,
+            "learning_rate": 0.0005,
+            "epoch": 0.047283178057019926,
+            "step": 1055
+        },
+        {
+            "loss": 14.2248,
+            "grad_norm": 2.190242290496826,
+            "learning_rate": 0.0005,
+            "epoch": 0.04750726894828543,
+            "step": 1060
+        },
+        {
+            "loss": 14.2402,
+            "grad_norm": 2.30466628074646,
+            "learning_rate": 0.0005,
+            "epoch": 0.04773135983955092,
+            "step": 1065
+        },
+        {
+            "loss": 14.1104,
+            "grad_norm": 2.035266876220703,
+            "learning_rate": 0.0005,
+            "epoch": 0.04795545073081642,
+            "step": 1070
+        },
+        {
+            "loss": 14.1833,
+            "grad_norm": 2.0810415744781494,
+            "learning_rate": 0.0005,
+            "epoch": 0.04817954162208191,
+            "step": 1075
+        },
+        {
+            "loss": 14.2342,
+            "grad_norm": 2.0724008083343506,
+            "learning_rate": 0.0005,
+            "epoch": 0.048403632513347415,
+            "step": 1080
+        },
+        {
+            "loss": 14.2719,
+            "grad_norm": 1.9949944019317627,
+            "learning_rate": 0.0005,
+            "epoch": 0.04862772340461291,
+            "step": 1085
+        },
+        {
+            "loss": 14.2835,
+            "grad_norm": 2.058635950088501,
+            "learning_rate": 0.0005,
+            "epoch": 0.048851814295878405,
+            "step": 1090
+        },
+        {
+            "loss": 14.1962,
+            "grad_norm": 1.8565870523452759,
+            "learning_rate": 0.0005,
+            "epoch": 0.04907590518714391,
+            "step": 1095
+        },
+        {
+            "loss": 14.2335,
+            "grad_norm": 1.9683058261871338,
+            "learning_rate": 0.0005,
+            "epoch": 0.0492999960784094,
+            "step": 1100
+        },
+        {
+            "loss": 14.2524,
+            "grad_norm": 1.9165276288986206,
+            "learning_rate": 0.0005,
+            "epoch": 0.0495240869696749,
+            "step": 1105
+        },
+        {
+            "loss": 14.1797,
+            "grad_norm": 1.9913203716278076,
+            "learning_rate": 0.0005,
+            "epoch": 0.0497481778609404,
+            "step": 1110
+        },
+        {
+            "loss": 14.1801,
+            "grad_norm": 2.0073564052581787,
+            "learning_rate": 0.0005,
+            "epoch": 0.049972268752205895,
+            "step": 1115
+        },
+        {
+            "loss": 14.2034,
+            "grad_norm": 2.0174570083618164,
+            "learning_rate": 0.0005,
+            "epoch": 0.05019635964347139,
+            "step": 1120
+        },
+        {
+            "loss": 14.2805,
+            "grad_norm": 2.1646080017089844,
+            "learning_rate": 0.0005,
+            "epoch": 0.05042045053473689,
+            "step": 1125
+        },
+        {
+            "loss": 14.3219,
+            "grad_norm": 1.9945608377456665,
+            "learning_rate": 0.0005,
+            "epoch": 0.05064454142600239,
+            "step": 1130
+        },
+        {
+            "loss": 14.3191,
+            "grad_norm": 2.1684176921844482,
+            "learning_rate": 0.0005,
+            "epoch": 0.05086863231726788,
+            "step": 1135
+        },
+        {
+            "loss": 14.1874,
+            "grad_norm": 2.009509325027466,
+            "learning_rate": 0.0005,
+            "epoch": 0.05109272320853338,
+            "step": 1140
+        },
+        {
+            "loss": 14.2761,
+            "grad_norm": 1.8036493062973022,
+            "learning_rate": 0.0005,
+            "epoch": 0.05131681409979888,
+            "step": 1145
+        },
+        {
+            "loss": 14.1475,
+            "grad_norm": 1.93241548538208,
+            "learning_rate": 0.0005,
+            "epoch": 0.051540904991064375,
+            "step": 1150
+        },
+        {
+            "loss": 14.1615,
+            "grad_norm": 1.9887784719467163,
+            "learning_rate": 0.0005,
+            "epoch": 0.05176499588232987,
+            "step": 1155
+        },
+        {
+            "loss": 14.1796,
+            "grad_norm": 2.110835552215576,
+            "learning_rate": 0.0005,
+            "epoch": 0.05198908677359537,
+            "step": 1160
+        },
+        {
+            "loss": 14.3305,
+            "grad_norm": 1.9870234727859497,
+            "learning_rate": 0.0005,
+            "epoch": 0.05221317766486087,
+            "step": 1165
+        },
+        {
+            "loss": 14.2401,
+            "grad_norm": 2.033839225769043,
+            "learning_rate": 0.0005,
+            "epoch": 0.05243726855612636,
+            "step": 1170
+        },
+        {
+            "loss": 14.2614,
+            "grad_norm": 2.0544512271881104,
+            "learning_rate": 0.0005,
+            "epoch": 0.052661359447391864,
+            "step": 1175
+        },
+        {
+            "loss": 14.2981,
+            "grad_norm": 2.0397698879241943,
+            "learning_rate": 0.0005,
+            "epoch": 0.05288545033865736,
+            "step": 1180
+        },
+        {
+            "loss": 14.2508,
+            "grad_norm": 2.0876481533050537,
+            "learning_rate": 0.0005,
+            "epoch": 0.053109541229922855,
+            "step": 1185
+        },
+        {
+            "loss": 14.2421,
+            "grad_norm": 2.0604329109191895,
+            "learning_rate": 0.0005,
+            "epoch": 0.05333363212118836,
+            "step": 1190
+        },
+        {
+            "loss": 14.2969,
+            "grad_norm": 2.262632131576538,
+            "learning_rate": 0.0005,
+            "epoch": 0.05355772301245385,
+            "step": 1195
+        },
+        {
+            "loss": 14.2506,
+            "grad_norm": 2.0581042766571045,
+            "learning_rate": 0.0005,
+            "epoch": 0.05378181390371935,
+            "step": 1200
+        },
+        {
+            "loss": 14.2416,
+            "grad_norm": 1.9745397567749023,
+            "learning_rate": 0.0005,
+            "epoch": 0.05400590479498485,
+            "step": 1205
+        },
+        {
+            "loss": 14.1699,
+            "grad_norm": 1.9858930110931396,
+            "learning_rate": 0.0005,
+            "epoch": 0.054229995686250344,
+            "step": 1210
+        },
+        {
+            "loss": 14.3006,
+            "grad_norm": 2.047473907470703,
+            "learning_rate": 0.0005,
+            "epoch": 0.05445408657751584,
+            "step": 1215
+        },
+        {
+            "loss": 14.3145,
+            "grad_norm": 2.091883420944214,
+            "learning_rate": 0.0005,
+            "epoch": 0.054678177468781335,
+            "step": 1220
+        },
+        {
+            "loss": 14.1586,
+            "grad_norm": 1.9428058862686157,
+            "learning_rate": 0.0005,
+            "epoch": 0.05490226836004684,
+            "step": 1225
+        },
+        {
+            "loss": 14.3759,
+            "grad_norm": 1.914755940437317,
+            "learning_rate": 0.0005,
+            "epoch": 0.05512635925131233,
+            "step": 1230
+        },
+        {
+            "loss": 14.3249,
+            "grad_norm": 1.89765465259552,
+            "learning_rate": 0.0005,
+            "epoch": 0.05535045014257783,
+            "step": 1235
+        },
+        {
+            "loss": 14.2483,
+            "grad_norm": 2.0791454315185547,
+            "learning_rate": 0.0005,
+            "epoch": 0.05557454103384333,
+            "step": 1240
+        },
+        {
+            "loss": 14.2269,
+            "grad_norm": 2.1523244380950928,
+            "learning_rate": 0.0005,
+            "epoch": 0.055798631925108824,
+            "step": 1245
+        },
+        {
+            "loss": 14.2138,
+            "grad_norm": 2.0883595943450928,
+            "learning_rate": 0.0005,
+            "epoch": 0.05602272281637432,
+            "step": 1250
+        },
+        {
+            "loss": 14.1858,
+            "grad_norm": 2.0705084800720215,
+            "learning_rate": 0.0005,
+            "epoch": 0.05624681370763982,
+            "step": 1255
+        },
+        {
+            "loss": 14.1589,
+            "grad_norm": 1.935402274131775,
+            "learning_rate": 0.0005,
+            "epoch": 0.05647090459890532,
+            "step": 1260
+        },
+        {
+            "loss": 14.1945,
+            "grad_norm": 2.0258562564849854,
+            "learning_rate": 0.0005,
+            "epoch": 0.05669499549017081,
+            "step": 1265
+        },
+        {
+            "loss": 14.2134,
+            "grad_norm": 2.161383628845215,
+            "learning_rate": 0.0005,
+            "epoch": 0.056919086381436314,
+            "step": 1270
+        },
+        {
+            "loss": 14.1014,
+            "grad_norm": 1.868166446685791,
+            "learning_rate": 0.0005,
+            "epoch": 0.05714317727270181,
+            "step": 1275
+        },
+        {
+            "loss": 14.1909,
+            "grad_norm": 2.053658962249756,
+            "learning_rate": 0.0005,
+            "epoch": 0.057367268163967304,
+            "step": 1280
+        },
+        {
+            "loss": 14.1878,
+            "grad_norm": 1.993074655532837,
+            "learning_rate": 0.0005,
+            "epoch": 0.0575913590552328,
+            "step": 1285
+        },
+        {
+            "loss": 14.1786,
+            "grad_norm": 1.9569101333618164,
+            "learning_rate": 0.0005,
+            "epoch": 0.0578154499464983,
+            "step": 1290
+        },
+        {
+            "loss": 14.1818,
+            "grad_norm": 1.9939252138137817,
+            "learning_rate": 0.0005,
+            "epoch": 0.058039540837763796,
+            "step": 1295
+        },
+        {
+            "loss": 14.2783,
+            "grad_norm": 1.876535177230835,
+            "learning_rate": 0.0005,
+            "epoch": 0.05826363172902929,
+            "step": 1300
+        },
+        {
+            "loss": 14.2104,
+            "grad_norm": 2.0808701515197754,
+            "learning_rate": 0.0005,
+            "epoch": 0.058487722620294794,
+            "step": 1305
+        },
+        {
+            "loss": 14.3005,
+            "grad_norm": 2.094132900238037,
+            "learning_rate": 0.0005,
+            "epoch": 0.05871181351156029,
+            "step": 1310
+        },
+        {
+            "loss": 14.2414,
+            "grad_norm": 2.143573522567749,
+            "learning_rate": 0.0005,
+            "epoch": 0.058935904402825784,
+            "step": 1315
+        },
+        {
+            "loss": 14.1906,
+            "grad_norm": 1.943663477897644,
+            "learning_rate": 0.0005,
+            "epoch": 0.059159995294091286,
+            "step": 1320
+        },
+        {
+            "loss": 14.233,
+            "grad_norm": 1.9655219316482544,
+            "learning_rate": 0.0005,
+            "epoch": 0.05938408618535678,
+            "step": 1325
+        },
+        {
+            "loss": 14.1973,
+            "grad_norm": 1.943224549293518,
+            "learning_rate": 0.0005,
+            "epoch": 0.059608177076622276,
+            "step": 1330
+        },
+        {
+            "loss": 14.2429,
+            "grad_norm": 2.0061824321746826,
+            "learning_rate": 0.0005,
+            "epoch": 0.05983226796788778,
+            "step": 1335
+        },
+        {
+            "loss": 14.2179,
+            "grad_norm": 1.8484687805175781,
+            "learning_rate": 0.0005,
+            "epoch": 0.060056358859153273,
+            "step": 1340
+        },
+        {
+            "loss": 14.2897,
+            "grad_norm": 2.1448802947998047,
+            "learning_rate": 0.0005,
+            "epoch": 0.06028044975041877,
+            "step": 1345
+        },
+        {
+            "loss": 14.3255,
+            "grad_norm": 2.0925004482269287,
+            "learning_rate": 0.0005,
+            "epoch": 0.060504540641684264,
+            "step": 1350
+        },
+        {
+            "loss": 14.2182,
+            "grad_norm": 2.072070598602295,
+            "learning_rate": 0.0005,
+            "epoch": 0.060728631532949766,
+            "step": 1355
+        },
+        {
+            "loss": 14.2734,
+            "grad_norm": 2.3992390632629395,
+            "learning_rate": 0.0005,
+            "epoch": 0.06095272242421526,
+            "step": 1360
+        },
+        {
+            "loss": 14.2596,
+            "grad_norm": 2.0187509059906006,
+            "learning_rate": 0.0005,
+            "epoch": 0.061176813315480756,
+            "step": 1365
+        },
+        {
+            "loss": 14.2198,
+            "grad_norm": 2.126812219619751,
+            "learning_rate": 0.0005,
+            "epoch": 0.06140090420674626,
+            "step": 1370
+        },
+        {
+            "loss": 14.3122,
+            "grad_norm": 1.9537678956985474,
+            "learning_rate": 0.0005,
+            "epoch": 0.06162499509801175,
+            "step": 1375
+        },
+        {
+            "loss": 14.1976,
+            "grad_norm": 1.9577600955963135,
+            "learning_rate": 0.0005,
+            "epoch": 0.06184908598927725,
+            "step": 1380
+        },
+        {
+            "loss": 14.1317,
+            "grad_norm": 1.8684802055358887,
+            "learning_rate": 0.0005,
+            "epoch": 0.06207317688054275,
+            "step": 1385
+        },
+        {
+            "loss": 14.2812,
+            "grad_norm": 2.074007511138916,
+            "learning_rate": 0.0005,
+            "epoch": 0.062297267771808246,
+            "step": 1390
+        },
+        {
+            "loss": 14.2356,
+            "grad_norm": 2.1334567070007324,
+            "learning_rate": 0.0005,
+            "epoch": 0.06252135866307375,
+            "step": 1395
+        },
+        {
+            "loss": 14.1932,
+            "grad_norm": 2.039788007736206,
+            "learning_rate": 0.0005,
+            "epoch": 0.06274544955433924,
+            "step": 1400
+        },
+        {
+            "loss": 14.2866,
+            "grad_norm": 1.9726324081420898,
+            "learning_rate": 0.0005,
+            "epoch": 0.06296954044560474,
+            "step": 1405
+        },
+        {
+            "loss": 14.2492,
+            "grad_norm": 2.065645217895508,
+            "learning_rate": 0.0005,
+            "epoch": 0.06319363133687024,
+            "step": 1410
+        },
+        {
+            "loss": 14.3002,
+            "grad_norm": 2.189345598220825,
+            "learning_rate": 0.0005,
+            "epoch": 0.06341772222813573,
+            "step": 1415
+        },
+        {
+            "loss": 14.1713,
+            "grad_norm": 1.8538919687271118,
+            "learning_rate": 0.0005,
+            "epoch": 0.06364181311940123,
+            "step": 1420
+        },
+        {
+            "loss": 14.2846,
+            "grad_norm": 2.042672634124756,
+            "learning_rate": 0.0005,
+            "epoch": 0.06386590401066673,
+            "step": 1425
+        },
+        {
+            "loss": 14.2022,
+            "grad_norm": 2.0098628997802734,
+            "learning_rate": 0.0005,
+            "epoch": 0.06408999490193222,
+            "step": 1430
+        },
+        {
+            "loss": 14.2969,
+            "grad_norm": 1.9617176055908203,
+            "learning_rate": 0.0005,
+            "epoch": 0.06431408579319772,
+            "step": 1435
+        },
+        {
+            "loss": 14.282,
+            "grad_norm": 2.022548198699951,
+            "learning_rate": 0.0005,
+            "epoch": 0.06453817668446322,
+            "step": 1440
+        },
+        {
+            "loss": 14.157,
+            "grad_norm": 1.9941786527633667,
+            "learning_rate": 0.0005,
+            "epoch": 0.06476226757572871,
+            "step": 1445
+        },
+        {
+            "loss": 14.297,
+            "grad_norm": 2.028557777404785,
+            "learning_rate": 0.0005,
+            "epoch": 0.06498635846699422,
+            "step": 1450
+        },
+        {
+            "loss": 14.1691,
+            "grad_norm": 1.9549771547317505,
+            "learning_rate": 0.0005,
+            "epoch": 0.0652104493582597,
+            "step": 1455
+        },
+        {
+            "loss": 14.3092,
+            "grad_norm": 2.1454029083251953,
+            "learning_rate": 0.0005,
+            "epoch": 0.0654345402495252,
+            "step": 1460
+        },
+        {
+            "loss": 14.2483,
+            "grad_norm": 2.0175013542175293,
+            "learning_rate": 0.0005,
+            "epoch": 0.06565863114079071,
+            "step": 1465
+        },
+        {
+            "loss": 14.3039,
+            "grad_norm": 1.8648537397384644,
+            "learning_rate": 0.0005,
+            "epoch": 0.0658827220320562,
+            "step": 1470
+        },
+        {
+            "loss": 14.2233,
+            "grad_norm": 2.2553863525390625,
+            "learning_rate": 0.0005,
+            "epoch": 0.0661068129233217,
+            "step": 1475
+        },
+        {
+            "loss": 14.2414,
+            "grad_norm": 2.15773606300354,
+            "learning_rate": 0.0005,
+            "epoch": 0.0663309038145872,
+            "step": 1480
+        },
+        {
+            "loss": 14.1754,
+            "grad_norm": 1.9377132654190063,
+            "learning_rate": 0.0005,
+            "epoch": 0.06655499470585269,
+            "step": 1485
+        },
+        {
+            "loss": 14.2484,
+            "grad_norm": 2.170342445373535,
+            "learning_rate": 0.0005,
+            "epoch": 0.06677908559711819,
+            "step": 1490
+        },
+        {
+            "loss": 14.2683,
+            "grad_norm": 1.8721458911895752,
+            "learning_rate": 0.0005,
+            "epoch": 0.06700317648838369,
+            "step": 1495
+        },
+        {
+            "loss": 14.1945,
+            "grad_norm": 1.9463225603103638,
+            "learning_rate": 0.0005,
+            "epoch": 0.06722726737964918,
+            "step": 1500
+        },
+        {
+            "eval_loss": 1.7756863832473755,
+            "eval_runtime": 18.6591,
+            "eval_samples_per_second": 878.068,
+            "eval_steps_per_second": 7.878,
+            "epoch": 0.06722726737964918,
+            "step": 1500
+        },
+        {
+            "loss": 14.1917,
+            "grad_norm": 1.997578501701355,
+            "learning_rate": 0.0005,
+            "epoch": 0.06745135827091468,
+            "step": 1505
+        },
+        {
+            "loss": 14.3386,
+            "grad_norm": 2.0206363201141357,
+            "learning_rate": 0.0005,
+            "epoch": 0.06767544916218018,
+            "step": 1510
+        },
+        {
+            "loss": 14.2589,
+            "grad_norm": 2.2400431632995605,
+            "learning_rate": 0.0005,
+            "epoch": 0.06789954005344567,
+            "step": 1515
+        },
+        {
+            "loss": 14.2423,
+            "grad_norm": 2.181386709213257,
+            "learning_rate": 0.0005,
+            "epoch": 0.06812363094471117,
+            "step": 1520
+        },
+        {
+            "loss": 14.1931,
+            "grad_norm": 1.9889214038848877,
+            "learning_rate": 0.0005,
+            "epoch": 0.06834772183597668,
+            "step": 1525
+        },
+        {
+            "loss": 14.3355,
+            "grad_norm": 2.175074577331543,
+            "learning_rate": 0.0005,
+            "epoch": 0.06857181272724217,
+            "step": 1530
+        },
+        {
+            "loss": 14.1702,
+            "grad_norm": 2.0602669715881348,
+            "learning_rate": 0.0005,
+            "epoch": 0.06879590361850767,
+            "step": 1535
+        },
+        {
+            "loss": 14.2198,
+            "grad_norm": 2.050475597381592,
+            "learning_rate": 0.0005,
+            "epoch": 0.06901999450977317,
+            "step": 1540
+        },
+        {
+            "loss": 14.2265,
+            "grad_norm": 2.1217849254608154,
+            "learning_rate": 0.0005,
+            "epoch": 0.06924408540103866,
+            "step": 1545
+        },
+        {
+            "loss": 14.212,
+            "grad_norm": 2.244414806365967,
+            "learning_rate": 0.0005,
+            "epoch": 0.06946817629230416,
+            "step": 1550
+        },
+        {
+            "loss": 14.2356,
+            "grad_norm": 2.1231167316436768,
+            "learning_rate": 0.0005,
+            "epoch": 0.06969226718356966,
+            "step": 1555
+        },
+        {
+            "loss": 14.2127,
+            "grad_norm": 1.9955865144729614,
+            "learning_rate": 0.0005,
+            "epoch": 0.06991635807483515,
+            "step": 1560
+        },
+        {
+            "loss": 14.2554,
+            "grad_norm": 2.26926851272583,
+            "learning_rate": 0.0005,
+            "epoch": 0.07014044896610065,
+            "step": 1565
+        },
+        {
+            "loss": 14.087,
+            "grad_norm": 1.944664478302002,
+            "learning_rate": 0.0005,
+            "epoch": 0.07036453985736615,
+            "step": 1570
+        },
+        {
+            "loss": 14.1392,
+            "grad_norm": 2.1763548851013184,
+            "learning_rate": 0.0005,
+            "epoch": 0.07058863074863164,
+            "step": 1575
+        },
+        {
+            "loss": 14.1809,
+            "grad_norm": 2.0459864139556885,
+            "learning_rate": 0.0005,
+            "epoch": 0.07081272163989714,
+            "step": 1580
+        },
+        {
+            "loss": 14.1929,
+            "grad_norm": 1.9332237243652344,
+            "learning_rate": 0.0005,
+            "epoch": 0.07103681253116265,
+            "step": 1585
+        },
+        {
+            "loss": 14.1885,
+            "grad_norm": 1.9467579126358032,
+            "learning_rate": 0.0005,
+            "epoch": 0.07126090342242813,
+            "step": 1590
+        },
+        {
+            "loss": 14.1799,
+            "grad_norm": 1.8630837202072144,
+            "learning_rate": 0.0005,
+            "epoch": 0.07148499431369364,
+            "step": 1595
+        },
+        {
+            "loss": 14.1846,
+            "grad_norm": 2.0461738109588623,
+            "learning_rate": 0.0005,
+            "epoch": 0.07170908520495912,
+            "step": 1600
+        },
+        {
+            "loss": 14.3046,
+            "grad_norm": 1.9873324632644653,
+            "learning_rate": 0.0005,
+            "epoch": 0.07193317609622463,
+            "step": 1605
+        },
+        {
+            "loss": 14.1908,
+            "grad_norm": 2.094851493835449,
+            "learning_rate": 0.0005,
+            "epoch": 0.07215726698749013,
+            "step": 1610
+        },
+        {
+            "loss": 14.269,
+            "grad_norm": 2.1090633869171143,
+            "learning_rate": 0.0005,
+            "epoch": 0.07238135787875562,
+            "step": 1615
+        },
+        {
+            "loss": 14.2144,
+            "grad_norm": 2.027017831802368,
+            "learning_rate": 0.0005,
+            "epoch": 0.07260544877002112,
+            "step": 1620
+        },
+        {
+            "loss": 14.287,
+            "grad_norm": 2.1863245964050293,
+            "learning_rate": 0.0005,
+            "epoch": 0.07282953966128662,
+            "step": 1625
+        },
+        {
+            "loss": 14.264,
+            "grad_norm": 2.1034693717956543,
+            "learning_rate": 0.0005,
+            "epoch": 0.07305363055255211,
+            "step": 1630
+        },
+        {
+            "loss": 14.1631,
+            "grad_norm": 1.9393339157104492,
+            "learning_rate": 0.0005,
+            "epoch": 0.07327772144381761,
+            "step": 1635
+        },
+        {
+            "loss": 14.2625,
+            "grad_norm": 1.9777086973190308,
+            "learning_rate": 0.0005,
+            "epoch": 0.07350181233508311,
+            "step": 1640
+        },
+        {
+            "loss": 14.201,
+            "grad_norm": 2.212585687637329,
+            "learning_rate": 0.0005,
+            "epoch": 0.0737259032263486,
+            "step": 1645
+        },
+        {
+            "loss": 14.1729,
+            "grad_norm": 2.052217483520508,
+            "learning_rate": 0.0005,
+            "epoch": 0.0739499941176141,
+            "step": 1650
+        },
+        {
+            "loss": 14.1748,
+            "grad_norm": 2.1643266677856445,
+            "learning_rate": 0.0005,
+            "epoch": 0.0741740850088796,
+            "step": 1655
+        },
+        {
+            "loss": 14.2851,
+            "grad_norm": 2.2242891788482666,
+            "learning_rate": 0.0005,
+            "epoch": 0.0743981759001451,
+            "step": 1660
+        },
+        {
+            "loss": 14.2938,
+            "grad_norm": 2.176030158996582,
+            "learning_rate": 0.0005,
+            "epoch": 0.0746222667914106,
+            "step": 1665
+        },
+        {
+            "loss": 14.4237,
+            "grad_norm": 2.0001564025878906,
+            "learning_rate": 0.0005,
+            "epoch": 0.0748463576826761,
+            "step": 1670
+        },
+        {
+            "loss": 14.2654,
+            "grad_norm": 2.1057164669036865,
+            "learning_rate": 0.0005,
+            "epoch": 0.07507044857394159,
+            "step": 1675
+        },
+        {
+            "loss": 14.2158,
+            "grad_norm": 2.168222427368164,
+            "learning_rate": 0.0005,
+            "epoch": 0.07529453946520709,
+            "step": 1680
+        },
+        {
+            "loss": 14.3471,
+            "grad_norm": 2.0431439876556396,
+            "learning_rate": 0.0005,
+            "epoch": 0.07551863035647259,
+            "step": 1685
+        },
+        {
+            "loss": 14.2348,
+            "grad_norm": 2.1417438983917236,
+            "learning_rate": 0.0005,
+            "epoch": 0.07574272124773808,
+            "step": 1690
+        },
+        {
+            "loss": 14.3011,
+            "grad_norm": 1.8830145597457886,
+            "learning_rate": 0.0005,
+            "epoch": 0.07596681213900358,
+            "step": 1695
+        },
+        {
+            "loss": 14.1981,
+            "grad_norm": 2.20562744140625,
+            "learning_rate": 0.0005,
+            "epoch": 0.07619090303026908,
+            "step": 1700
+        },
+        {
+            "loss": 14.1587,
+            "grad_norm": 2.0206198692321777,
+            "learning_rate": 0.0005,
+            "epoch": 0.07641499392153457,
+            "step": 1705
+        },
+        {
+            "loss": 14.2649,
+            "grad_norm": 1.962499976158142,
+            "learning_rate": 0.0005,
+            "epoch": 0.07663908481280007,
+            "step": 1710
+        },
+        {
+            "loss": 14.2564,
+            "grad_norm": 2.036729574203491,
+            "learning_rate": 0.0005,
+            "epoch": 0.07686317570406558,
+            "step": 1715
+        },
+        {
+            "loss": 14.1581,
+            "grad_norm": 2.094439744949341,
+            "learning_rate": 0.0005,
+            "epoch": 0.07708726659533106,
+            "step": 1720
+        },
+        {
+            "loss": 14.1704,
+            "grad_norm": 2.143383502960205,
+            "learning_rate": 0.0005,
+            "epoch": 0.07731135748659657,
+            "step": 1725
+        },
+        {
+            "loss": 14.2263,
+            "grad_norm": 1.9057375192642212,
+            "learning_rate": 0.0005,
+            "epoch": 0.07753544837786205,
+            "step": 1730
+        },
+        {
+            "loss": 14.3187,
+            "grad_norm": 1.9055688381195068,
+            "learning_rate": 0.0005,
+            "epoch": 0.07775953926912756,
+            "step": 1735
+        },
+        {
+            "loss": 14.3005,
+            "grad_norm": 2.220723867416382,
+            "learning_rate": 0.0005,
+            "epoch": 0.07798363016039306,
+            "step": 1740
+        },
+        {
+            "loss": 14.2552,
+            "grad_norm": 2.2899348735809326,
+            "learning_rate": 0.0005,
+            "epoch": 0.07820772105165855,
+            "step": 1745
+        },
+        {
+            "loss": 14.1751,
+            "grad_norm": 1.9819159507751465,
+            "learning_rate": 0.0005,
+            "epoch": 0.07843181194292405,
+            "step": 1750
+        },
+        {
+            "loss": 14.1918,
+            "grad_norm": 1.9067010879516602,
+            "learning_rate": 0.0005,
+            "epoch": 0.07865590283418955,
+            "step": 1755
+        },
+        {
+            "loss": 14.2655,
+            "grad_norm": 1.8417835235595703,
+            "learning_rate": 0.0005,
+            "epoch": 0.07887999372545504,
+            "step": 1760
+        },
+        {
+            "loss": 14.2191,
+            "grad_norm": 2.03226637840271,
+            "learning_rate": 0.0005,
+            "epoch": 0.07910408461672054,
+            "step": 1765
+        },
+        {
+            "loss": 14.2526,
+            "grad_norm": 1.9686570167541504,
+            "learning_rate": 0.0005,
+            "epoch": 0.07932817550798604,
+            "step": 1770
+        },
+        {
+            "loss": 14.2191,
+            "grad_norm": 1.992022156715393,
+            "learning_rate": 0.0005,
+            "epoch": 0.07955226639925153,
+            "step": 1775
+        },
+        {
+            "loss": 14.222,
+            "grad_norm": 1.9003782272338867,
+            "learning_rate": 0.0005,
+            "epoch": 0.07977635729051703,
+            "step": 1780
+        },
+        {
+            "loss": 14.2151,
+            "grad_norm": 2.0168862342834473,
+            "learning_rate": 0.0005,
+            "epoch": 0.08000044818178254,
+            "step": 1785
+        },
+        {
+            "loss": 14.2644,
+            "grad_norm": 2.1465072631835938,
+            "learning_rate": 0.0005,
+            "epoch": 0.08022453907304802,
+            "step": 1790
+        },
+        {
+            "loss": 14.2751,
+            "grad_norm": 1.987831473350525,
+            "learning_rate": 0.0005,
+            "epoch": 0.08044862996431353,
+            "step": 1795
+        },
+        {
+            "loss": 14.3681,
+            "grad_norm": 2.0240437984466553,
+            "learning_rate": 0.0005,
+            "epoch": 0.08067272085557903,
+            "step": 1800
+        },
+        {
+            "loss": 14.3111,
+            "grad_norm": 2.202488422393799,
+            "learning_rate": 0.0005,
+            "epoch": 0.08089681174684452,
+            "step": 1805
+        },
+        {
+            "loss": 14.1726,
+            "grad_norm": 2.3818199634552,
+            "learning_rate": 0.0005,
+            "epoch": 0.08112090263811002,
+            "step": 1810
+        },
+        {
+            "loss": 14.2812,
+            "grad_norm": 2.1590795516967773,
+            "learning_rate": 0.0005,
+            "epoch": 0.08134499352937552,
+            "step": 1815
+        },
+        {
+            "loss": 14.1505,
+            "grad_norm": 2.025921106338501,
+            "learning_rate": 0.0005,
+            "epoch": 0.08156908442064101,
+            "step": 1820
+        },
+        {
+            "loss": 14.2326,
+            "grad_norm": 1.9268773794174194,
+            "learning_rate": 0.0005,
+            "epoch": 0.08179317531190651,
+            "step": 1825
+        },
+        {
+            "loss": 14.2276,
+            "grad_norm": 1.9694938659667969,
+            "learning_rate": 0.0005,
+            "epoch": 0.08201726620317201,
+            "step": 1830
+        },
+        {
+            "loss": 14.244,
+            "grad_norm": 2.059058904647827,
+            "learning_rate": 0.0005,
+            "epoch": 0.0822413570944375,
+            "step": 1835
+        },
+        {
+            "loss": 14.264,
+            "grad_norm": 1.882689118385315,
+            "learning_rate": 0.0005,
+            "epoch": 0.082465447985703,
+            "step": 1840
+        },
+        {
+            "loss": 14.2633,
+            "grad_norm": 2.167781114578247,
+            "learning_rate": 0.0005,
+            "epoch": 0.0826895388769685,
+            "step": 1845
+        },
+        {
+            "loss": 14.2088,
+            "grad_norm": 2.1264965534210205,
+            "learning_rate": 0.0005,
+            "epoch": 0.08291362976823399,
+            "step": 1850
+        },
+        {
+            "loss": 14.1594,
+            "grad_norm": 2.0307443141937256,
+            "learning_rate": 0.0005,
+            "epoch": 0.0831377206594995,
+            "step": 1855
+        },
+        {
+            "loss": 14.3297,
+            "grad_norm": 2.053874969482422,
+            "learning_rate": 0.0005,
+            "epoch": 0.083361811550765,
+            "step": 1860
+        },
+        {
+            "loss": 14.1621,
+            "grad_norm": 1.9825525283813477,
+            "learning_rate": 0.0005,
+            "epoch": 0.08358590244203049,
+            "step": 1865
+        },
+        {
+            "loss": 14.18,
+            "grad_norm": 1.9070043563842773,
+            "learning_rate": 0.0005,
+            "epoch": 0.08380999333329599,
+            "step": 1870
+        },
+        {
+            "loss": 14.2022,
+            "grad_norm": 1.9118187427520752,
+            "learning_rate": 0.0005,
+            "epoch": 0.08403408422456148,
+            "step": 1875
+        },
+        {
+            "loss": 14.3087,
+            "grad_norm": 2.05552339553833,
+            "learning_rate": 0.0005,
+            "epoch": 0.08425817511582698,
+            "step": 1880
+        },
+        {
+            "loss": 14.1496,
+            "grad_norm": 2.0132551193237305,
+            "learning_rate": 0.0005,
+            "epoch": 0.08448226600709248,
+            "step": 1885
+        },
+        {
+            "loss": 14.1777,
+            "grad_norm": 1.9496753215789795,
+            "learning_rate": 0.0005,
+            "epoch": 0.08470635689835797,
+            "step": 1890
+        },
+        {
+            "loss": 14.2595,
+            "grad_norm": 1.9793522357940674,
+            "learning_rate": 0.0005,
+            "epoch": 0.08493044778962347,
+            "step": 1895
+        },
+        {
+            "loss": 14.1772,
+            "grad_norm": 2.1838932037353516,
+            "learning_rate": 0.0005,
+            "epoch": 0.08515453868088897,
+            "step": 1900
+        },
+        {
+            "loss": 14.2392,
+            "grad_norm": 2.0018186569213867,
+            "learning_rate": 0.0005,
+            "epoch": 0.08537862957215446,
+            "step": 1905
+        },
+        {
+            "loss": 14.2405,
+            "grad_norm": 1.9946718215942383,
+            "learning_rate": 0.0005,
+            "epoch": 0.08560272046341996,
+            "step": 1910
+        },
+        {
+            "loss": 14.1868,
+            "grad_norm": 1.9519858360290527,
+            "learning_rate": 0.0005,
+            "epoch": 0.08582681135468546,
+            "step": 1915
+        },
+        {
+            "loss": 14.2879,
+            "grad_norm": 2.1353161334991455,
+            "learning_rate": 0.0005,
+            "epoch": 0.08605090224595095,
+            "step": 1920
+        },
+        {
+            "loss": 14.2344,
+            "grad_norm": 1.9332703351974487,
+            "learning_rate": 0.0005,
+            "epoch": 0.08627499313721645,
+            "step": 1925
+        },
+        {
+            "loss": 14.1706,
+            "grad_norm": 1.8814945220947266,
+            "learning_rate": 0.0005,
+            "epoch": 0.08649908402848196,
+            "step": 1930
+        },
+        {
+            "loss": 14.3248,
+            "grad_norm": 2.1232025623321533,
+            "learning_rate": 0.0005,
+            "epoch": 0.08672317491974744,
+            "step": 1935
+        },
+        {
+            "loss": 14.2242,
+            "grad_norm": 2.008812665939331,
+            "learning_rate": 0.0005,
+            "epoch": 0.08694726581101295,
+            "step": 1940
+        },
+        {
+            "loss": 14.3041,
+            "grad_norm": 2.014477014541626,
+            "learning_rate": 0.0005,
+            "epoch": 0.08717135670227845,
+            "step": 1945
+        },
+        {
+            "loss": 14.2186,
+            "grad_norm": 1.984898328781128,
+            "learning_rate": 0.0005,
+            "epoch": 0.08739544759354394,
+            "step": 1950
+        },
+        {
+            "loss": 14.1764,
+            "grad_norm": 1.9279865026474,
+            "learning_rate": 0.0005,
+            "epoch": 0.08761953848480944,
+            "step": 1955
+        },
+        {
+            "loss": 14.2195,
+            "grad_norm": 1.8694874048233032,
+            "learning_rate": 0.0005,
+            "epoch": 0.08784362937607494,
+            "step": 1960
+        },
+        {
+            "loss": 14.2286,
+            "grad_norm": 1.8479489088058472,
+            "learning_rate": 0.0005,
+            "epoch": 0.08806772026734043,
+            "step": 1965
+        },
+        {
+            "loss": 14.3218,
+            "grad_norm": 1.998494267463684,
+            "learning_rate": 0.0005,
+            "epoch": 0.08829181115860593,
+            "step": 1970
+        },
+        {
+            "loss": 14.1551,
+            "grad_norm": 1.9892079830169678,
+            "learning_rate": 0.0005,
+            "epoch": 0.08851590204987143,
+            "step": 1975
+        },
+        {
+            "loss": 14.2086,
+            "grad_norm": 1.9579588174819946,
+            "learning_rate": 0.0005,
+            "epoch": 0.08873999294113692,
+            "step": 1980
+        },
+        {
+            "loss": 14.2593,
+            "grad_norm": 2.2442996501922607,
+            "learning_rate": 0.0005,
+            "epoch": 0.08896408383240242,
+            "step": 1985
+        },
+        {
+            "loss": 14.2019,
+            "grad_norm": 2.1113250255584717,
+            "learning_rate": 0.0005,
+            "epoch": 0.08918817472366793,
+            "step": 1990
+        },
+        {
+            "loss": 14.3084,
+            "grad_norm": 1.873713731765747,
+            "learning_rate": 0.0005,
+            "epoch": 0.08941226561493341,
+            "step": 1995
+        },
+        {
+            "loss": 14.2092,
+            "grad_norm": 1.936440110206604,
+            "learning_rate": 0.0005,
+            "epoch": 0.08963635650619892,
+            "step": 2000
+        },
+        {
+            "eval_loss": 1.7694858312606812,
+            "eval_runtime": 18.5104,
+            "eval_samples_per_second": 885.125,
+            "eval_steps_per_second": 7.941,
+            "epoch": 0.08963635650619892,
+            "step": 2000
+        },
+        {
+            "loss": 14.2115,
+            "grad_norm": 2.13944673538208,
+            "learning_rate": 0.0005,
+            "epoch": 0.0898604473974644,
+            "step": 2005
+        },
+        {
+            "loss": 14.2865,
+            "grad_norm": 2.0662755966186523,
+            "learning_rate": 0.0005,
+            "epoch": 0.0900845382887299,
+            "step": 2010
+        },
+        {
+            "loss": 14.2349,
+            "grad_norm": 2.0008041858673096,
+            "learning_rate": 0.0005,
+            "epoch": 0.09030862917999541,
+            "step": 2015
+        },
+        {
+            "loss": 14.315,
+            "grad_norm": 2.110163688659668,
+            "learning_rate": 0.0005,
+            "epoch": 0.0905327200712609,
+            "step": 2020
+        },
+        {
+            "loss": 14.249,
+            "grad_norm": 2.0696797370910645,
+            "learning_rate": 0.0005,
+            "epoch": 0.0907568109625264,
+            "step": 2025
+        },
+        {
+            "loss": 14.2679,
+            "grad_norm": 2.0506837368011475,
+            "learning_rate": 0.0005,
+            "epoch": 0.0909809018537919,
+            "step": 2030
+        },
+        {
+            "loss": 14.2995,
+            "grad_norm": 1.9853943586349487,
+            "learning_rate": 0.0005,
+            "epoch": 0.09120499274505739,
+            "step": 2035
+        },
+        {
+            "loss": 14.1841,
+            "grad_norm": 1.9887142181396484,
+            "learning_rate": 0.0005,
+            "epoch": 0.09142908363632289,
+            "step": 2040
+        },
+        {
+            "loss": 14.1799,
+            "grad_norm": 1.916782259941101,
+            "learning_rate": 0.0005,
+            "epoch": 0.0916531745275884,
+            "step": 2045
+        },
+        {
+            "loss": 14.2815,
+            "grad_norm": 1.9294121265411377,
+            "learning_rate": 0.0005,
+            "epoch": 0.09187726541885388,
+            "step": 2050
+        },
+        {
+            "loss": 14.3043,
+            "grad_norm": 2.086944580078125,
+            "learning_rate": 0.0005,
+            "epoch": 0.09210135631011938,
+            "step": 2055
+        },
+        {
+            "loss": 14.0751,
+            "grad_norm": 2.0607666969299316,
+            "learning_rate": 0.0005,
+            "epoch": 0.09232544720138489,
+            "step": 2060
+        },
+        {
+            "loss": 14.0893,
+            "grad_norm": 1.8890804052352905,
+            "learning_rate": 0.0005,
+            "epoch": 0.09254953809265037,
+            "step": 2065
+        },
+        {
+            "loss": 14.2719,
+            "grad_norm": 2.1385536193847656,
+            "learning_rate": 0.0005,
+            "epoch": 0.09277362898391588,
+            "step": 2070
+        },
+        {
+            "loss": 14.1656,
+            "grad_norm": 2.054858446121216,
+            "learning_rate": 0.0005,
+            "epoch": 0.09299771987518138,
+            "step": 2075
+        },
+        {
+            "loss": 14.1492,
+            "grad_norm": 2.176318645477295,
+            "learning_rate": 0.0005,
+            "epoch": 0.09322181076644687,
+            "step": 2080
+        },
+        {
+            "loss": 14.1967,
+            "grad_norm": 2.424211263656616,
+            "learning_rate": 0.0005,
+            "epoch": 0.09344590165771237,
+            "step": 2085
+        },
+        {
+            "loss": 14.2942,
+            "grad_norm": 2.1716487407684326,
+            "learning_rate": 0.0005,
+            "epoch": 0.09366999254897787,
+            "step": 2090
+        },
+        {
+            "loss": 14.2212,
+            "grad_norm": 1.8533152341842651,
+            "learning_rate": 0.0005,
+            "epoch": 0.09389408344024336,
+            "step": 2095
+        },
+        {
+            "loss": 14.2678,
+            "grad_norm": 1.9658597707748413,
+            "learning_rate": 0.0005,
+            "epoch": 0.09411817433150886,
+            "step": 2100
+        },
+        {
+            "loss": 14.3178,
+            "grad_norm": 1.9647737741470337,
+            "learning_rate": 0.0005,
+            "epoch": 0.09434226522277436,
+            "step": 2105
+        },
+        {
+            "loss": 14.307,
+            "grad_norm": 1.9707311391830444,
+            "learning_rate": 0.0005,
+            "epoch": 0.09456635611403985,
+            "step": 2110
+        },
+        {
+            "loss": 14.2303,
+            "grad_norm": 1.9236079454421997,
+            "learning_rate": 0.0005,
+            "epoch": 0.09479044700530535,
+            "step": 2115
+        },
+        {
+            "loss": 14.1953,
+            "grad_norm": 1.9698580503463745,
+            "learning_rate": 0.0005,
+            "epoch": 0.09501453789657086,
+            "step": 2120
+        },
+        {
+            "loss": 14.2416,
+            "grad_norm": 2.0320985317230225,
+            "learning_rate": 0.0005,
+            "epoch": 0.09523862878783634,
+            "step": 2125
+        },
+        {
+            "loss": 14.3115,
+            "grad_norm": 2.0945701599121094,
+            "learning_rate": 0.0005,
+            "epoch": 0.09546271967910185,
+            "step": 2130
+        },
+        {
+            "loss": 14.2766,
+            "grad_norm": 1.9204962253570557,
+            "learning_rate": 0.0005,
+            "epoch": 0.09568681057036735,
+            "step": 2135
+        },
+        {
+            "loss": 14.1301,
+            "grad_norm": 2.120319366455078,
+            "learning_rate": 0.0005,
+            "epoch": 0.09591090146163284,
+            "step": 2140
+        },
+        {
+            "loss": 14.1893,
+            "grad_norm": 2.18697452545166,
+            "learning_rate": 0.0005,
+            "epoch": 0.09613499235289834,
+            "step": 2145
+        },
+        {
+            "loss": 14.2259,
+            "grad_norm": 2.278101682662964,
+            "learning_rate": 0.0005,
+            "epoch": 0.09635908324416383,
+            "step": 2150
+        },
+        {
+            "loss": 14.2309,
+            "grad_norm": 2.2018258571624756,
+            "learning_rate": 0.0005,
+            "epoch": 0.09658317413542933,
+            "step": 2155
+        },
+        {
+            "loss": 14.2609,
+            "grad_norm": 2.056438684463501,
+            "learning_rate": 0.0005,
+            "epoch": 0.09680726502669483,
+            "step": 2160
+        },
+        {
+            "loss": 14.1959,
+            "grad_norm": 2.1732802391052246,
+            "learning_rate": 0.0005,
+            "epoch": 0.09703135591796032,
+            "step": 2165
+        },
+        {
+            "loss": 14.2884,
+            "grad_norm": 2.277907133102417,
+            "learning_rate": 0.0005,
+            "epoch": 0.09725544680922582,
+            "step": 2170
+        },
+        {
+            "loss": 14.1614,
+            "grad_norm": 2.2980709075927734,
+            "learning_rate": 0.0005,
+            "epoch": 0.09747953770049132,
+            "step": 2175
+        },
+        {
+            "loss": 14.1949,
+            "grad_norm": 2.000433921813965,
+            "learning_rate": 0.0005,
+            "epoch": 0.09770362859175681,
+            "step": 2180
+        },
+        {
+            "loss": 14.1517,
+            "grad_norm": 2.084157705307007,
+            "learning_rate": 0.0005,
+            "epoch": 0.09792771948302231,
+            "step": 2185
+        },
+        {
+            "loss": 14.0785,
+            "grad_norm": 2.104269504547119,
+            "learning_rate": 0.0005,
+            "epoch": 0.09815181037428782,
+            "step": 2190
+        },
+        {
+            "loss": 14.24,
+            "grad_norm": 2.062222719192505,
+            "learning_rate": 0.0005,
+            "epoch": 0.0983759012655533,
+            "step": 2195
+        },
+        {
+            "loss": 14.2399,
+            "grad_norm": 1.9244170188903809,
+            "learning_rate": 0.0005,
+            "epoch": 0.0985999921568188,
+            "step": 2200
+        },
+        {
+            "loss": 14.2444,
+            "grad_norm": 2.064002752304077,
+            "learning_rate": 0.0005,
+            "epoch": 0.09882408304808431,
+            "step": 2205
+        },
+        {
+            "loss": 14.2101,
+            "grad_norm": 1.9590471982955933,
+            "learning_rate": 0.0005,
+            "epoch": 0.0990481739393498,
+            "step": 2210
+        },
+        {
+            "loss": 14.3417,
+            "grad_norm": 2.0932776927948,
+            "learning_rate": 0.0005,
+            "epoch": 0.0992722648306153,
+            "step": 2215
+        },
+        {
+            "loss": 14.1232,
+            "grad_norm": 2.11871600151062,
+            "learning_rate": 0.0005,
+            "epoch": 0.0994963557218808,
+            "step": 2220
+        },
+        {
+            "loss": 14.2301,
+            "grad_norm": 2.137220859527588,
+            "learning_rate": 0.0005,
+            "epoch": 0.09972044661314629,
+            "step": 2225
+        },
+        {
+            "loss": 14.204,
+            "grad_norm": 2.1775405406951904,
+            "learning_rate": 0.0005,
+            "epoch": 0.09994453750441179,
+            "step": 2230
+        },
+        {
+            "loss": 14.1153,
+            "grad_norm": 1.8135520219802856,
+            "learning_rate": 0.0005,
+            "epoch": 0.10016862839567729,
+            "step": 2235
+        },
+        {
+            "loss": 14.0921,
+            "grad_norm": 1.9595603942871094,
+            "learning_rate": 0.0005,
+            "epoch": 0.10039271928694278,
+            "step": 2240
+        },
+        {
+            "loss": 14.1785,
+            "grad_norm": 2.0768790245056152,
+            "learning_rate": 0.0005,
+            "epoch": 0.10061681017820828,
+            "step": 2245
+        },
+        {
+            "loss": 14.3008,
+            "grad_norm": 1.988590955734253,
+            "learning_rate": 0.0005,
+            "epoch": 0.10084090106947378,
+            "step": 2250
+        },
+        {
+            "loss": 14.2555,
+            "grad_norm": 2.1865451335906982,
+            "learning_rate": 0.0005,
+            "epoch": 0.10106499196073927,
+            "step": 2255
+        },
+        {
+            "loss": 14.1862,
+            "grad_norm": 2.0482969284057617,
+            "learning_rate": 0.0005,
+            "epoch": 0.10128908285200477,
+            "step": 2260
+        },
+        {
+            "loss": 14.2719,
+            "grad_norm": 2.082878351211548,
+            "learning_rate": 0.0005,
+            "epoch": 0.10151317374327028,
+            "step": 2265
+        },
+        {
+            "loss": 14.1601,
+            "grad_norm": 2.1429758071899414,
+            "learning_rate": 0.0005,
+            "epoch": 0.10173726463453577,
+            "step": 2270
+        },
+        {
+            "loss": 14.3036,
+            "grad_norm": 1.9471749067306519,
+            "learning_rate": 0.0005,
+            "epoch": 0.10196135552580127,
+            "step": 2275
+        },
+        {
+            "loss": 14.1512,
+            "grad_norm": 1.9544740915298462,
+            "learning_rate": 0.0005,
+            "epoch": 0.10218544641706676,
+            "step": 2280
+        },
+        {
+            "loss": 14.1185,
+            "grad_norm": 1.9881949424743652,
+            "learning_rate": 0.0005,
+            "epoch": 0.10240953730833226,
+            "step": 2285
+        },
+        {
+            "loss": 14.1819,
+            "grad_norm": 2.0492491722106934,
+            "learning_rate": 0.0005,
+            "epoch": 0.10263362819959776,
+            "step": 2290
+        },
+        {
+            "loss": 14.1772,
+            "grad_norm": 1.8486850261688232,
+            "learning_rate": 0.0005,
+            "epoch": 0.10285771909086325,
+            "step": 2295
+        },
+        {
+            "loss": 14.2211,
+            "grad_norm": 1.8952183723449707,
+            "learning_rate": 0.0005,
+            "epoch": 0.10308180998212875,
+            "step": 2300
+        },
+        {
+            "loss": 14.2313,
+            "grad_norm": 2.058659076690674,
+            "learning_rate": 0.0005,
+            "epoch": 0.10330590087339425,
+            "step": 2305
+        },
+        {
+            "loss": 14.1636,
+            "grad_norm": 2.078249454498291,
+            "learning_rate": 0.0005,
+            "epoch": 0.10352999176465974,
+            "step": 2310
+        },
+        {
+            "loss": 14.2255,
+            "grad_norm": 2.116319417953491,
+            "learning_rate": 0.0005,
+            "epoch": 0.10375408265592524,
+            "step": 2315
+        },
+        {
+            "loss": 14.1742,
+            "grad_norm": 1.9836976528167725,
+            "learning_rate": 0.0005,
+            "epoch": 0.10397817354719074,
+            "step": 2320
+        },
+        {
+            "loss": 14.2929,
+            "grad_norm": 1.8880807161331177,
+            "learning_rate": 0.0005,
+            "epoch": 0.10420226443845623,
+            "step": 2325
+        },
+        {
+            "loss": 14.2256,
+            "grad_norm": 2.033595323562622,
+            "learning_rate": 0.0005,
+            "epoch": 0.10442635532972173,
+            "step": 2330
+        },
+        {
+            "loss": 14.2659,
+            "grad_norm": 1.9538575410842896,
+            "learning_rate": 0.0005,
+            "epoch": 0.10465044622098724,
+            "step": 2335
+        },
+        {
+            "loss": 14.2364,
+            "grad_norm": 2.214423179626465,
+            "learning_rate": 0.0005,
+            "epoch": 0.10487453711225272,
+            "step": 2340
+        },
+        {
+            "loss": 14.2227,
+            "grad_norm": 2.110269546508789,
+            "learning_rate": 0.0005,
+            "epoch": 0.10509862800351823,
+            "step": 2345
+        },
+        {
+            "loss": 14.2708,
+            "grad_norm": 2.208690643310547,
+            "learning_rate": 0.0005,
+            "epoch": 0.10532271889478373,
+            "step": 2350
+        },
+        {
+            "loss": 14.2478,
+            "grad_norm": 2.119637966156006,
+            "learning_rate": 0.0005,
+            "epoch": 0.10554680978604922,
+            "step": 2355
+        },
+        {
+            "loss": 14.2098,
+            "grad_norm": 1.8906817436218262,
+            "learning_rate": 0.0005,
+            "epoch": 0.10577090067731472,
+            "step": 2360
+        },
+        {
+            "loss": 14.2196,
+            "grad_norm": 1.96303129196167,
+            "learning_rate": 0.0005,
+            "epoch": 0.10599499156858022,
+            "step": 2365
+        },
+        {
+            "loss": 14.3115,
+            "grad_norm": 1.972203254699707,
+            "learning_rate": 0.0005,
+            "epoch": 0.10621908245984571,
+            "step": 2370
+        },
+        {
+            "loss": 14.2014,
+            "grad_norm": 2.158946990966797,
+            "learning_rate": 0.0005,
+            "epoch": 0.10644317335111121,
+            "step": 2375
+        },
+        {
+            "loss": 14.2611,
+            "grad_norm": 2.0782086849212646,
+            "learning_rate": 0.0005,
+            "epoch": 0.10666726424237671,
+            "step": 2380
+        },
+        {
+            "loss": 14.1658,
+            "grad_norm": 1.9304105043411255,
+            "learning_rate": 0.0005,
+            "epoch": 0.1068913551336422,
+            "step": 2385
+        },
+        {
+            "loss": 14.1135,
+            "grad_norm": 1.9931169748306274,
+            "learning_rate": 0.0005,
+            "epoch": 0.1071154460249077,
+            "step": 2390
+        },
+        {
+            "loss": 14.2804,
+            "grad_norm": 2.020009994506836,
+            "learning_rate": 0.0005,
+            "epoch": 0.1073395369161732,
+            "step": 2395
+        },
+        {
+            "loss": 14.2276,
+            "grad_norm": 2.0295846462249756,
+            "learning_rate": 0.0005,
+            "epoch": 0.1075636278074387,
+            "step": 2400
+        },
+        {
+            "loss": 14.1819,
+            "grad_norm": 1.9559621810913086,
+            "learning_rate": 0.0005,
+            "epoch": 0.1077877186987042,
+            "step": 2405
+        },
+        {
+            "loss": 14.2124,
+            "grad_norm": 2.1522719860076904,
+            "learning_rate": 0.0005,
+            "epoch": 0.1080118095899697,
+            "step": 2410
+        },
+        {
+            "loss": 14.2209,
+            "grad_norm": 1.9611594676971436,
+            "learning_rate": 0.0005,
+            "epoch": 0.10823590048123519,
+            "step": 2415
+        },
+        {
+            "loss": 14.2033,
+            "grad_norm": 2.0471651554107666,
+            "learning_rate": 0.0005,
+            "epoch": 0.10845999137250069,
+            "step": 2420
+        },
+        {
+            "loss": 14.1608,
+            "grad_norm": 2.0396370887756348,
+            "learning_rate": 0.0005,
+            "epoch": 0.10868408226376618,
+            "step": 2425
+        },
+        {
+            "loss": 14.2029,
+            "grad_norm": 1.9847370386123657,
+            "learning_rate": 0.0005,
+            "epoch": 0.10890817315503168,
+            "step": 2430
+        },
+        {
+            "loss": 14.3484,
+            "grad_norm": 1.928900122642517,
+            "learning_rate": 0.0005,
+            "epoch": 0.10913226404629718,
+            "step": 2435
+        },
+        {
+            "loss": 14.1839,
+            "grad_norm": 1.9697222709655762,
+            "learning_rate": 0.0005,
+            "epoch": 0.10935635493756267,
+            "step": 2440
+        },
+        {
+            "loss": 14.3031,
+            "grad_norm": 1.9415478706359863,
+            "learning_rate": 0.0005,
+            "epoch": 0.10958044582882817,
+            "step": 2445
+        },
+        {
+            "loss": 14.1281,
+            "grad_norm": 2.1670243740081787,
+            "learning_rate": 0.0005,
+            "epoch": 0.10980453672009367,
+            "step": 2450
+        },
+        {
+            "loss": 14.2247,
+            "grad_norm": 2.067964792251587,
+            "learning_rate": 0.0005,
+            "epoch": 0.11002862761135916,
+            "step": 2455
+        },
+        {
+            "loss": 14.1618,
+            "grad_norm": 1.9133504629135132,
+            "learning_rate": 0.0005,
+            "epoch": 0.11025271850262466,
+            "step": 2460
+        },
+        {
+            "loss": 14.1767,
+            "grad_norm": 2.117112159729004,
+            "learning_rate": 0.0005,
+            "epoch": 0.11047680939389017,
+            "step": 2465
+        },
+        {
+            "loss": 14.2068,
+            "grad_norm": 2.132375717163086,
+            "learning_rate": 0.0005,
+            "epoch": 0.11070090028515565,
+            "step": 2470
+        },
+        {
+            "loss": 14.1987,
+            "grad_norm": 1.9348459243774414,
+            "learning_rate": 0.0005,
+            "epoch": 0.11092499117642116,
+            "step": 2475
+        },
+        {
+            "loss": 14.0825,
+            "grad_norm": 2.089294910430908,
+            "learning_rate": 0.0005,
+            "epoch": 0.11114908206768666,
+            "step": 2480
+        },
+        {
+            "loss": 14.2549,
+            "grad_norm": 2.2194576263427734,
+            "learning_rate": 0.0005,
+            "epoch": 0.11137317295895215,
+            "step": 2485
+        },
+        {
+            "loss": 14.3019,
+            "grad_norm": 2.019630193710327,
+            "learning_rate": 0.0005,
+            "epoch": 0.11159726385021765,
+            "step": 2490
+        },
+        {
+            "loss": 14.219,
+            "grad_norm": 1.942373275756836,
+            "learning_rate": 0.0005,
+            "epoch": 0.11182135474148315,
+            "step": 2495
+        },
+        {
+            "loss": 14.1376,
+            "grad_norm": 2.0612740516662598,
+            "learning_rate": 0.0005,
+            "epoch": 0.11204544563274864,
+            "step": 2500
+        },
+        {
+            "eval_loss": 1.7685763835906982,
+            "eval_runtime": 18.3338,
+            "eval_samples_per_second": 893.65,
+            "eval_steps_per_second": 8.018,
+            "epoch": 0.11204544563274864,
+            "step": 2500
+        },
+        {
+            "loss": 14.202,
+            "grad_norm": 2.05464768409729,
+            "learning_rate": 0.0005,
+            "epoch": 0.11226953652401414,
+            "step": 2505
+        },
+        {
+            "loss": 14.1914,
+            "grad_norm": 1.9068650007247925,
+            "learning_rate": 0.0005,
+            "epoch": 0.11249362741527964,
+            "step": 2510
+        },
+        {
+            "loss": 14.1411,
+            "grad_norm": 1.9879950284957886,
+            "learning_rate": 0.0005,
+            "epoch": 0.11271771830654513,
+            "step": 2515
+        },
+        {
+            "loss": 14.1364,
+            "grad_norm": 2.096574306488037,
+            "learning_rate": 0.0005,
+            "epoch": 0.11294180919781063,
+            "step": 2520
+        },
+        {
+            "loss": 14.1867,
+            "grad_norm": 1.994155764579773,
+            "learning_rate": 0.0005,
+            "epoch": 0.11316590008907614,
+            "step": 2525
+        },
+        {
+            "loss": 14.1919,
+            "grad_norm": 2.00042724609375,
+            "learning_rate": 0.0005,
+            "epoch": 0.11338999098034162,
+            "step": 2530
+        },
+        {
+            "loss": 14.1076,
+            "grad_norm": 2.0063529014587402,
+            "learning_rate": 0.0005,
+            "epoch": 0.11361408187160713,
+            "step": 2535
+        },
+        {
+            "loss": 14.2082,
+            "grad_norm": 1.990536093711853,
+            "learning_rate": 0.0005,
+            "epoch": 0.11383817276287263,
+            "step": 2540
+        },
+        {
+            "loss": 14.2478,
+            "grad_norm": 2.0046889781951904,
+            "learning_rate": 0.0005,
+            "epoch": 0.11406226365413812,
+            "step": 2545
+        },
+        {
+            "loss": 14.223,
+            "grad_norm": 1.8907625675201416,
+            "learning_rate": 0.0005,
+            "epoch": 0.11428635454540362,
+            "step": 2550
+        },
+        {
+            "loss": 14.1406,
+            "grad_norm": 1.927342176437378,
+            "learning_rate": 0.0005,
+            "epoch": 0.11451044543666912,
+            "step": 2555
+        },
+        {
+            "loss": 14.1912,
+            "grad_norm": 1.8561776876449585,
+            "learning_rate": 0.0005,
+            "epoch": 0.11473453632793461,
+            "step": 2560
+        },
+        {
+            "loss": 14.218,
+            "grad_norm": 1.9795992374420166,
+            "learning_rate": 0.0005,
+            "epoch": 0.11495862721920011,
+            "step": 2565
+        },
+        {
+            "loss": 14.1991,
+            "grad_norm": 2.0302844047546387,
+            "learning_rate": 0.0005,
+            "epoch": 0.1151827181104656,
+            "step": 2570
+        },
+        {
+            "loss": 14.1874,
+            "grad_norm": 1.9552420377731323,
+            "learning_rate": 0.0005,
+            "epoch": 0.1154068090017311,
+            "step": 2575
+        },
+        {
+            "loss": 14.1369,
+            "grad_norm": 2.0299878120422363,
+            "learning_rate": 0.0005,
+            "epoch": 0.1156308998929966,
+            "step": 2580
+        },
+        {
+            "loss": 14.2594,
+            "grad_norm": 2.055478096008301,
+            "learning_rate": 0.0005,
+            "epoch": 0.11585499078426209,
+            "step": 2585
+        },
+        {
+            "loss": 14.2619,
+            "grad_norm": 1.9360374212265015,
+            "learning_rate": 0.0005,
+            "epoch": 0.11607908167552759,
+            "step": 2590
+        },
+        {
+            "loss": 14.0849,
+            "grad_norm": 2.006324529647827,
+            "learning_rate": 0.0005,
+            "epoch": 0.1163031725667931,
+            "step": 2595
+        },
+        {
+            "loss": 14.1619,
+            "grad_norm": 1.85250985622406,
+            "learning_rate": 0.0005,
+            "epoch": 0.11652726345805858,
+            "step": 2600
+        },
+        {
+            "loss": 14.1659,
+            "grad_norm": 1.9738417863845825,
+            "learning_rate": 0.0005,
+            "epoch": 0.11675135434932409,
+            "step": 2605
+        },
+        {
+            "loss": 14.2338,
+            "grad_norm": 1.9255121946334839,
+            "learning_rate": 0.0005,
+            "epoch": 0.11697544524058959,
+            "step": 2610
+        },
+        {
+            "loss": 14.176,
+            "grad_norm": 1.98264741897583,
+            "learning_rate": 0.0005,
+            "epoch": 0.11719953613185508,
+            "step": 2615
+        },
+        {
+            "loss": 14.1467,
+            "grad_norm": 1.8770341873168945,
+            "learning_rate": 0.0005,
+            "epoch": 0.11742362702312058,
+            "step": 2620
+        },
+        {
+            "loss": 14.3661,
+            "grad_norm": 2.094547986984253,
+            "learning_rate": 0.0005,
+            "epoch": 0.11764771791438608,
+            "step": 2625
+        },
+        {
+            "loss": 14.246,
+            "grad_norm": 2.261746644973755,
+            "learning_rate": 0.0005,
+            "epoch": 0.11787180880565157,
+            "step": 2630
+        },
+        {
+            "loss": 14.2903,
+            "grad_norm": 1.9414910078048706,
+            "learning_rate": 0.0005,
+            "epoch": 0.11809589969691707,
+            "step": 2635
+        },
+        {
+            "loss": 14.1791,
+            "grad_norm": 1.837276577949524,
+            "learning_rate": 0.0005,
+            "epoch": 0.11831999058818257,
+            "step": 2640
+        },
+        {
+            "loss": 14.2062,
+            "grad_norm": 2.1287856101989746,
+            "learning_rate": 0.0005,
+            "epoch": 0.11854408147944806,
+            "step": 2645
+        },
+        {
+            "loss": 14.2858,
+            "grad_norm": 2.1157543659210205,
+            "learning_rate": 0.0005,
+            "epoch": 0.11876817237071356,
+            "step": 2650
+        },
+        {
+            "loss": 14.142,
+            "grad_norm": 2.023026704788208,
+            "learning_rate": 0.0005,
+            "epoch": 0.11899226326197906,
+            "step": 2655
+        },
+        {
+            "loss": 14.1913,
+            "grad_norm": 1.9373962879180908,
+            "learning_rate": 0.0005,
+            "epoch": 0.11921635415324455,
+            "step": 2660
+        },
+        {
+            "loss": 14.1843,
+            "grad_norm": 2.1472551822662354,
+            "learning_rate": 0.0005,
+            "epoch": 0.11944044504451005,
+            "step": 2665
+        },
+        {
+            "loss": 14.1469,
+            "grad_norm": 2.104508876800537,
+            "learning_rate": 0.0005,
+            "epoch": 0.11966453593577556,
+            "step": 2670
+        },
+        {
+            "loss": 14.2021,
+            "grad_norm": 1.955299735069275,
+            "learning_rate": 0.0005,
+            "epoch": 0.11988862682704104,
+            "step": 2675
+        },
+        {
+            "loss": 14.2431,
+            "grad_norm": 2.3675167560577393,
+            "learning_rate": 0.0005,
+            "epoch": 0.12011271771830655,
+            "step": 2680
+        },
+        {
+            "loss": 14.2884,
+            "grad_norm": 2.3444430828094482,
+            "learning_rate": 0.0005,
+            "epoch": 0.12033680860957205,
+            "step": 2685
+        },
+        {
+            "loss": 14.187,
+            "grad_norm": 1.919924259185791,
+            "learning_rate": 0.0005,
+            "epoch": 0.12056089950083754,
+            "step": 2690
+        },
+        {
+            "loss": 14.153,
+            "grad_norm": 1.9042751789093018,
+            "learning_rate": 0.0005,
+            "epoch": 0.12078499039210304,
+            "step": 2695
+        },
+        {
+            "loss": 14.2259,
+            "grad_norm": 1.8784211874008179,
+            "learning_rate": 0.0005,
+            "epoch": 0.12100908128336853,
+            "step": 2700
+        },
+        {
+            "loss": 14.1874,
+            "grad_norm": 2.0663719177246094,
+            "learning_rate": 0.0005,
+            "epoch": 0.12123317217463403,
+            "step": 2705
+        },
+        {
+            "loss": 14.2434,
+            "grad_norm": 1.931662917137146,
+            "learning_rate": 0.0005,
+            "epoch": 0.12145726306589953,
+            "step": 2710
+        },
+        {
+            "loss": 14.0864,
+            "grad_norm": 2.011249303817749,
+            "learning_rate": 0.0005,
+            "epoch": 0.12168135395716502,
+            "step": 2715
+        },
+        {
+            "loss": 14.1069,
+            "grad_norm": 2.1195309162139893,
+            "learning_rate": 0.0005,
+            "epoch": 0.12190544484843052,
+            "step": 2720
+        },
+        {
+            "loss": 14.2632,
+            "grad_norm": 2.1942484378814697,
+            "learning_rate": 0.0005,
+            "epoch": 0.12212953573969602,
+            "step": 2725
+        },
+        {
+            "loss": 14.1839,
+            "grad_norm": 2.2845983505249023,
+            "learning_rate": 0.0005,
+            "epoch": 0.12235362663096151,
+            "step": 2730
+        },
+        {
+            "loss": 14.1554,
+            "grad_norm": 1.9804823398590088,
+            "learning_rate": 0.0005,
+            "epoch": 0.12257771752222701,
+            "step": 2735
+        },
+        {
+            "loss": 14.2848,
+            "grad_norm": 2.084580421447754,
+            "learning_rate": 0.0005,
+            "epoch": 0.12280180841349252,
+            "step": 2740
+        },
+        {
+            "loss": 14.2366,
+            "grad_norm": 1.894793152809143,
+            "learning_rate": 0.0005,
+            "epoch": 0.123025899304758,
+            "step": 2745
+        },
+        {
+            "loss": 14.3029,
+            "grad_norm": 2.0080316066741943,
+            "learning_rate": 0.0005,
+            "epoch": 0.1232499901960235,
+            "step": 2750
+        },
+        {
+            "loss": 14.1409,
+            "grad_norm": 1.9836593866348267,
+            "learning_rate": 0.0005,
+            "epoch": 0.12347408108728901,
+            "step": 2755
+        },
+        {
+            "loss": 14.1903,
+            "grad_norm": 2.1319408416748047,
+            "learning_rate": 0.0005,
+            "epoch": 0.1236981719785545,
+            "step": 2760
+        },
+        {
+            "loss": 14.1102,
+            "grad_norm": 2.1290483474731445,
+            "learning_rate": 0.0005,
+            "epoch": 0.12392226286982,
+            "step": 2765
+        },
+        {
+            "loss": 14.2279,
+            "grad_norm": 1.875629186630249,
+            "learning_rate": 0.0005,
+            "epoch": 0.1241463537610855,
+            "step": 2770
+        },
+        {
+            "loss": 14.1881,
+            "grad_norm": 2.0723519325256348,
+            "learning_rate": 0.0005,
+            "epoch": 0.12437044465235099,
+            "step": 2775
+        },
+        {
+            "loss": 14.1952,
+            "grad_norm": 1.951786756515503,
+            "learning_rate": 0.0005,
+            "epoch": 0.12459453554361649,
+            "step": 2780
+        },
+        {
+            "loss": 14.2668,
+            "grad_norm": 2.084223985671997,
+            "learning_rate": 0.0005,
+            "epoch": 0.124818626434882,
+            "step": 2785
+        },
+        {
+            "loss": 14.2497,
+            "grad_norm": 1.9113136529922485,
+            "learning_rate": 0.0005,
+            "epoch": 0.1250427173261475,
+            "step": 2790
+        },
+        {
+            "loss": 14.2409,
+            "grad_norm": 2.0159406661987305,
+            "learning_rate": 0.0005,
+            "epoch": 0.12526680821741298,
+            "step": 2795
+        },
+        {
+            "loss": 14.2057,
+            "grad_norm": 2.006678819656372,
+            "learning_rate": 0.0005,
+            "epoch": 0.12549089910867847,
+            "step": 2800
+        },
+        {
+            "loss": 14.2058,
+            "grad_norm": 1.7884042263031006,
+            "learning_rate": 0.0005,
+            "epoch": 0.125714989999944,
+            "step": 2805
+        },
+        {
+            "loss": 14.2556,
+            "grad_norm": 1.9461385011672974,
+            "learning_rate": 0.0005,
+            "epoch": 0.12593908089120948,
+            "step": 2810
+        },
+        {
+            "loss": 14.1465,
+            "grad_norm": 1.9860858917236328,
+            "learning_rate": 0.0005,
+            "epoch": 0.12616317178247496,
+            "step": 2815
+        },
+        {
+            "loss": 14.2526,
+            "grad_norm": 2.0649843215942383,
+            "learning_rate": 0.0005,
+            "epoch": 0.12638726267374048,
+            "step": 2820
+        },
+        {
+            "loss": 14.1878,
+            "grad_norm": 1.8778871297836304,
+            "learning_rate": 0.0005,
+            "epoch": 0.12661135356500597,
+            "step": 2825
+        },
+        {
+            "loss": 14.1535,
+            "grad_norm": 2.0370824337005615,
+            "learning_rate": 0.0005,
+            "epoch": 0.12683544445627146,
+            "step": 2830
+        },
+        {
+            "loss": 14.2271,
+            "grad_norm": 2.323042154312134,
+            "learning_rate": 0.0005,
+            "epoch": 0.12705953534753697,
+            "step": 2835
+        },
+        {
+            "loss": 14.154,
+            "grad_norm": 2.1211202144622803,
+            "learning_rate": 0.0005,
+            "epoch": 0.12728362623880246,
+            "step": 2840
+        },
+        {
+            "loss": 14.4302,
+            "grad_norm": 2.069763660430908,
+            "learning_rate": 0.0005,
+            "epoch": 0.12750771713006795,
+            "step": 2845
+        },
+        {
+            "loss": 14.1709,
+            "grad_norm": 2.1002161502838135,
+            "learning_rate": 0.0005,
+            "epoch": 0.12773180802133347,
+            "step": 2850
+        },
+        {
+            "loss": 14.1417,
+            "grad_norm": 1.8982101678848267,
+            "learning_rate": 0.0005,
+            "epoch": 0.12795589891259895,
+            "step": 2855
+        },
+        {
+            "loss": 14.1567,
+            "grad_norm": 2.0299010276794434,
+            "learning_rate": 0.0005,
+            "epoch": 0.12817998980386444,
+            "step": 2860
+        },
+        {
+            "loss": 14.2246,
+            "grad_norm": 1.9968671798706055,
+            "learning_rate": 0.0005,
+            "epoch": 0.12840408069512996,
+            "step": 2865
+        },
+        {
+            "loss": 14.1685,
+            "grad_norm": 2.1153697967529297,
+            "learning_rate": 0.0005,
+            "epoch": 0.12862817158639545,
+            "step": 2870
+        },
+        {
+            "loss": 14.1018,
+            "grad_norm": 1.9605722427368164,
+            "learning_rate": 0.0005,
+            "epoch": 0.12885226247766093,
+            "step": 2875
+        },
+        {
+            "loss": 14.2357,
+            "grad_norm": 2.1596014499664307,
+            "learning_rate": 0.0005,
+            "epoch": 0.12907635336892645,
+            "step": 2880
+        },
+        {
+            "loss": 14.1059,
+            "grad_norm": 2.115142583847046,
+            "learning_rate": 0.0005,
+            "epoch": 0.12930044426019194,
+            "step": 2885
+        },
+        {
+            "loss": 14.2208,
+            "grad_norm": 1.9209206104278564,
+            "learning_rate": 0.0005,
+            "epoch": 0.12952453515145743,
+            "step": 2890
+        },
+        {
+            "loss": 14.0913,
+            "grad_norm": 2.015691041946411,
+            "learning_rate": 0.0005,
+            "epoch": 0.12974862604272294,
+            "step": 2895
+        },
+        {
+            "loss": 14.2086,
+            "grad_norm": 2.330225944519043,
+            "learning_rate": 0.0005,
+            "epoch": 0.12997271693398843,
+            "step": 2900
+        },
+        {
+            "loss": 14.1818,
+            "grad_norm": 2.37501859664917,
+            "learning_rate": 0.0005,
+            "epoch": 0.13019680782525392,
+            "step": 2905
+        },
+        {
+            "loss": 14.302,
+            "grad_norm": 1.8006333112716675,
+            "learning_rate": 0.0005,
+            "epoch": 0.1304208987165194,
+            "step": 2910
+        },
+        {
+            "loss": 14.1393,
+            "grad_norm": 1.9731910228729248,
+            "learning_rate": 0.0005,
+            "epoch": 0.13064498960778492,
+            "step": 2915
+        },
+        {
+            "loss": 14.1278,
+            "grad_norm": 2.001185894012451,
+            "learning_rate": 0.0005,
+            "epoch": 0.1308690804990504,
+            "step": 2920
+        },
+        {
+            "loss": 14.2546,
+            "grad_norm": 2.10178542137146,
+            "learning_rate": 0.0005,
+            "epoch": 0.1310931713903159,
+            "step": 2925
+        },
+        {
+            "loss": 14.2663,
+            "grad_norm": 1.9712029695510864,
+            "learning_rate": 0.0005,
+            "epoch": 0.13131726228158142,
+            "step": 2930
+        },
+        {
+            "loss": 14.248,
+            "grad_norm": 2.0637047290802,
+            "learning_rate": 0.0005,
+            "epoch": 0.1315413531728469,
+            "step": 2935
+        },
+        {
+            "loss": 14.2202,
+            "grad_norm": 1.9773246049880981,
+            "learning_rate": 0.0005,
+            "epoch": 0.1317654440641124,
+            "step": 2940
+        },
+        {
+            "loss": 14.1655,
+            "grad_norm": 1.958693027496338,
+            "learning_rate": 0.0005,
+            "epoch": 0.1319895349553779,
+            "step": 2945
+        },
+        {
+            "loss": 14.1329,
+            "grad_norm": 1.9771935939788818,
+            "learning_rate": 0.0005,
+            "epoch": 0.1322136258466434,
+            "step": 2950
+        },
+        {
+            "loss": 14.1144,
+            "grad_norm": 2.0255813598632812,
+            "learning_rate": 0.0005,
+            "epoch": 0.13243771673790888,
+            "step": 2955
+        },
+        {
+            "loss": 14.1486,
+            "grad_norm": 1.958465337753296,
+            "learning_rate": 0.0005,
+            "epoch": 0.1326618076291744,
+            "step": 2960
+        },
+        {
+            "loss": 14.1397,
+            "grad_norm": 1.9276237487792969,
+            "learning_rate": 0.0005,
+            "epoch": 0.1328858985204399,
+            "step": 2965
+        },
+        {
+            "loss": 14.2054,
+            "grad_norm": 2.038614511489868,
+            "learning_rate": 0.0005,
+            "epoch": 0.13310998941170538,
+            "step": 2970
+        },
+        {
+            "loss": 14.1933,
+            "grad_norm": 2.030726671218872,
+            "learning_rate": 0.0005,
+            "epoch": 0.1333340803029709,
+            "step": 2975
+        },
+        {
+            "loss": 14.1404,
+            "grad_norm": 2.0541908740997314,
+            "learning_rate": 0.0005,
+            "epoch": 0.13355817119423638,
+            "step": 2980
+        },
+        {
+            "loss": 14.1175,
+            "grad_norm": 1.962197184562683,
+            "learning_rate": 0.0005,
+            "epoch": 0.13378226208550187,
+            "step": 2985
+        },
+        {
+            "loss": 14.227,
+            "grad_norm": 1.9372152090072632,
+            "learning_rate": 0.0005,
+            "epoch": 0.13400635297676738,
+            "step": 2990
+        },
+        {
+            "loss": 14.2248,
+            "grad_norm": 1.856194019317627,
+            "learning_rate": 0.0005,
+            "epoch": 0.13423044386803287,
+            "step": 2995
+        },
+        {
+            "loss": 14.2369,
+            "grad_norm": 2.008424758911133,
+            "learning_rate": 0.0005,
+            "epoch": 0.13445453475929836,
+            "step": 3000
+        },
+        {
+            "eval_loss": 1.766704797744751,
+            "eval_runtime": 18.3043,
+            "eval_samples_per_second": 895.089,
+            "eval_steps_per_second": 8.031,
+            "epoch": 0.13445453475929836,
+            "step": 3000
+        },
+        {
+            "loss": 14.2562,
+            "grad_norm": 1.9771119356155396,
+            "learning_rate": 0.0005,
+            "epoch": 0.13467862565056388,
+            "step": 3005
+        },
+        {
+            "loss": 14.1959,
+            "grad_norm": 2.1328437328338623,
+            "learning_rate": 0.0005,
+            "epoch": 0.13490271654182937,
+            "step": 3010
+        },
+        {
+            "loss": 14.1737,
+            "grad_norm": 2.1573870182037354,
+            "learning_rate": 0.0005,
+            "epoch": 0.13512680743309485,
+            "step": 3015
+        },
+        {
+            "loss": 14.165,
+            "grad_norm": 1.8604843616485596,
+            "learning_rate": 0.0005,
+            "epoch": 0.13535089832436037,
+            "step": 3020
+        },
+        {
+            "loss": 14.2836,
+            "grad_norm": 1.8671879768371582,
+            "learning_rate": 0.0005,
+            "epoch": 0.13557498921562586,
+            "step": 3025
+        },
+        {
+            "loss": 14.2551,
+            "grad_norm": 2.0154945850372314,
+            "learning_rate": 0.0005,
+            "epoch": 0.13579908010689135,
+            "step": 3030
+        },
+        {
+            "loss": 14.2651,
+            "grad_norm": 1.9040732383728027,
+            "learning_rate": 0.0005,
+            "epoch": 0.13602317099815686,
+            "step": 3035
+        },
+        {
+            "loss": 14.1997,
+            "grad_norm": 2.0263047218322754,
+            "learning_rate": 0.0005,
+            "epoch": 0.13624726188942235,
+            "step": 3040
+        },
+        {
+            "loss": 14.2974,
+            "grad_norm": 1.902062177658081,
+            "learning_rate": 0.0005,
+            "epoch": 0.13647135278068784,
+            "step": 3045
+        },
+        {
+            "loss": 14.1647,
+            "grad_norm": 1.8822523355484009,
+            "learning_rate": 0.0005,
+            "epoch": 0.13669544367195335,
+            "step": 3050
+        },
+        {
+            "loss": 14.1171,
+            "grad_norm": 1.9626497030258179,
+            "learning_rate": 0.0005,
+            "epoch": 0.13691953456321884,
+            "step": 3055
+        },
+        {
+            "loss": 14.1639,
+            "grad_norm": 1.9847556352615356,
+            "learning_rate": 0.0005,
+            "epoch": 0.13714362545448433,
+            "step": 3060
+        },
+        {
+            "loss": 14.3451,
+            "grad_norm": 2.1711082458496094,
+            "learning_rate": 0.0005,
+            "epoch": 0.13736771634574985,
+            "step": 3065
+        },
+        {
+            "loss": 14.2022,
+            "grad_norm": 1.9800173044204712,
+            "learning_rate": 0.0005,
+            "epoch": 0.13759180723701533,
+            "step": 3070
+        },
+        {
+            "loss": 14.3151,
+            "grad_norm": 1.9356379508972168,
+            "learning_rate": 0.0005,
+            "epoch": 0.13781589812828082,
+            "step": 3075
+        },
+        {
+            "loss": 14.194,
+            "grad_norm": 2.027348518371582,
+            "learning_rate": 0.0005,
+            "epoch": 0.13803998901954634,
+            "step": 3080
+        },
+        {
+            "loss": 14.202,
+            "grad_norm": 2.1039271354675293,
+            "learning_rate": 0.0005,
+            "epoch": 0.13826407991081183,
+            "step": 3085
+        },
+        {
+            "loss": 14.2074,
+            "grad_norm": 1.9769928455352783,
+            "learning_rate": 0.0005,
+            "epoch": 0.13848817080207732,
+            "step": 3090
+        },
+        {
+            "loss": 14.1429,
+            "grad_norm": 1.9912936687469482,
+            "learning_rate": 0.0005,
+            "epoch": 0.13871226169334283,
+            "step": 3095
+        },
+        {
+            "loss": 14.2189,
+            "grad_norm": 1.9611599445343018,
+            "learning_rate": 0.0005,
+            "epoch": 0.13893635258460832,
+            "step": 3100
+        },
+        {
+            "loss": 14.2479,
+            "grad_norm": 1.8536235094070435,
+            "learning_rate": 0.0005,
+            "epoch": 0.1391604434758738,
+            "step": 3105
+        },
+        {
+            "loss": 14.2304,
+            "grad_norm": 1.8334366083145142,
+            "learning_rate": 0.0005,
+            "epoch": 0.13938453436713932,
+            "step": 3110
+        },
+        {
+            "loss": 14.1922,
+            "grad_norm": 2.020998239517212,
+            "learning_rate": 0.0005,
+            "epoch": 0.1396086252584048,
+            "step": 3115
+        },
+        {
+            "loss": 14.2457,
+            "grad_norm": 2.0635135173797607,
+            "learning_rate": 0.0005,
+            "epoch": 0.1398327161496703,
+            "step": 3120
+        },
+        {
+            "loss": 14.2277,
+            "grad_norm": 1.9335616827011108,
+            "learning_rate": 0.0005,
+            "epoch": 0.14005680704093582,
+            "step": 3125
+        },
+        {
+            "loss": 14.1735,
+            "grad_norm": 2.0182814598083496,
+            "learning_rate": 0.0005,
+            "epoch": 0.1402808979322013,
+            "step": 3130
+        },
+        {
+            "loss": 14.1644,
+            "grad_norm": 1.9300106763839722,
+            "learning_rate": 0.0005,
+            "epoch": 0.1405049888234668,
+            "step": 3135
+        },
+        {
+            "loss": 14.105,
+            "grad_norm": 1.9784296751022339,
+            "learning_rate": 0.0005,
+            "epoch": 0.1407290797147323,
+            "step": 3140
+        },
+        {
+            "loss": 14.2727,
+            "grad_norm": 1.8297818899154663,
+            "learning_rate": 0.0005,
+            "epoch": 0.1409531706059978,
+            "step": 3145
+        },
+        {
+            "loss": 14.2687,
+            "grad_norm": 2.1447770595550537,
+            "learning_rate": 0.0005,
+            "epoch": 0.14117726149726328,
+            "step": 3150
+        },
+        {
+            "loss": 14.2114,
+            "grad_norm": 1.9135013818740845,
+            "learning_rate": 0.0005,
+            "epoch": 0.1414013523885288,
+            "step": 3155
+        },
+        {
+            "loss": 14.1907,
+            "grad_norm": 2.194002389907837,
+            "learning_rate": 0.0005,
+            "epoch": 0.1416254432797943,
+            "step": 3160
+        },
+        {
+            "loss": 14.1548,
+            "grad_norm": 2.1245758533477783,
+            "learning_rate": 0.0005,
+            "epoch": 0.14184953417105978,
+            "step": 3165
+        },
+        {
+            "loss": 14.1323,
+            "grad_norm": 2.1875216960906982,
+            "learning_rate": 0.0005,
+            "epoch": 0.1420736250623253,
+            "step": 3170
+        },
+        {
+            "loss": 14.1936,
+            "grad_norm": 2.0083394050598145,
+            "learning_rate": 0.0005,
+            "epoch": 0.14229771595359078,
+            "step": 3175
+        },
+        {
+            "loss": 14.1293,
+            "grad_norm": 2.0901074409484863,
+            "learning_rate": 0.0005,
+            "epoch": 0.14252180684485627,
+            "step": 3180
+        },
+        {
+            "loss": 14.2272,
+            "grad_norm": 2.2300620079040527,
+            "learning_rate": 0.0005,
+            "epoch": 0.14274589773612176,
+            "step": 3185
+        },
+        {
+            "loss": 14.178,
+            "grad_norm": 1.8876138925552368,
+            "learning_rate": 0.0005,
+            "epoch": 0.14296998862738727,
+            "step": 3190
+        },
+        {
+            "loss": 14.0993,
+            "grad_norm": 1.8890095949172974,
+            "learning_rate": 0.0005,
+            "epoch": 0.14319407951865276,
+            "step": 3195
+        },
+        {
+            "loss": 14.2384,
+            "grad_norm": 2.007490396499634,
+            "learning_rate": 0.0005,
+            "epoch": 0.14341817040991825,
+            "step": 3200
+        },
+        {
+            "loss": 14.1718,
+            "grad_norm": 1.9662436246871948,
+            "learning_rate": 0.0005,
+            "epoch": 0.14364226130118377,
+            "step": 3205
+        },
+        {
+            "loss": 14.2349,
+            "grad_norm": 2.0031325817108154,
+            "learning_rate": 0.0005,
+            "epoch": 0.14386635219244925,
+            "step": 3210
+        },
+        {
+            "loss": 14.2658,
+            "grad_norm": 2.1796069145202637,
+            "learning_rate": 0.0005,
+            "epoch": 0.14409044308371474,
+            "step": 3215
+        },
+        {
+            "loss": 14.1664,
+            "grad_norm": 2.0716702938079834,
+            "learning_rate": 0.0005,
+            "epoch": 0.14431453397498026,
+            "step": 3220
+        },
+        {
+            "loss": 14.2758,
+            "grad_norm": 1.9255475997924805,
+            "learning_rate": 0.0005,
+            "epoch": 0.14453862486624575,
+            "step": 3225
+        },
+        {
+            "loss": 14.1191,
+            "grad_norm": 2.288928508758545,
+            "learning_rate": 0.0005,
+            "epoch": 0.14476271575751123,
+            "step": 3230
+        },
+        {
+            "loss": 14.0554,
+            "grad_norm": 2.1082675457000732,
+            "learning_rate": 0.0005,
+            "epoch": 0.14498680664877675,
+            "step": 3235
+        },
+        {
+            "loss": 14.1859,
+            "grad_norm": 2.0232932567596436,
+            "learning_rate": 0.0005,
+            "epoch": 0.14521089754004224,
+            "step": 3240
+        },
+        {
+            "loss": 14.1005,
+            "grad_norm": 1.8354517221450806,
+            "learning_rate": 0.0005,
+            "epoch": 0.14543498843130773,
+            "step": 3245
+        },
+        {
+            "loss": 14.1885,
+            "grad_norm": 2.031891345977783,
+            "learning_rate": 0.0005,
+            "epoch": 0.14565907932257324,
+            "step": 3250
+        },
+        {
+            "loss": 14.1925,
+            "grad_norm": 1.9492149353027344,
+            "learning_rate": 0.0005,
+            "epoch": 0.14588317021383873,
+            "step": 3255
+        },
+        {
+            "loss": 14.294,
+            "grad_norm": 2.073660373687744,
+            "learning_rate": 0.0005,
+            "epoch": 0.14610726110510422,
+            "step": 3260
+        },
+        {
+            "loss": 14.1864,
+            "grad_norm": 1.9290986061096191,
+            "learning_rate": 0.0005,
+            "epoch": 0.14633135199636974,
+            "step": 3265
+        },
+        {
+            "loss": 14.1303,
+            "grad_norm": 2.0566937923431396,
+            "learning_rate": 0.0005,
+            "epoch": 0.14655544288763522,
+            "step": 3270
+        },
+        {
+            "loss": 14.027,
+            "grad_norm": 1.9406923055648804,
+            "learning_rate": 0.0005,
+            "epoch": 0.1467795337789007,
+            "step": 3275
+        },
+        {
+            "loss": 14.0723,
+            "grad_norm": 2.021090507507324,
+            "learning_rate": 0.0005,
+            "epoch": 0.14700362467016623,
+            "step": 3280
+        },
+        {
+            "loss": 14.162,
+            "grad_norm": 2.272169589996338,
+            "learning_rate": 0.0005,
+            "epoch": 0.14722771556143172,
+            "step": 3285
+        },
+        {
+            "loss": 14.2405,
+            "grad_norm": 1.9472681283950806,
+            "learning_rate": 0.0005,
+            "epoch": 0.1474518064526972,
+            "step": 3290
+        },
+        {
+            "loss": 14.229,
+            "grad_norm": 1.8901585340499878,
+            "learning_rate": 0.0005,
+            "epoch": 0.14767589734396272,
+            "step": 3295
+        },
+        {
+            "loss": 14.0907,
+            "grad_norm": 1.9590394496917725,
+            "learning_rate": 0.0005,
+            "epoch": 0.1478999882352282,
+            "step": 3300
+        },
+        {
+            "loss": 14.1363,
+            "grad_norm": 2.068774461746216,
+            "learning_rate": 0.0005,
+            "epoch": 0.1481240791264937,
+            "step": 3305
+        },
+        {
+            "loss": 14.2616,
+            "grad_norm": 2.1157450675964355,
+            "learning_rate": 0.0005,
+            "epoch": 0.1483481700177592,
+            "step": 3310
+        },
+        {
+            "loss": 14.167,
+            "grad_norm": 2.029632091522217,
+            "learning_rate": 0.0005,
+            "epoch": 0.1485722609090247,
+            "step": 3315
+        },
+        {
+            "loss": 14.1601,
+            "grad_norm": 2.0851261615753174,
+            "learning_rate": 0.0005,
+            "epoch": 0.1487963518002902,
+            "step": 3320
+        },
+        {
+            "loss": 14.225,
+            "grad_norm": 2.05302357673645,
+            "learning_rate": 0.0005,
+            "epoch": 0.1490204426915557,
+            "step": 3325
+        },
+        {
+            "loss": 14.0864,
+            "grad_norm": 1.9321824312210083,
+            "learning_rate": 0.0005,
+            "epoch": 0.1492445335828212,
+            "step": 3330
+        },
+        {
+            "loss": 14.1827,
+            "grad_norm": 1.9291865825653076,
+            "learning_rate": 0.0005,
+            "epoch": 0.14946862447408668,
+            "step": 3335
+        },
+        {
+            "loss": 14.2404,
+            "grad_norm": 2.1532223224639893,
+            "learning_rate": 0.0005,
+            "epoch": 0.1496927153653522,
+            "step": 3340
+        },
+        {
+            "loss": 14.2338,
+            "grad_norm": 2.166241407394409,
+            "learning_rate": 0.0005,
+            "epoch": 0.14991680625661769,
+            "step": 3345
+        },
+        {
+            "loss": 14.1989,
+            "grad_norm": 2.0716960430145264,
+            "learning_rate": 0.0005,
+            "epoch": 0.15014089714788317,
+            "step": 3350
+        },
+        {
+            "loss": 14.2566,
+            "grad_norm": 2.158968448638916,
+            "learning_rate": 0.0005,
+            "epoch": 0.1503649880391487,
+            "step": 3355
+        },
+        {
+            "loss": 14.2069,
+            "grad_norm": 2.1141438484191895,
+            "learning_rate": 0.0005,
+            "epoch": 0.15058907893041418,
+            "step": 3360
+        },
+        {
+            "loss": 14.151,
+            "grad_norm": 2.0348339080810547,
+            "learning_rate": 0.0005,
+            "epoch": 0.15081316982167967,
+            "step": 3365
+        },
+        {
+            "loss": 14.1337,
+            "grad_norm": 2.09519100189209,
+            "learning_rate": 0.0005,
+            "epoch": 0.15103726071294518,
+            "step": 3370
+        },
+        {
+            "loss": 14.1808,
+            "grad_norm": 2.012434720993042,
+            "learning_rate": 0.0005,
+            "epoch": 0.15126135160421067,
+            "step": 3375
+        },
+        {
+            "loss": 14.0764,
+            "grad_norm": 2.0529088973999023,
+            "learning_rate": 0.0005,
+            "epoch": 0.15148544249547616,
+            "step": 3380
+        },
+        {
+            "loss": 14.1444,
+            "grad_norm": 2.0532617568969727,
+            "learning_rate": 0.0005,
+            "epoch": 0.15170953338674167,
+            "step": 3385
+        },
+        {
+            "loss": 14.2286,
+            "grad_norm": 1.7985155582427979,
+            "learning_rate": 0.0005,
+            "epoch": 0.15193362427800716,
+            "step": 3390
+        },
+        {
+            "loss": 14.1809,
+            "grad_norm": 1.9286701679229736,
+            "learning_rate": 0.0005,
+            "epoch": 0.15215771516927265,
+            "step": 3395
+        },
+        {
+            "loss": 14.1947,
+            "grad_norm": 2.126664161682129,
+            "learning_rate": 0.0005,
+            "epoch": 0.15238180606053817,
+            "step": 3400
+        },
+        {
+            "loss": 14.0935,
+            "grad_norm": 2.0395383834838867,
+            "learning_rate": 0.0005,
+            "epoch": 0.15260589695180365,
+            "step": 3405
+        },
+        {
+            "loss": 14.1364,
+            "grad_norm": 2.0696709156036377,
+            "learning_rate": 0.0005,
+            "epoch": 0.15282998784306914,
+            "step": 3410
+        },
+        {
+            "loss": 14.21,
+            "grad_norm": 1.9847348928451538,
+            "learning_rate": 0.0005,
+            "epoch": 0.15305407873433466,
+            "step": 3415
+        },
+        {
+            "loss": 14.1458,
+            "grad_norm": 1.9850157499313354,
+            "learning_rate": 0.0005,
+            "epoch": 0.15327816962560015,
+            "step": 3420
+        },
+        {
+            "loss": 14.2044,
+            "grad_norm": 2.1267752647399902,
+            "learning_rate": 0.0005,
+            "epoch": 0.15350226051686564,
+            "step": 3425
+        },
+        {
+            "loss": 14.246,
+            "grad_norm": 2.027884006500244,
+            "learning_rate": 0.0005,
+            "epoch": 0.15372635140813115,
+            "step": 3430
+        },
+        {
+            "loss": 14.1692,
+            "grad_norm": 2.0583317279815674,
+            "learning_rate": 0.0005,
+            "epoch": 0.15395044229939664,
+            "step": 3435
+        },
+        {
+            "loss": 14.2038,
+            "grad_norm": 2.07718825340271,
+            "learning_rate": 0.0005,
+            "epoch": 0.15417453319066213,
+            "step": 3440
+        },
+        {
+            "loss": 14.1482,
+            "grad_norm": 1.9352562427520752,
+            "learning_rate": 0.0005,
+            "epoch": 0.15439862408192764,
+            "step": 3445
+        },
+        {
+            "loss": 14.2268,
+            "grad_norm": 1.9634325504302979,
+            "learning_rate": 0.0005,
+            "epoch": 0.15462271497319313,
+            "step": 3450
+        },
+        {
+            "loss": 14.2241,
+            "grad_norm": 2.0379881858825684,
+            "learning_rate": 0.0005,
+            "epoch": 0.15484680586445862,
+            "step": 3455
+        },
+        {
+            "loss": 14.1091,
+            "grad_norm": 1.9497126340866089,
+            "learning_rate": 0.0005,
+            "epoch": 0.1550708967557241,
+            "step": 3460
+        },
+        {
+            "loss": 14.0976,
+            "grad_norm": 2.1084482669830322,
+            "learning_rate": 0.0005,
+            "epoch": 0.15529498764698962,
+            "step": 3465
+        },
+        {
+            "loss": 14.2549,
+            "grad_norm": 1.9887515306472778,
+            "learning_rate": 0.0005,
+            "epoch": 0.1555190785382551,
+            "step": 3470
+        },
+        {
+            "loss": 14.209,
+            "grad_norm": 2.1256349086761475,
+            "learning_rate": 0.0005,
+            "epoch": 0.1557431694295206,
+            "step": 3475
+        },
+        {
+            "loss": 14.1249,
+            "grad_norm": 2.1030185222625732,
+            "learning_rate": 0.0005,
+            "epoch": 0.15596726032078612,
+            "step": 3480
+        },
+        {
+            "loss": 14.2972,
+            "grad_norm": 1.9382493495941162,
+            "learning_rate": 0.0005,
+            "epoch": 0.1561913512120516,
+            "step": 3485
+        },
+        {
+            "loss": 14.226,
+            "grad_norm": 1.9482671022415161,
+            "learning_rate": 0.0005,
+            "epoch": 0.1564154421033171,
+            "step": 3490
+        },
+        {
+            "loss": 14.1851,
+            "grad_norm": 1.9536066055297852,
+            "learning_rate": 0.0005,
+            "epoch": 0.1566395329945826,
+            "step": 3495
+        },
+        {
+            "loss": 14.2262,
+            "grad_norm": 2.197573184967041,
+            "learning_rate": 0.0005,
+            "epoch": 0.1568636238858481,
+            "step": 3500
+        },
+        {
+            "eval_loss": 1.7696559429168701,
+            "eval_runtime": 18.3621,
+            "eval_samples_per_second": 892.271,
+            "eval_steps_per_second": 8.006,
+            "epoch": 0.1568636238858481,
+            "step": 3500
+        },
+        {
+            "loss": 14.1501,
+            "grad_norm": 2.1332786083221436,
+            "learning_rate": 0.0005,
+            "epoch": 0.15708771477711359,
+            "step": 3505
+        },
+        {
+            "loss": 14.1858,
+            "grad_norm": 2.1712486743927,
+            "learning_rate": 0.0005,
+            "epoch": 0.1573118056683791,
+            "step": 3510
+        },
+        {
+            "loss": 14.2522,
+            "grad_norm": 1.9922101497650146,
+            "learning_rate": 0.0005,
+            "epoch": 0.1575358965596446,
+            "step": 3515
+        },
+        {
+            "loss": 14.1805,
+            "grad_norm": 2.2779488563537598,
+            "learning_rate": 0.0005,
+            "epoch": 0.15775998745091008,
+            "step": 3520
+        },
+        {
+            "loss": 14.1793,
+            "grad_norm": 1.9923431873321533,
+            "learning_rate": 0.0005,
+            "epoch": 0.1579840783421756,
+            "step": 3525
+        },
+        {
+            "loss": 14.2578,
+            "grad_norm": 2.1077773571014404,
+            "learning_rate": 0.0005,
+            "epoch": 0.15820816923344108,
+            "step": 3530
+        },
+        {
+            "loss": 14.2033,
+            "grad_norm": 2.0996055603027344,
+            "learning_rate": 0.0005,
+            "epoch": 0.15843226012470657,
+            "step": 3535
+        },
+        {
+            "loss": 14.2472,
+            "grad_norm": 1.8849451541900635,
+            "learning_rate": 0.0005,
+            "epoch": 0.15865635101597209,
+            "step": 3540
+        },
+        {
+            "loss": 14.1729,
+            "grad_norm": 1.9497867822647095,
+            "learning_rate": 0.0005,
+            "epoch": 0.15888044190723757,
+            "step": 3545
+        },
+        {
+            "loss": 14.1113,
+            "grad_norm": 2.1263110637664795,
+            "learning_rate": 0.0005,
+            "epoch": 0.15910453279850306,
+            "step": 3550
+        },
+        {
+            "loss": 14.3157,
+            "grad_norm": 2.093867063522339,
+            "learning_rate": 0.0005,
+            "epoch": 0.15932862368976858,
+            "step": 3555
+        },
+        {
+            "loss": 14.1378,
+            "grad_norm": 1.9930814504623413,
+            "learning_rate": 0.0005,
+            "epoch": 0.15955271458103407,
+            "step": 3560
+        },
+        {
+            "loss": 14.0981,
+            "grad_norm": 1.9953348636627197,
+            "learning_rate": 0.0005,
+            "epoch": 0.15977680547229955,
+            "step": 3565
+        },
+        {
+            "loss": 14.3871,
+            "grad_norm": 2.055655002593994,
+            "learning_rate": 0.0005,
+            "epoch": 0.16000089636356507,
+            "step": 3570
+        },
+        {
+            "loss": 14.2805,
+            "grad_norm": 1.884310007095337,
+            "learning_rate": 0.0005,
+            "epoch": 0.16022498725483056,
+            "step": 3575
+        },
+        {
+            "loss": 14.0668,
+            "grad_norm": 1.9147753715515137,
+            "learning_rate": 0.0005,
+            "epoch": 0.16044907814609605,
+            "step": 3580
+        },
+        {
+            "loss": 14.2115,
+            "grad_norm": 2.0674216747283936,
+            "learning_rate": 0.0005,
+            "epoch": 0.16067316903736156,
+            "step": 3585
+        },
+        {
+            "loss": 14.1922,
+            "grad_norm": 2.0138802528381348,
+            "learning_rate": 0.0005,
+            "epoch": 0.16089725992862705,
+            "step": 3590
+        },
+        {
+            "loss": 14.1058,
+            "grad_norm": 1.9173427820205688,
+            "learning_rate": 0.0005,
+            "epoch": 0.16112135081989254,
+            "step": 3595
+        },
+        {
+            "loss": 14.0634,
+            "grad_norm": 1.9749242067337036,
+            "learning_rate": 0.0005,
+            "epoch": 0.16134544171115806,
+            "step": 3600
+        },
+        {
+            "loss": 14.2739,
+            "grad_norm": 1.9603943824768066,
+            "learning_rate": 0.0005,
+            "epoch": 0.16156953260242354,
+            "step": 3605
+        },
+        {
+            "loss": 14.1966,
+            "grad_norm": 2.004307508468628,
+            "learning_rate": 0.0005,
+            "epoch": 0.16179362349368903,
+            "step": 3610
+        },
+        {
+            "loss": 14.1685,
+            "grad_norm": 1.9626902341842651,
+            "learning_rate": 0.0005,
+            "epoch": 0.16201771438495455,
+            "step": 3615
+        },
+        {
+            "loss": 14.2082,
+            "grad_norm": 1.9217684268951416,
+            "learning_rate": 0.0005,
+            "epoch": 0.16224180527622004,
+            "step": 3620
+        },
+        {
+            "loss": 14.1972,
+            "grad_norm": 2.047498941421509,
+            "learning_rate": 0.0005,
+            "epoch": 0.16246589616748552,
+            "step": 3625
+        },
+        {
+            "loss": 14.1222,
+            "grad_norm": 1.9391909837722778,
+            "learning_rate": 0.0005,
+            "epoch": 0.16268998705875104,
+            "step": 3630
+        },
+        {
+            "loss": 14.1984,
+            "grad_norm": 2.028310775756836,
+            "learning_rate": 0.0005,
+            "epoch": 0.16291407795001653,
+            "step": 3635
+        },
+        {
+            "loss": 14.217,
+            "grad_norm": 2.000718355178833,
+            "learning_rate": 0.0005,
+            "epoch": 0.16313816884128202,
+            "step": 3640
+        },
+        {
+            "loss": 14.2582,
+            "grad_norm": 1.9638919830322266,
+            "learning_rate": 0.0005,
+            "epoch": 0.16336225973254753,
+            "step": 3645
+        },
+        {
+            "loss": 14.1799,
+            "grad_norm": 1.9330832958221436,
+            "learning_rate": 0.0005,
+            "epoch": 0.16358635062381302,
+            "step": 3650
+        },
+        {
+            "loss": 14.1262,
+            "grad_norm": 1.9113759994506836,
+            "learning_rate": 0.0005,
+            "epoch": 0.1638104415150785,
+            "step": 3655
+        },
+        {
+            "loss": 14.2052,
+            "grad_norm": 1.8158360719680786,
+            "learning_rate": 0.0005,
+            "epoch": 0.16403453240634402,
+            "step": 3660
+        },
+        {
+            "loss": 14.201,
+            "grad_norm": 2.2180416584014893,
+            "learning_rate": 0.0005,
+            "epoch": 0.1642586232976095,
+            "step": 3665
+        },
+        {
+            "loss": 14.1913,
+            "grad_norm": 2.4512360095977783,
+            "learning_rate": 0.0005,
+            "epoch": 0.164482714188875,
+            "step": 3670
+        },
+        {
+            "loss": 14.1491,
+            "grad_norm": 2.017465829849243,
+            "learning_rate": 0.0005,
+            "epoch": 0.16470680508014052,
+            "step": 3675
+        },
+        {
+            "loss": 14.1776,
+            "grad_norm": 2.021597146987915,
+            "learning_rate": 0.0005,
+            "epoch": 0.164930895971406,
+            "step": 3680
+        },
+        {
+            "loss": 14.2879,
+            "grad_norm": 2.0513081550598145,
+            "learning_rate": 0.0005,
+            "epoch": 0.1651549868626715,
+            "step": 3685
+        },
+        {
+            "loss": 14.2206,
+            "grad_norm": 1.86613929271698,
+            "learning_rate": 0.0005,
+            "epoch": 0.165379077753937,
+            "step": 3690
+        },
+        {
+            "loss": 14.2824,
+            "grad_norm": 1.9146097898483276,
+            "learning_rate": 0.0005,
+            "epoch": 0.1656031686452025,
+            "step": 3695
+        },
+        {
+            "loss": 14.2447,
+            "grad_norm": 1.893530249595642,
+            "learning_rate": 0.0005,
+            "epoch": 0.16582725953646799,
+            "step": 3700
+        },
+        {
+            "loss": 14.2238,
+            "grad_norm": 2.099381685256958,
+            "learning_rate": 0.0005,
+            "epoch": 0.1660513504277335,
+            "step": 3705
+        },
+        {
+            "loss": 14.2613,
+            "grad_norm": 1.8816338777542114,
+            "learning_rate": 0.0005,
+            "epoch": 0.166275441318999,
+            "step": 3710
+        },
+        {
+            "loss": 14.2852,
+            "grad_norm": 1.8955073356628418,
+            "learning_rate": 0.0005,
+            "epoch": 0.16649953221026448,
+            "step": 3715
+        },
+        {
+            "loss": 14.183,
+            "grad_norm": 2.0466463565826416,
+            "learning_rate": 0.0005,
+            "epoch": 0.16672362310153,
+            "step": 3720
+        },
+        {
+            "loss": 14.2067,
+            "grad_norm": 1.945846438407898,
+            "learning_rate": 0.0005,
+            "epoch": 0.16694771399279548,
+            "step": 3725
+        },
+        {
+            "loss": 14.2256,
+            "grad_norm": 2.201847553253174,
+            "learning_rate": 0.0005,
+            "epoch": 0.16717180488406097,
+            "step": 3730
+        },
+        {
+            "loss": 14.224,
+            "grad_norm": 1.9679917097091675,
+            "learning_rate": 0.0005,
+            "epoch": 0.16739589577532646,
+            "step": 3735
+        },
+        {
+            "loss": 14.2026,
+            "grad_norm": 2.029130697250366,
+            "learning_rate": 0.0005,
+            "epoch": 0.16761998666659197,
+            "step": 3740
+        },
+        {
+            "loss": 14.2399,
+            "grad_norm": 2.055572271347046,
+            "learning_rate": 0.0005,
+            "epoch": 0.16784407755785746,
+            "step": 3745
+        },
+        {
+            "loss": 14.2184,
+            "grad_norm": 2.423161506652832,
+            "learning_rate": 0.0005,
+            "epoch": 0.16806816844912295,
+            "step": 3750
+        },
+        {
+            "loss": 14.2343,
+            "grad_norm": 1.8970690965652466,
+            "learning_rate": 0.0005,
+            "epoch": 0.16829225934038847,
+            "step": 3755
+        },
+        {
+            "loss": 14.1596,
+            "grad_norm": 1.9343767166137695,
+            "learning_rate": 0.0005,
+            "epoch": 0.16851635023165396,
+            "step": 3760
+        },
+        {
+            "loss": 14.1946,
+            "grad_norm": 1.874110221862793,
+            "learning_rate": 0.0005,
+            "epoch": 0.16874044112291944,
+            "step": 3765
+        },
+        {
+            "loss": 14.1475,
+            "grad_norm": 1.7971092462539673,
+            "learning_rate": 0.0005,
+            "epoch": 0.16896453201418496,
+            "step": 3770
+        },
+        {
+            "loss": 14.161,
+            "grad_norm": 1.8645752668380737,
+            "learning_rate": 0.0005,
+            "epoch": 0.16918862290545045,
+            "step": 3775
+        },
+        {
+            "loss": 14.104,
+            "grad_norm": 1.9900041818618774,
+            "learning_rate": 0.0005,
+            "epoch": 0.16941271379671594,
+            "step": 3780
+        },
+        {
+            "loss": 14.1374,
+            "grad_norm": 2.0856997966766357,
+            "learning_rate": 0.0005,
+            "epoch": 0.16963680468798145,
+            "step": 3785
+        },
+        {
+            "loss": 14.1522,
+            "grad_norm": 2.178510904312134,
+            "learning_rate": 0.0005,
+            "epoch": 0.16986089557924694,
+            "step": 3790
+        },
+        {
+            "loss": 14.1531,
+            "grad_norm": 2.0625481605529785,
+            "learning_rate": 0.0005,
+            "epoch": 0.17008498647051243,
+            "step": 3795
+        },
+        {
+            "loss": 14.2027,
+            "grad_norm": 2.0520877838134766,
+            "learning_rate": 0.0005,
+            "epoch": 0.17030907736177794,
+            "step": 3800
+        },
+        {
+            "loss": 14.2521,
+            "grad_norm": 2.1085047721862793,
+            "learning_rate": 0.0005,
+            "epoch": 0.17053316825304343,
+            "step": 3805
+        },
+        {
+            "loss": 14.1758,
+            "grad_norm": 1.987500548362732,
+            "learning_rate": 0.0005,
+            "epoch": 0.17075725914430892,
+            "step": 3810
+        },
+        {
+            "loss": 14.1554,
+            "grad_norm": 2.0057876110076904,
+            "learning_rate": 0.0005,
+            "epoch": 0.17098135003557444,
+            "step": 3815
+        },
+        {
+            "loss": 14.2989,
+            "grad_norm": 1.9037858247756958,
+            "learning_rate": 0.0005,
+            "epoch": 0.17120544092683992,
+            "step": 3820
+        },
+        {
+            "loss": 14.1809,
+            "grad_norm": 1.977766513824463,
+            "learning_rate": 0.0005,
+            "epoch": 0.1714295318181054,
+            "step": 3825
+        },
+        {
+            "loss": 14.3139,
+            "grad_norm": 1.839013934135437,
+            "learning_rate": 0.0005,
+            "epoch": 0.17165362270937093,
+            "step": 3830
+        },
+        {
+            "loss": 14.1497,
+            "grad_norm": 1.875419020652771,
+            "learning_rate": 0.0005,
+            "epoch": 0.17187771360063642,
+            "step": 3835
+        },
+        {
+            "loss": 14.3065,
+            "grad_norm": 1.92861807346344,
+            "learning_rate": 0.0005,
+            "epoch": 0.1721018044919019,
+            "step": 3840
+        },
+        {
+            "loss": 14.1558,
+            "grad_norm": 1.94759202003479,
+            "learning_rate": 0.0005,
+            "epoch": 0.17232589538316742,
+            "step": 3845
+        },
+        {
+            "loss": 14.1194,
+            "grad_norm": 1.9916903972625732,
+            "learning_rate": 0.0005,
+            "epoch": 0.1725499862744329,
+            "step": 3850
+        },
+        {
+            "loss": 14.1304,
+            "grad_norm": 2.2031240463256836,
+            "learning_rate": 0.0005,
+            "epoch": 0.1727740771656984,
+            "step": 3855
+        },
+        {
+            "loss": 14.1791,
+            "grad_norm": 2.153580904006958,
+            "learning_rate": 0.0005,
+            "epoch": 0.1729981680569639,
+            "step": 3860
+        },
+        {
+            "loss": 14.1746,
+            "grad_norm": 2.063750743865967,
+            "learning_rate": 0.0005,
+            "epoch": 0.1732222589482294,
+            "step": 3865
+        },
+        {
+            "loss": 14.1255,
+            "grad_norm": 1.8618475198745728,
+            "learning_rate": 0.0005,
+            "epoch": 0.1734463498394949,
+            "step": 3870
+        },
+        {
+            "loss": 14.1702,
+            "grad_norm": 1.898120641708374,
+            "learning_rate": 0.0005,
+            "epoch": 0.1736704407307604,
+            "step": 3875
+        },
+        {
+            "loss": 14.2209,
+            "grad_norm": 2.050619125366211,
+            "learning_rate": 0.0005,
+            "epoch": 0.1738945316220259,
+            "step": 3880
+        },
+        {
+            "loss": 14.2072,
+            "grad_norm": 2.2624967098236084,
+            "learning_rate": 0.0005,
+            "epoch": 0.17411862251329138,
+            "step": 3885
+        },
+        {
+            "loss": 14.3059,
+            "grad_norm": 1.9534329175949097,
+            "learning_rate": 0.0005,
+            "epoch": 0.1743427134045569,
+            "step": 3890
+        },
+        {
+            "loss": 14.0746,
+            "grad_norm": 1.9066346883773804,
+            "learning_rate": 0.0005,
+            "epoch": 0.1745668042958224,
+            "step": 3895
+        },
+        {
+            "loss": 14.0933,
+            "grad_norm": 2.0286290645599365,
+            "learning_rate": 0.0005,
+            "epoch": 0.17479089518708787,
+            "step": 3900
+        },
+        {
+            "loss": 14.272,
+            "grad_norm": 1.9771586656570435,
+            "learning_rate": 0.0005,
+            "epoch": 0.1750149860783534,
+            "step": 3905
+        },
+        {
+            "loss": 14.1956,
+            "grad_norm": 2.023655652999878,
+            "learning_rate": 0.0005,
+            "epoch": 0.17523907696961888,
+            "step": 3910
+        },
+        {
+            "loss": 14.1492,
+            "grad_norm": 2.0308282375335693,
+            "learning_rate": 0.0005,
+            "epoch": 0.17546316786088437,
+            "step": 3915
+        },
+        {
+            "loss": 14.116,
+            "grad_norm": 1.9718906879425049,
+            "learning_rate": 0.0005,
+            "epoch": 0.17568725875214988,
+            "step": 3920
+        },
+        {
+            "loss": 14.1377,
+            "grad_norm": 2.0543808937072754,
+            "learning_rate": 0.0005,
+            "epoch": 0.17591134964341537,
+            "step": 3925
+        },
+        {
+            "loss": 14.2587,
+            "grad_norm": 1.874050259590149,
+            "learning_rate": 0.0005,
+            "epoch": 0.17613544053468086,
+            "step": 3930
+        },
+        {
+            "loss": 14.0803,
+            "grad_norm": 2.186093807220459,
+            "learning_rate": 0.0005,
+            "epoch": 0.17635953142594638,
+            "step": 3935
+        },
+        {
+            "loss": 14.1585,
+            "grad_norm": 2.1141676902770996,
+            "learning_rate": 0.0005,
+            "epoch": 0.17658362231721186,
+            "step": 3940
+        },
+        {
+            "loss": 14.1433,
+            "grad_norm": 2.111907482147217,
+            "learning_rate": 0.0005,
+            "epoch": 0.17680771320847735,
+            "step": 3945
+        },
+        {
+            "loss": 14.1331,
+            "grad_norm": 2.1772894859313965,
+            "learning_rate": 0.0005,
+            "epoch": 0.17703180409974287,
+            "step": 3950
+        },
+        {
+            "loss": 14.1855,
+            "grad_norm": 2.2366697788238525,
+            "learning_rate": 0.0005,
+            "epoch": 0.17725589499100836,
+            "step": 3955
+        },
+        {
+            "loss": 14.1047,
+            "grad_norm": 2.192366600036621,
+            "learning_rate": 0.0005,
+            "epoch": 0.17747998588227384,
+            "step": 3960
+        },
+        {
+            "loss": 14.1799,
+            "grad_norm": 2.024726152420044,
+            "learning_rate": 0.0005,
+            "epoch": 0.17770407677353936,
+            "step": 3965
+        },
+        {
+            "loss": 14.0991,
+            "grad_norm": 1.8307974338531494,
+            "learning_rate": 0.0005,
+            "epoch": 0.17792816766480485,
+            "step": 3970
+        },
+        {
+            "loss": 14.1659,
+            "grad_norm": 1.9299472570419312,
+            "learning_rate": 0.0005,
+            "epoch": 0.17815225855607034,
+            "step": 3975
+        },
+        {
+            "loss": 14.1707,
+            "grad_norm": 1.914947748184204,
+            "learning_rate": 0.0005,
+            "epoch": 0.17837634944733585,
+            "step": 3980
+        },
+        {
+            "loss": 14.1733,
+            "grad_norm": 2.022740602493286,
+            "learning_rate": 0.0005,
+            "epoch": 0.17860044033860134,
+            "step": 3985
+        },
+        {
+            "loss": 14.2043,
+            "grad_norm": 2.0922834873199463,
+            "learning_rate": 0.0005,
+            "epoch": 0.17882453122986683,
+            "step": 3990
+        },
+        {
+            "loss": 14.2815,
+            "grad_norm": 1.8319189548492432,
+            "learning_rate": 0.0005,
+            "epoch": 0.17904862212113234,
+            "step": 3995
+        },
+        {
+            "loss": 14.2127,
+            "grad_norm": 1.9187498092651367,
+            "learning_rate": 0.0005,
+            "epoch": 0.17927271301239783,
+            "step": 4000
+        },
+        {
+            "eval_loss": 1.7681186199188232,
+            "eval_runtime": 18.1442,
+            "eval_samples_per_second": 902.987,
+            "eval_steps_per_second": 8.102,
+            "epoch": 0.17927271301239783,
+            "step": 4000
+        },
+        {
+            "loss": 14.1477,
+            "grad_norm": 2.096644401550293,
+            "learning_rate": 0.0005,
+            "epoch": 0.17949680390366332,
+            "step": 4005
+        },
+        {
+            "loss": 14.1563,
+            "grad_norm": 2.2901999950408936,
+            "learning_rate": 0.0005,
+            "epoch": 0.1797208947949288,
+            "step": 4010
+        },
+        {
+            "loss": 14.0974,
+            "grad_norm": 2.0580334663391113,
+            "learning_rate": 0.0005,
+            "epoch": 0.17994498568619433,
+            "step": 4015
+        },
+        {
+            "loss": 14.1193,
+            "grad_norm": 1.9283686876296997,
+            "learning_rate": 0.0005,
+            "epoch": 0.1801690765774598,
+            "step": 4020
+        },
+        {
+            "loss": 14.2224,
+            "grad_norm": 2.0037317276000977,
+            "learning_rate": 0.0005,
+            "epoch": 0.1803931674687253,
+            "step": 4025
+        },
+        {
+            "loss": 14.1634,
+            "grad_norm": 2.235990285873413,
+            "learning_rate": 0.0005,
+            "epoch": 0.18061725835999082,
+            "step": 4030
+        },
+        {
+            "loss": 14.1742,
+            "grad_norm": 2.029142141342163,
+            "learning_rate": 0.0005,
+            "epoch": 0.1808413492512563,
+            "step": 4035
+        },
+        {
+            "loss": 14.1899,
+            "grad_norm": 1.978773832321167,
+            "learning_rate": 0.0005,
+            "epoch": 0.1810654401425218,
+            "step": 4040
+        },
+        {
+            "loss": 14.1891,
+            "grad_norm": 2.089905261993408,
+            "learning_rate": 0.0005,
+            "epoch": 0.1812895310337873,
+            "step": 4045
+        },
+        {
+            "loss": 14.218,
+            "grad_norm": 1.9286154508590698,
+            "learning_rate": 0.0005,
+            "epoch": 0.1815136219250528,
+            "step": 4050
+        },
+        {
+            "loss": 14.2234,
+            "grad_norm": 1.9573795795440674,
+            "learning_rate": 0.0005,
+            "epoch": 0.1817377128163183,
+            "step": 4055
+        },
+        {
+            "loss": 14.1842,
+            "grad_norm": 1.9163328409194946,
+            "learning_rate": 0.0005,
+            "epoch": 0.1819618037075838,
+            "step": 4060
+        },
+        {
+            "loss": 14.2729,
+            "grad_norm": 1.882692575454712,
+            "learning_rate": 0.0005,
+            "epoch": 0.1821858945988493,
+            "step": 4065
+        },
+        {
+            "loss": 14.1909,
+            "grad_norm": 2.1156723499298096,
+            "learning_rate": 0.0005,
+            "epoch": 0.18240998549011478,
+            "step": 4070
+        },
+        {
+            "loss": 14.2205,
+            "grad_norm": 1.8549857139587402,
+            "learning_rate": 0.0005,
+            "epoch": 0.1826340763813803,
+            "step": 4075
+        },
+        {
+            "loss": 14.143,
+            "grad_norm": 1.9605876207351685,
+            "learning_rate": 0.0005,
+            "epoch": 0.18285816727264578,
+            "step": 4080
+        },
+        {
+            "loss": 14.1288,
+            "grad_norm": 2.1343271732330322,
+            "learning_rate": 0.0005,
+            "epoch": 0.18308225816391127,
+            "step": 4085
+        },
+        {
+            "loss": 14.2586,
+            "grad_norm": 1.905978798866272,
+            "learning_rate": 0.0005,
+            "epoch": 0.1833063490551768,
+            "step": 4090
+        },
+        {
+            "loss": 14.2387,
+            "grad_norm": 1.9031362533569336,
+            "learning_rate": 0.0005,
+            "epoch": 0.18353043994644228,
+            "step": 4095
+        },
+        {
+            "loss": 14.2686,
+            "grad_norm": 2.0088846683502197,
+            "learning_rate": 0.0005,
+            "epoch": 0.18375453083770776,
+            "step": 4100
+        },
+        {
+            "loss": 14.2028,
+            "grad_norm": 1.8141952753067017,
+            "learning_rate": 0.0005,
+            "epoch": 0.18397862172897328,
+            "step": 4105
+        },
+        {
+            "loss": 14.1979,
+            "grad_norm": 1.85496985912323,
+            "learning_rate": 0.0005,
+            "epoch": 0.18420271262023877,
+            "step": 4110
+        },
+        {
+            "loss": 14.0222,
+            "grad_norm": 1.909380316734314,
+            "learning_rate": 0.0005,
+            "epoch": 0.18442680351150426,
+            "step": 4115
+        },
+        {
+            "loss": 14.0876,
+            "grad_norm": 1.9430705308914185,
+            "learning_rate": 0.0005,
+            "epoch": 0.18465089440276977,
+            "step": 4120
+        },
+        {
+            "loss": 14.1747,
+            "grad_norm": 2.063159704208374,
+            "learning_rate": 0.0005,
+            "epoch": 0.18487498529403526,
+            "step": 4125
+        },
+        {
+            "loss": 14.1689,
+            "grad_norm": 1.9237209558486938,
+            "learning_rate": 0.0005,
+            "epoch": 0.18509907618530075,
+            "step": 4130
+        },
+        {
+            "loss": 14.1428,
+            "grad_norm": 1.9635990858078003,
+            "learning_rate": 0.0005,
+            "epoch": 0.18532316707656626,
+            "step": 4135
+        },
+        {
+            "loss": 14.2495,
+            "grad_norm": 1.9368425607681274,
+            "learning_rate": 0.0005,
+            "epoch": 0.18554725796783175,
+            "step": 4140
+        },
+        {
+            "loss": 14.2331,
+            "grad_norm": 1.946762204170227,
+            "learning_rate": 0.0005,
+            "epoch": 0.18577134885909724,
+            "step": 4145
+        },
+        {
+            "loss": 14.1397,
+            "grad_norm": 1.9915481805801392,
+            "learning_rate": 0.0005,
+            "epoch": 0.18599543975036276,
+            "step": 4150
+        },
+        {
+            "loss": 14.2306,
+            "grad_norm": 2.0128185749053955,
+            "learning_rate": 0.0005,
+            "epoch": 0.18621953064162824,
+            "step": 4155
+        },
+        {
+            "loss": 14.3031,
+            "grad_norm": 1.8977795839309692,
+            "learning_rate": 0.0005,
+            "epoch": 0.18644362153289373,
+            "step": 4160
+        },
+        {
+            "loss": 14.0981,
+            "grad_norm": 1.9194146394729614,
+            "learning_rate": 0.0005,
+            "epoch": 0.18666771242415925,
+            "step": 4165
+        },
+        {
+            "loss": 14.1679,
+            "grad_norm": 1.9983175992965698,
+            "learning_rate": 0.0005,
+            "epoch": 0.18689180331542474,
+            "step": 4170
+        },
+        {
+            "loss": 14.0315,
+            "grad_norm": 2.0112874507904053,
+            "learning_rate": 0.0005,
+            "epoch": 0.18711589420669023,
+            "step": 4175
+        },
+        {
+            "loss": 14.2212,
+            "grad_norm": 2.002476453781128,
+            "learning_rate": 0.0005,
+            "epoch": 0.18733998509795574,
+            "step": 4180
+        },
+        {
+            "loss": 13.9643,
+            "grad_norm": 1.9648174047470093,
+            "learning_rate": 0.0005,
+            "epoch": 0.18756407598922123,
+            "step": 4185
+        },
+        {
+            "loss": 14.2087,
+            "grad_norm": 1.9648782014846802,
+            "learning_rate": 0.0005,
+            "epoch": 0.18778816688048672,
+            "step": 4190
+        },
+        {
+            "loss": 14.1372,
+            "grad_norm": 1.9486149549484253,
+            "learning_rate": 0.0005,
+            "epoch": 0.18801225777175223,
+            "step": 4195
+        },
+        {
+            "loss": 14.0997,
+            "grad_norm": 1.9666614532470703,
+            "learning_rate": 0.0005,
+            "epoch": 0.18823634866301772,
+            "step": 4200
+        },
+        {
+            "loss": 14.1642,
+            "grad_norm": 2.030273675918579,
+            "learning_rate": 0.0005,
+            "epoch": 0.1884604395542832,
+            "step": 4205
+        },
+        {
+            "loss": 14.1444,
+            "grad_norm": 1.8118653297424316,
+            "learning_rate": 0.0005,
+            "epoch": 0.18868453044554873,
+            "step": 4210
+        },
+        {
+            "loss": 14.0818,
+            "grad_norm": 1.9407061338424683,
+            "learning_rate": 0.0005,
+            "epoch": 0.18890862133681421,
+            "step": 4215
+        },
+        {
+            "loss": 14.277,
+            "grad_norm": 1.949091911315918,
+            "learning_rate": 0.0005,
+            "epoch": 0.1891327122280797,
+            "step": 4220
+        },
+        {
+            "loss": 14.1532,
+            "grad_norm": 1.961521863937378,
+            "learning_rate": 0.0005,
+            "epoch": 0.18935680311934522,
+            "step": 4225
+        },
+        {
+            "loss": 14.1745,
+            "grad_norm": 2.182128667831421,
+            "learning_rate": 0.0005,
+            "epoch": 0.1895808940106107,
+            "step": 4230
+        },
+        {
+            "loss": 14.2068,
+            "grad_norm": 2.066096305847168,
+            "learning_rate": 0.0005,
+            "epoch": 0.1898049849018762,
+            "step": 4235
+        },
+        {
+            "loss": 14.1577,
+            "grad_norm": 1.9394439458847046,
+            "learning_rate": 0.0005,
+            "epoch": 0.1900290757931417,
+            "step": 4240
+        },
+        {
+            "loss": 14.1978,
+            "grad_norm": 1.9918142557144165,
+            "learning_rate": 0.0005,
+            "epoch": 0.1902531666844072,
+            "step": 4245
+        },
+        {
+            "loss": 14.1594,
+            "grad_norm": 2.0450599193573,
+            "learning_rate": 0.0005,
+            "epoch": 0.1904772575756727,
+            "step": 4250
+        },
+        {
+            "loss": 14.1575,
+            "grad_norm": 2.207885980606079,
+            "learning_rate": 0.0005,
+            "epoch": 0.1907013484669382,
+            "step": 4255
+        },
+        {
+            "loss": 14.1581,
+            "grad_norm": 1.9571201801300049,
+            "learning_rate": 0.0005,
+            "epoch": 0.1909254393582037,
+            "step": 4260
+        },
+        {
+            "loss": 14.1871,
+            "grad_norm": 2.14985728263855,
+            "learning_rate": 0.0005,
+            "epoch": 0.19114953024946918,
+            "step": 4265
+        },
+        {
+            "loss": 14.2034,
+            "grad_norm": 2.0554187297821045,
+            "learning_rate": 0.0005,
+            "epoch": 0.1913736211407347,
+            "step": 4270
+        },
+        {
+            "loss": 14.07,
+            "grad_norm": 1.8668701648712158,
+            "learning_rate": 0.0005,
+            "epoch": 0.19159771203200018,
+            "step": 4275
+        },
+        {
+            "loss": 14.0869,
+            "grad_norm": 1.7568072080612183,
+            "learning_rate": 0.0005,
+            "epoch": 0.19182180292326567,
+            "step": 4280
+        },
+        {
+            "loss": 14.1629,
+            "grad_norm": 1.8690704107284546,
+            "learning_rate": 0.0005,
+            "epoch": 0.19204589381453116,
+            "step": 4285
+        },
+        {
+            "loss": 14.1132,
+            "grad_norm": 2.0354042053222656,
+            "learning_rate": 0.0005,
+            "epoch": 0.19226998470579668,
+            "step": 4290
+        },
+        {
+            "loss": 14.1111,
+            "grad_norm": 2.053537368774414,
+            "learning_rate": 0.0005,
+            "epoch": 0.19249407559706216,
+            "step": 4295
+        },
+        {
+            "loss": 14.2885,
+            "grad_norm": 1.9747086763381958,
+            "learning_rate": 0.0005,
+            "epoch": 0.19271816648832765,
+            "step": 4300
+        },
+        {
+            "loss": 14.2183,
+            "grad_norm": 2.169306755065918,
+            "learning_rate": 0.0005,
+            "epoch": 0.19294225737959317,
+            "step": 4305
+        },
+        {
+            "loss": 14.1,
+            "grad_norm": 2.0829429626464844,
+            "learning_rate": 0.0005,
+            "epoch": 0.19316634827085866,
+            "step": 4310
+        },
+        {
+            "loss": 14.1872,
+            "grad_norm": 2.137615919113159,
+            "learning_rate": 0.0005,
+            "epoch": 0.19339043916212414,
+            "step": 4315
+        },
+        {
+            "loss": 14.194,
+            "grad_norm": 2.0153584480285645,
+            "learning_rate": 0.0005,
+            "epoch": 0.19361453005338966,
+            "step": 4320
+        },
+        {
+            "loss": 14.2101,
+            "grad_norm": 1.9413926601409912,
+            "learning_rate": 0.0005,
+            "epoch": 0.19383862094465515,
+            "step": 4325
+        },
+        {
+            "loss": 14.1603,
+            "grad_norm": 1.9266632795333862,
+            "learning_rate": 0.0005,
+            "epoch": 0.19406271183592064,
+            "step": 4330
+        },
+        {
+            "loss": 14.1888,
+            "grad_norm": 1.8601921796798706,
+            "learning_rate": 0.0005,
+            "epoch": 0.19428680272718615,
+            "step": 4335
+        },
+        {
+            "loss": 14.0737,
+            "grad_norm": 2.0529356002807617,
+            "learning_rate": 0.0005,
+            "epoch": 0.19451089361845164,
+            "step": 4340
+        },
+        {
+            "loss": 14.1909,
+            "grad_norm": 2.106046438217163,
+            "learning_rate": 0.0005,
+            "epoch": 0.19473498450971713,
+            "step": 4345
+        },
+        {
+            "loss": 14.141,
+            "grad_norm": 1.9242192506790161,
+            "learning_rate": 0.0005,
+            "epoch": 0.19495907540098265,
+            "step": 4350
+        },
+        {
+            "loss": 14.2139,
+            "grad_norm": 2.0597519874572754,
+            "learning_rate": 0.0005,
+            "epoch": 0.19518316629224813,
+            "step": 4355
+        },
+        {
+            "loss": 14.1508,
+            "grad_norm": 2.0324819087982178,
+            "learning_rate": 0.0005,
+            "epoch": 0.19540725718351362,
+            "step": 4360
+        },
+        {
+            "loss": 14.0463,
+            "grad_norm": 1.981508731842041,
+            "learning_rate": 0.0005,
+            "epoch": 0.19563134807477914,
+            "step": 4365
+        },
+        {
+            "loss": 14.1125,
+            "grad_norm": 1.9829838275909424,
+            "learning_rate": 0.0005,
+            "epoch": 0.19585543896604463,
+            "step": 4370
+        },
+        {
+            "loss": 14.1608,
+            "grad_norm": 2.0116312503814697,
+            "learning_rate": 0.0005,
+            "epoch": 0.19607952985731011,
+            "step": 4375
+        },
+        {
+            "loss": 14.1478,
+            "grad_norm": 1.931014895439148,
+            "learning_rate": 0.0005,
+            "epoch": 0.19630362074857563,
+            "step": 4380
+        },
+        {
+            "loss": 14.1709,
+            "grad_norm": 2.123720407485962,
+            "learning_rate": 0.0005,
+            "epoch": 0.19652771163984112,
+            "step": 4385
+        },
+        {
+            "loss": 14.1629,
+            "grad_norm": 2.0556654930114746,
+            "learning_rate": 0.0005,
+            "epoch": 0.1967518025311066,
+            "step": 4390
+        },
+        {
+            "loss": 14.2394,
+            "grad_norm": 1.9314510822296143,
+            "learning_rate": 0.0005,
+            "epoch": 0.19697589342237212,
+            "step": 4395
+        },
+        {
+            "loss": 14.1073,
+            "grad_norm": 2.0276386737823486,
+            "learning_rate": 0.0005,
+            "epoch": 0.1971999843136376,
+            "step": 4400
+        },
+        {
+            "loss": 14.1801,
+            "grad_norm": 1.922781229019165,
+            "learning_rate": 0.0005,
+            "epoch": 0.1974240752049031,
+            "step": 4405
+        },
+        {
+            "loss": 14.2932,
+            "grad_norm": 1.8780455589294434,
+            "learning_rate": 0.0005,
+            "epoch": 0.19764816609616861,
+            "step": 4410
+        },
+        {
+            "loss": 14.2366,
+            "grad_norm": 2.770186185836792,
+            "learning_rate": 0.0005,
+            "epoch": 0.1978722569874341,
+            "step": 4415
+        },
+        {
+            "loss": 14.1106,
+            "grad_norm": 1.9934104681015015,
+            "learning_rate": 0.0005,
+            "epoch": 0.1980963478786996,
+            "step": 4420
+        },
+        {
+            "loss": 14.1122,
+            "grad_norm": 1.9693650007247925,
+            "learning_rate": 0.0005,
+            "epoch": 0.1983204387699651,
+            "step": 4425
+        },
+        {
+            "loss": 14.1088,
+            "grad_norm": 2.167682647705078,
+            "learning_rate": 0.0005,
+            "epoch": 0.1985445296612306,
+            "step": 4430
+        },
+        {
+            "loss": 14.1959,
+            "grad_norm": 1.8895680904388428,
+            "learning_rate": 0.0005,
+            "epoch": 0.19876862055249608,
+            "step": 4435
+        },
+        {
+            "loss": 14.2289,
+            "grad_norm": 1.9719932079315186,
+            "learning_rate": 0.0005,
+            "epoch": 0.1989927114437616,
+            "step": 4440
+        },
+        {
+            "loss": 14.112,
+            "grad_norm": 1.9747672080993652,
+            "learning_rate": 0.0005,
+            "epoch": 0.1992168023350271,
+            "step": 4445
+        },
+        {
+            "loss": 14.0386,
+            "grad_norm": 2.0173239707946777,
+            "learning_rate": 0.0005,
+            "epoch": 0.19944089322629258,
+            "step": 4450
+        },
+        {
+            "loss": 14.1635,
+            "grad_norm": 1.942533254623413,
+            "learning_rate": 0.0005,
+            "epoch": 0.1996649841175581,
+            "step": 4455
+        },
+        {
+            "loss": 14.2371,
+            "grad_norm": 1.8232591152191162,
+            "learning_rate": 0.0005,
+            "epoch": 0.19988907500882358,
+            "step": 4460
+        },
+        {
+            "loss": 14.225,
+            "grad_norm": 1.9957948923110962,
+            "learning_rate": 0.0005,
+            "epoch": 0.20011316590008907,
+            "step": 4465
+        },
+        {
+            "loss": 14.1964,
+            "grad_norm": 1.9297046661376953,
+            "learning_rate": 0.0005,
+            "epoch": 0.20033725679135458,
+            "step": 4470
+        },
+        {
+            "loss": 14.1573,
+            "grad_norm": 2.030123472213745,
+            "learning_rate": 0.0005,
+            "epoch": 0.20056134768262007,
+            "step": 4475
+        },
+        {
+            "loss": 14.2168,
+            "grad_norm": 2.060288906097412,
+            "learning_rate": 0.0005,
+            "epoch": 0.20078543857388556,
+            "step": 4480
+        },
+        {
+            "loss": 14.1218,
+            "grad_norm": 2.0320048332214355,
+            "learning_rate": 0.0005,
+            "epoch": 0.20100952946515108,
+            "step": 4485
+        },
+        {
+            "loss": 14.2131,
+            "grad_norm": 2.079967975616455,
+            "learning_rate": 0.0005,
+            "epoch": 0.20123362035641656,
+            "step": 4490
+        },
+        {
+            "loss": 14.1652,
+            "grad_norm": 2.0145998001098633,
+            "learning_rate": 0.0005,
+            "epoch": 0.20145771124768205,
+            "step": 4495
+        },
+        {
+            "loss": 14.3127,
+            "grad_norm": 1.9463013410568237,
+            "learning_rate": 0.0005,
+            "epoch": 0.20168180213894757,
+            "step": 4500
+        },
+        {
+            "eval_loss": 1.7675774097442627,
+            "eval_runtime": 18.483,
+            "eval_samples_per_second": 886.435,
+            "eval_steps_per_second": 7.953,
+            "epoch": 0.20168180213894757,
+            "step": 4500
+        },
+        {
+            "loss": 14.1472,
+            "grad_norm": 2.087540626525879,
+            "learning_rate": 0.0005,
+            "epoch": 0.20190589303021306,
+            "step": 4505
+        },
+        {
+            "loss": 14.2054,
+            "grad_norm": 1.9299111366271973,
+            "learning_rate": 0.0005,
+            "epoch": 0.20212998392147855,
+            "step": 4510
+        },
+        {
+            "loss": 14.2457,
+            "grad_norm": 1.780938744544983,
+            "learning_rate": 0.0005,
+            "epoch": 0.20235407481274406,
+            "step": 4515
+        },
+        {
+            "loss": 14.075,
+            "grad_norm": 1.9856529235839844,
+            "learning_rate": 0.0005,
+            "epoch": 0.20257816570400955,
+            "step": 4520
+        },
+        {
+            "loss": 14.1967,
+            "grad_norm": 1.8442602157592773,
+            "learning_rate": 0.0005,
+            "epoch": 0.20280225659527504,
+            "step": 4525
+        },
+        {
+            "loss": 14.1787,
+            "grad_norm": 1.90349280834198,
+            "learning_rate": 0.0005,
+            "epoch": 0.20302634748654055,
+            "step": 4530
+        },
+        {
+            "loss": 14.1255,
+            "grad_norm": 2.013941526412964,
+            "learning_rate": 0.0005,
+            "epoch": 0.20325043837780604,
+            "step": 4535
+        },
+        {
+            "loss": 14.208,
+            "grad_norm": 2.1116743087768555,
+            "learning_rate": 0.0005,
+            "epoch": 0.20347452926907153,
+            "step": 4540
+        },
+        {
+            "loss": 14.2598,
+            "grad_norm": 2.4251699447631836,
+            "learning_rate": 0.0005,
+            "epoch": 0.20369862016033705,
+            "step": 4545
+        },
+        {
+            "loss": 14.1809,
+            "grad_norm": 1.930247187614441,
+            "learning_rate": 0.0005,
+            "epoch": 0.20392271105160253,
+            "step": 4550
+        },
+        {
+            "loss": 14.1545,
+            "grad_norm": 1.8743910789489746,
+            "learning_rate": 0.0005,
+            "epoch": 0.20414680194286802,
+            "step": 4555
+        },
+        {
+            "loss": 14.0868,
+            "grad_norm": 2.250181198120117,
+            "learning_rate": 0.0005,
+            "epoch": 0.2043708928341335,
+            "step": 4560
+        },
+        {
+            "loss": 14.1351,
+            "grad_norm": 1.8496135473251343,
+            "learning_rate": 0.0005,
+            "epoch": 0.20459498372539903,
+            "step": 4565
+        },
+        {
+            "loss": 14.2764,
+            "grad_norm": 2.0180184841156006,
+            "learning_rate": 0.0005,
+            "epoch": 0.20481907461666451,
+            "step": 4570
+        },
+        {
+            "loss": 14.2002,
+            "grad_norm": 1.9459477663040161,
+            "learning_rate": 0.0005,
+            "epoch": 0.20504316550793,
+            "step": 4575
+        },
+        {
+            "loss": 14.1513,
+            "grad_norm": 1.914696455001831,
+            "learning_rate": 0.0005,
+            "epoch": 0.20526725639919552,
+            "step": 4580
+        },
+        {
+            "loss": 14.2149,
+            "grad_norm": 1.929095983505249,
+            "learning_rate": 0.0005,
+            "epoch": 0.205491347290461,
+            "step": 4585
+        },
+        {
+            "loss": 14.2086,
+            "grad_norm": 2.1694610118865967,
+            "learning_rate": 0.0005,
+            "epoch": 0.2057154381817265,
+            "step": 4590
+        },
+        {
+            "loss": 14.117,
+            "grad_norm": 1.9861886501312256,
+            "learning_rate": 0.0005,
+            "epoch": 0.205939529072992,
+            "step": 4595
+        },
+        {
+            "loss": 14.1129,
+            "grad_norm": 2.0941860675811768,
+            "learning_rate": 0.0005,
+            "epoch": 0.2061636199642575,
+            "step": 4600
+        },
+        {
+            "loss": 14.2026,
+            "grad_norm": 1.8497081995010376,
+            "learning_rate": 0.0005,
+            "epoch": 0.206387710855523,
+            "step": 4605
+        },
+        {
+            "loss": 14.2216,
+            "grad_norm": 2.1611194610595703,
+            "learning_rate": 0.0005,
+            "epoch": 0.2066118017467885,
+            "step": 4610
+        },
+        {
+            "loss": 14.2318,
+            "grad_norm": 2.006192684173584,
+            "learning_rate": 0.0005,
+            "epoch": 0.206835892638054,
+            "step": 4615
+        },
+        {
+            "loss": 14.251,
+            "grad_norm": 2.033385992050171,
+            "learning_rate": 0.0005,
+            "epoch": 0.20705998352931948,
+            "step": 4620
+        },
+        {
+            "loss": 14.2138,
+            "grad_norm": 1.9639863967895508,
+            "learning_rate": 0.0005,
+            "epoch": 0.207284074420585,
+            "step": 4625
+        },
+        {
+            "loss": 14.1249,
+            "grad_norm": 1.7755391597747803,
+            "learning_rate": 0.0005,
+            "epoch": 0.20750816531185048,
+            "step": 4630
+        },
+        {
+            "loss": 14.2106,
+            "grad_norm": 2.0013914108276367,
+            "learning_rate": 0.0005,
+            "epoch": 0.20773225620311597,
+            "step": 4635
+        },
+        {
+            "loss": 14.2025,
+            "grad_norm": 2.1361021995544434,
+            "learning_rate": 0.0005,
+            "epoch": 0.2079563470943815,
+            "step": 4640
+        },
+        {
+            "loss": 14.0911,
+            "grad_norm": 1.9296854734420776,
+            "learning_rate": 0.0005,
+            "epoch": 0.20818043798564698,
+            "step": 4645
+        },
+        {
+            "loss": 14.1535,
+            "grad_norm": 1.9083620309829712,
+            "learning_rate": 0.0005,
+            "epoch": 0.20840452887691246,
+            "step": 4650
+        },
+        {
+            "loss": 14.1754,
+            "grad_norm": 2.1370511054992676,
+            "learning_rate": 0.0005,
+            "epoch": 0.20862861976817798,
+            "step": 4655
+        },
+        {
+            "loss": 14.1165,
+            "grad_norm": 2.0198771953582764,
+            "learning_rate": 0.0005,
+            "epoch": 0.20885271065944347,
+            "step": 4660
+        },
+        {
+            "loss": 14.1984,
+            "grad_norm": 2.125767230987549,
+            "learning_rate": 0.0005,
+            "epoch": 0.20907680155070896,
+            "step": 4665
+        },
+        {
+            "loss": 14.1492,
+            "grad_norm": 2.205829620361328,
+            "learning_rate": 0.0005,
+            "epoch": 0.20930089244197447,
+            "step": 4670
+        },
+        {
+            "loss": 14.2795,
+            "grad_norm": 2.207564353942871,
+            "learning_rate": 0.0005,
+            "epoch": 0.20952498333323996,
+            "step": 4675
+        },
+        {
+            "loss": 14.0954,
+            "grad_norm": 2.261016368865967,
+            "learning_rate": 0.0005,
+            "epoch": 0.20974907422450545,
+            "step": 4680
+        },
+        {
+            "loss": 14.1751,
+            "grad_norm": 2.4380059242248535,
+            "learning_rate": 0.0005,
+            "epoch": 0.20997316511577097,
+            "step": 4685
+        },
+        {
+            "loss": 14.1239,
+            "grad_norm": 2.4530105590820312,
+            "learning_rate": 0.0005,
+            "epoch": 0.21019725600703645,
+            "step": 4690
+        },
+        {
+            "loss": 14.0755,
+            "grad_norm": 2.1958963871002197,
+            "learning_rate": 0.0005,
+            "epoch": 0.21042134689830194,
+            "step": 4695
+        },
+        {
+            "loss": 14.1825,
+            "grad_norm": 1.959038257598877,
+            "learning_rate": 0.0005,
+            "epoch": 0.21064543778956746,
+            "step": 4700
+        },
+        {
+            "loss": 14.0358,
+            "grad_norm": 1.8986763954162598,
+            "learning_rate": 0.0005,
+            "epoch": 0.21086952868083295,
+            "step": 4705
+        },
+        {
+            "loss": 14.033,
+            "grad_norm": 1.932701826095581,
+            "learning_rate": 0.0005,
+            "epoch": 0.21109361957209843,
+            "step": 4710
+        },
+        {
+            "loss": 14.1792,
+            "grad_norm": 1.8619905710220337,
+            "learning_rate": 0.0005,
+            "epoch": 0.21131771046336395,
+            "step": 4715
+        },
+        {
+            "loss": 14.0447,
+            "grad_norm": 1.9640110731124878,
+            "learning_rate": 0.0005,
+            "epoch": 0.21154180135462944,
+            "step": 4720
+        },
+        {
+            "loss": 14.0722,
+            "grad_norm": 1.8437641859054565,
+            "learning_rate": 0.0005,
+            "epoch": 0.21176589224589493,
+            "step": 4725
+        },
+        {
+            "loss": 14.1856,
+            "grad_norm": 1.8629305362701416,
+            "learning_rate": 0.0005,
+            "epoch": 0.21198998313716044,
+            "step": 4730
+        },
+        {
+            "loss": 14.1092,
+            "grad_norm": 2.0573370456695557,
+            "learning_rate": 0.0005,
+            "epoch": 0.21221407402842593,
+            "step": 4735
+        },
+        {
+            "loss": 14.1854,
+            "grad_norm": 1.941266417503357,
+            "learning_rate": 0.0005,
+            "epoch": 0.21243816491969142,
+            "step": 4740
+        },
+        {
+            "loss": 14.2779,
+            "grad_norm": 1.9443109035491943,
+            "learning_rate": 0.0005,
+            "epoch": 0.21266225581095693,
+            "step": 4745
+        },
+        {
+            "loss": 14.1528,
+            "grad_norm": 2.157406806945801,
+            "learning_rate": 0.0005,
+            "epoch": 0.21288634670222242,
+            "step": 4750
+        },
+        {
+            "loss": 14.2898,
+            "grad_norm": 1.987005591392517,
+            "learning_rate": 0.0005,
+            "epoch": 0.2131104375934879,
+            "step": 4755
+        },
+        {
+            "loss": 14.2019,
+            "grad_norm": 1.8471060991287231,
+            "learning_rate": 0.0005,
+            "epoch": 0.21333452848475343,
+            "step": 4760
+        },
+        {
+            "loss": 14.1338,
+            "grad_norm": 1.946158766746521,
+            "learning_rate": 0.0005,
+            "epoch": 0.21355861937601892,
+            "step": 4765
+        },
+        {
+            "loss": 14.1958,
+            "grad_norm": 1.9621422290802002,
+            "learning_rate": 0.0005,
+            "epoch": 0.2137827102672844,
+            "step": 4770
+        },
+        {
+            "loss": 14.1579,
+            "grad_norm": 1.9724212884902954,
+            "learning_rate": 0.0005,
+            "epoch": 0.21400680115854992,
+            "step": 4775
+        },
+        {
+            "loss": 14.1578,
+            "grad_norm": 2.133413076400757,
+            "learning_rate": 0.0005,
+            "epoch": 0.2142308920498154,
+            "step": 4780
+        },
+        {
+            "loss": 14.1889,
+            "grad_norm": 2.089862823486328,
+            "learning_rate": 0.0005,
+            "epoch": 0.2144549829410809,
+            "step": 4785
+        },
+        {
+            "loss": 14.1868,
+            "grad_norm": 2.0941824913024902,
+            "learning_rate": 0.0005,
+            "epoch": 0.2146790738323464,
+            "step": 4790
+        },
+        {
+            "loss": 14.1852,
+            "grad_norm": 2.1381120681762695,
+            "learning_rate": 0.0005,
+            "epoch": 0.2149031647236119,
+            "step": 4795
+        },
+        {
+            "loss": 14.1893,
+            "grad_norm": 1.9739441871643066,
+            "learning_rate": 0.0005,
+            "epoch": 0.2151272556148774,
+            "step": 4800
+        },
+        {
+            "loss": 14.1212,
+            "grad_norm": 2.1892285346984863,
+            "learning_rate": 0.0005,
+            "epoch": 0.2153513465061429,
+            "step": 4805
+        },
+        {
+            "loss": 14.216,
+            "grad_norm": 1.9279417991638184,
+            "learning_rate": 0.0005,
+            "epoch": 0.2155754373974084,
+            "step": 4810
+        },
+        {
+            "loss": 14.0294,
+            "grad_norm": 1.9074068069458008,
+            "learning_rate": 0.0005,
+            "epoch": 0.21579952828867388,
+            "step": 4815
+        },
+        {
+            "loss": 14.2165,
+            "grad_norm": 1.9555555582046509,
+            "learning_rate": 0.0005,
+            "epoch": 0.2160236191799394,
+            "step": 4820
+        },
+        {
+            "loss": 14.1804,
+            "grad_norm": 1.9695144891738892,
+            "learning_rate": 0.0005,
+            "epoch": 0.21624771007120488,
+            "step": 4825
+        },
+        {
+            "loss": 14.0645,
+            "grad_norm": 2.063330888748169,
+            "learning_rate": 0.0005,
+            "epoch": 0.21647180096247037,
+            "step": 4830
+        },
+        {
+            "loss": 14.1582,
+            "grad_norm": 1.9090768098831177,
+            "learning_rate": 0.0005,
+            "epoch": 0.2166958918537359,
+            "step": 4835
+        },
+        {
+            "loss": 14.1237,
+            "grad_norm": 2.066309690475464,
+            "learning_rate": 0.0005,
+            "epoch": 0.21691998274500138,
+            "step": 4840
+        },
+        {
+            "loss": 14.121,
+            "grad_norm": 1.9688724279403687,
+            "learning_rate": 0.0005,
+            "epoch": 0.21714407363626687,
+            "step": 4845
+        },
+        {
+            "loss": 14.0986,
+            "grad_norm": 2.0572214126586914,
+            "learning_rate": 0.0005,
+            "epoch": 0.21736816452753235,
+            "step": 4850
+        },
+        {
+            "loss": 14.1856,
+            "grad_norm": 1.8715074062347412,
+            "learning_rate": 0.0005,
+            "epoch": 0.21759225541879787,
+            "step": 4855
+        },
+        {
+            "loss": 14.1442,
+            "grad_norm": 1.880017876625061,
+            "learning_rate": 0.0005,
+            "epoch": 0.21781634631006336,
+            "step": 4860
+        },
+        {
+            "loss": 14.0823,
+            "grad_norm": 2.1221022605895996,
+            "learning_rate": 0.0005,
+            "epoch": 0.21804043720132885,
+            "step": 4865
+        },
+        {
+            "loss": 14.1402,
+            "grad_norm": 1.8102777004241943,
+            "learning_rate": 0.0005,
+            "epoch": 0.21826452809259436,
+            "step": 4870
+        },
+        {
+            "loss": 14.0999,
+            "grad_norm": 2.0177509784698486,
+            "learning_rate": 0.0005,
+            "epoch": 0.21848861898385985,
+            "step": 4875
+        },
+        {
+            "loss": 14.156,
+            "grad_norm": 1.8971713781356812,
+            "learning_rate": 0.0005,
+            "epoch": 0.21871270987512534,
+            "step": 4880
+        },
+        {
+            "loss": 14.1406,
+            "grad_norm": 1.9269769191741943,
+            "learning_rate": 0.0005,
+            "epoch": 0.21893680076639085,
+            "step": 4885
+        },
+        {
+            "loss": 14.062,
+            "grad_norm": 1.8797369003295898,
+            "learning_rate": 0.0005,
+            "epoch": 0.21916089165765634,
+            "step": 4890
+        },
+        {
+            "loss": 14.1249,
+            "grad_norm": 1.9738274812698364,
+            "learning_rate": 0.0005,
+            "epoch": 0.21938498254892183,
+            "step": 4895
+        },
+        {
+            "loss": 14.0797,
+            "grad_norm": 1.9180667400360107,
+            "learning_rate": 0.0005,
+            "epoch": 0.21960907344018735,
+            "step": 4900
+        },
+        {
+            "loss": 14.1443,
+            "grad_norm": 1.7913436889648438,
+            "learning_rate": 0.0005,
+            "epoch": 0.21983316433145283,
+            "step": 4905
+        },
+        {
+            "loss": 14.1768,
+            "grad_norm": 2.0899739265441895,
+            "learning_rate": 0.0005,
+            "epoch": 0.22005725522271832,
+            "step": 4910
+        },
+        {
+            "loss": 14.1679,
+            "grad_norm": 2.132495164871216,
+            "learning_rate": 0.0005,
+            "epoch": 0.22028134611398384,
+            "step": 4915
+        },
+        {
+            "loss": 14.1287,
+            "grad_norm": 1.8545218706130981,
+            "learning_rate": 0.0005,
+            "epoch": 0.22050543700524933,
+            "step": 4920
+        },
+        {
+            "loss": 14.0069,
+            "grad_norm": 1.779749870300293,
+            "learning_rate": 0.0005,
+            "epoch": 0.22072952789651482,
+            "step": 4925
+        },
+        {
+            "loss": 14.0672,
+            "grad_norm": 1.7925384044647217,
+            "learning_rate": 0.0005,
+            "epoch": 0.22095361878778033,
+            "step": 4930
+        },
+        {
+            "loss": 14.0624,
+            "grad_norm": 1.9922071695327759,
+            "learning_rate": 0.0005,
+            "epoch": 0.22117770967904582,
+            "step": 4935
+        },
+        {
+            "loss": 14.0915,
+            "grad_norm": 1.8443610668182373,
+            "learning_rate": 0.0005,
+            "epoch": 0.2214018005703113,
+            "step": 4940
+        },
+        {
+            "loss": 14.1989,
+            "grad_norm": 2.111323833465576,
+            "learning_rate": 0.0005,
+            "epoch": 0.22162589146157682,
+            "step": 4945
+        },
+        {
+            "loss": 14.2195,
+            "grad_norm": 2.143103837966919,
+            "learning_rate": 0.0005,
+            "epoch": 0.2218499823528423,
+            "step": 4950
+        },
+        {
+            "loss": 14.0956,
+            "grad_norm": 1.8921234607696533,
+            "learning_rate": 0.0005,
+            "epoch": 0.2220740732441078,
+            "step": 4955
+        },
+        {
+            "loss": 14.1515,
+            "grad_norm": 1.8157905340194702,
+            "learning_rate": 0.0005,
+            "epoch": 0.22229816413537332,
+            "step": 4960
+        },
+        {
+            "loss": 14.192,
+            "grad_norm": 1.8960933685302734,
+            "learning_rate": 0.0005,
+            "epoch": 0.2225222550266388,
+            "step": 4965
+        },
+        {
+            "loss": 14.2119,
+            "grad_norm": 2.09665584564209,
+            "learning_rate": 0.0005,
+            "epoch": 0.2227463459179043,
+            "step": 4970
+        },
+        {
+            "loss": 14.0471,
+            "grad_norm": 1.929456114768982,
+            "learning_rate": 0.0005,
+            "epoch": 0.2229704368091698,
+            "step": 4975
+        },
+        {
+            "loss": 14.2054,
+            "grad_norm": 1.9969825744628906,
+            "learning_rate": 0.0005,
+            "epoch": 0.2231945277004353,
+            "step": 4980
+        },
+        {
+            "loss": 14.1176,
+            "grad_norm": 2.014631748199463,
+            "learning_rate": 0.0005,
+            "epoch": 0.22341861859170078,
+            "step": 4985
+        },
+        {
+            "loss": 14.1314,
+            "grad_norm": 2.0011658668518066,
+            "learning_rate": 0.0005,
+            "epoch": 0.2236427094829663,
+            "step": 4990
+        },
+        {
+            "loss": 14.0723,
+            "grad_norm": 1.908673882484436,
+            "learning_rate": 0.0005,
+            "epoch": 0.2238668003742318,
+            "step": 4995
+        },
+        {
+            "loss": 14.1676,
+            "grad_norm": 1.9253995418548584,
+            "learning_rate": 0.0005,
+            "epoch": 0.22409089126549728,
+            "step": 5000
+        },
+        {
+            "eval_loss": 1.75691819190979,
+            "eval_runtime": 18.4767,
+            "eval_samples_per_second": 886.736,
+            "eval_steps_per_second": 7.956,
+            "epoch": 0.22409089126549728,
+            "step": 5000
+        },
+        {
+            "loss": 14.1952,
+            "grad_norm": 1.9407540559768677,
+            "learning_rate": 0.0005,
+            "epoch": 0.2243149821567628,
+            "step": 5005
+        },
+        {
+            "loss": 14.2406,
+            "grad_norm": 1.813610315322876,
+            "learning_rate": 0.0005,
+            "epoch": 0.22453907304802828,
+            "step": 5010
+        },
+        {
+            "loss": 14.143,
+            "grad_norm": 1.9335557222366333,
+            "learning_rate": 0.0005,
+            "epoch": 0.22476316393929377,
+            "step": 5015
+        },
+        {
+            "loss": 14.1897,
+            "grad_norm": 1.9015038013458252,
+            "learning_rate": 0.0005,
+            "epoch": 0.22498725483055929,
+            "step": 5020
+        },
+        {
+            "loss": 14.2873,
+            "grad_norm": 2.3834633827209473,
+            "learning_rate": 0.0005,
+            "epoch": 0.22521134572182477,
+            "step": 5025
+        },
+        {
+            "loss": 14.0845,
+            "grad_norm": 1.96584951877594,
+            "learning_rate": 0.0005,
+            "epoch": 0.22543543661309026,
+            "step": 5030
+        },
+        {
+            "loss": 14.1636,
+            "grad_norm": 2.146692991256714,
+            "learning_rate": 0.0005,
+            "epoch": 0.22565952750435578,
+            "step": 5035
+        },
+        {
+            "loss": 14.2176,
+            "grad_norm": 1.9806510210037231,
+            "learning_rate": 0.0005,
+            "epoch": 0.22588361839562127,
+            "step": 5040
+        },
+        {
+            "loss": 14.2031,
+            "grad_norm": 1.9274729490280151,
+            "learning_rate": 0.0005,
+            "epoch": 0.22610770928688675,
+            "step": 5045
+        },
+        {
+            "loss": 14.1723,
+            "grad_norm": 1.7561886310577393,
+            "learning_rate": 0.0005,
+            "epoch": 0.22633180017815227,
+            "step": 5050
+        },
+        {
+            "loss": 14.1663,
+            "grad_norm": 1.9986470937728882,
+            "learning_rate": 0.0005,
+            "epoch": 0.22655589106941776,
+            "step": 5055
+        },
+        {
+            "loss": 14.1284,
+            "grad_norm": 2.1279730796813965,
+            "learning_rate": 0.0005,
+            "epoch": 0.22677998196068325,
+            "step": 5060
+        },
+        {
+            "loss": 14.0967,
+            "grad_norm": 1.9476654529571533,
+            "learning_rate": 0.0005,
+            "epoch": 0.22700407285194876,
+            "step": 5065
+        },
+        {
+            "loss": 14.1572,
+            "grad_norm": 1.9759037494659424,
+            "learning_rate": 0.0005,
+            "epoch": 0.22722816374321425,
+            "step": 5070
+        },
+        {
+            "loss": 14.0725,
+            "grad_norm": 2.016960620880127,
+            "learning_rate": 0.0005,
+            "epoch": 0.22745225463447974,
+            "step": 5075
+        },
+        {
+            "loss": 14.2644,
+            "grad_norm": 2.0249178409576416,
+            "learning_rate": 0.0005,
+            "epoch": 0.22767634552574526,
+            "step": 5080
+        },
+        {
+            "loss": 14.1046,
+            "grad_norm": 1.9025015830993652,
+            "learning_rate": 0.0005,
+            "epoch": 0.22790043641701074,
+            "step": 5085
+        },
+        {
+            "loss": 14.135,
+            "grad_norm": 2.043222665786743,
+            "learning_rate": 0.0005,
+            "epoch": 0.22812452730827623,
+            "step": 5090
+        },
+        {
+            "loss": 14.2218,
+            "grad_norm": 1.9227269887924194,
+            "learning_rate": 0.0005,
+            "epoch": 0.22834861819954175,
+            "step": 5095
+        },
+        {
+            "loss": 14.1908,
+            "grad_norm": 1.9823837280273438,
+            "learning_rate": 0.0005,
+            "epoch": 0.22857270909080724,
+            "step": 5100
+        },
+        {
+            "loss": 14.2474,
+            "grad_norm": 2.08957839012146,
+            "learning_rate": 0.0005,
+            "epoch": 0.22879679998207272,
+            "step": 5105
+        },
+        {
+            "loss": 14.1789,
+            "grad_norm": 1.8854748010635376,
+            "learning_rate": 0.0005,
+            "epoch": 0.22902089087333824,
+            "step": 5110
+        },
+        {
+            "loss": 14.143,
+            "grad_norm": 1.9586538076400757,
+            "learning_rate": 0.0005,
+            "epoch": 0.22924498176460373,
+            "step": 5115
+        },
+        {
+            "loss": 14.2054,
+            "grad_norm": 2.168851137161255,
+            "learning_rate": 0.0005,
+            "epoch": 0.22946907265586922,
+            "step": 5120
+        },
+        {
+            "loss": 14.2061,
+            "grad_norm": 1.938035249710083,
+            "learning_rate": 0.0005,
+            "epoch": 0.2296931635471347,
+            "step": 5125
+        },
+        {
+            "loss": 14.113,
+            "grad_norm": 1.8732202053070068,
+            "learning_rate": 0.0005,
+            "epoch": 0.22991725443840022,
+            "step": 5130
+        },
+        {
+            "loss": 14.0918,
+            "grad_norm": 1.9816375970840454,
+            "learning_rate": 0.0005,
+            "epoch": 0.2301413453296657,
+            "step": 5135
+        },
+        {
+            "loss": 14.1993,
+            "grad_norm": 2.076524257659912,
+            "learning_rate": 0.0005,
+            "epoch": 0.2303654362209312,
+            "step": 5140
+        },
+        {
+            "loss": 14.2595,
+            "grad_norm": 2.0364131927490234,
+            "learning_rate": 0.0005,
+            "epoch": 0.2305895271121967,
+            "step": 5145
+        },
+        {
+            "loss": 14.1889,
+            "grad_norm": 1.949928641319275,
+            "learning_rate": 0.0005,
+            "epoch": 0.2308136180034622,
+            "step": 5150
+        },
+        {
+            "loss": 14.0974,
+            "grad_norm": 1.966451644897461,
+            "learning_rate": 0.0005,
+            "epoch": 0.2310377088947277,
+            "step": 5155
+        },
+        {
+            "loss": 14.2536,
+            "grad_norm": 1.9429701566696167,
+            "learning_rate": 0.0005,
+            "epoch": 0.2312617997859932,
+            "step": 5160
+        },
+        {
+            "loss": 14.0996,
+            "grad_norm": 1.9530205726623535,
+            "learning_rate": 0.0005,
+            "epoch": 0.2314858906772587,
+            "step": 5165
+        },
+        {
+            "loss": 14.2591,
+            "grad_norm": 2.3076071739196777,
+            "learning_rate": 0.0005,
+            "epoch": 0.23170998156852418,
+            "step": 5170
+        },
+        {
+            "loss": 14.2217,
+            "grad_norm": 2.1827287673950195,
+            "learning_rate": 0.0005,
+            "epoch": 0.2319340724597897,
+            "step": 5175
+        },
+        {
+            "loss": 14.1556,
+            "grad_norm": 1.8181138038635254,
+            "learning_rate": 0.0005,
+            "epoch": 0.23215816335105519,
+            "step": 5180
+        },
+        {
+            "loss": 14.1436,
+            "grad_norm": 2.006176233291626,
+            "learning_rate": 0.0005,
+            "epoch": 0.23238225424232067,
+            "step": 5185
+        },
+        {
+            "loss": 14.1115,
+            "grad_norm": 1.8907212018966675,
+            "learning_rate": 0.0005,
+            "epoch": 0.2326063451335862,
+            "step": 5190
+        },
+        {
+            "loss": 14.1021,
+            "grad_norm": 1.937179684638977,
+            "learning_rate": 0.0005,
+            "epoch": 0.23283043602485168,
+            "step": 5195
+        },
+        {
+            "loss": 14.2557,
+            "grad_norm": 2.1188528537750244,
+            "learning_rate": 0.0005,
+            "epoch": 0.23305452691611717,
+            "step": 5200
+        },
+        {
+            "loss": 14.1039,
+            "grad_norm": 2.041637420654297,
+            "learning_rate": 0.0005,
+            "epoch": 0.23327861780738268,
+            "step": 5205
+        },
+        {
+            "loss": 14.3222,
+            "grad_norm": 1.9130808115005493,
+            "learning_rate": 0.0005,
+            "epoch": 0.23350270869864817,
+            "step": 5210
+        },
+        {
+            "loss": 14.1975,
+            "grad_norm": 1.881244421005249,
+            "learning_rate": 0.0005,
+            "epoch": 0.23372679958991366,
+            "step": 5215
+        },
+        {
+            "loss": 14.0401,
+            "grad_norm": 1.8800253868103027,
+            "learning_rate": 0.0005,
+            "epoch": 0.23395089048117917,
+            "step": 5220
+        },
+        {
+            "loss": 14.157,
+            "grad_norm": 2.026484489440918,
+            "learning_rate": 0.0005,
+            "epoch": 0.23417498137244466,
+            "step": 5225
+        },
+        {
+            "loss": 14.1267,
+            "grad_norm": 1.979641318321228,
+            "learning_rate": 0.0005,
+            "epoch": 0.23439907226371015,
+            "step": 5230
+        },
+        {
+            "loss": 14.1186,
+            "grad_norm": 1.9195035696029663,
+            "learning_rate": 0.0005,
+            "epoch": 0.23462316315497567,
+            "step": 5235
+        },
+        {
+            "loss": 14.2415,
+            "grad_norm": 1.9984934329986572,
+            "learning_rate": 0.0005,
+            "epoch": 0.23484725404624116,
+            "step": 5240
+        },
+        {
+            "loss": 14.2466,
+            "grad_norm": 1.9173070192337036,
+            "learning_rate": 0.0005,
+            "epoch": 0.23507134493750664,
+            "step": 5245
+        },
+        {
+            "loss": 14.2014,
+            "grad_norm": 2.020751714706421,
+            "learning_rate": 0.0005,
+            "epoch": 0.23529543582877216,
+            "step": 5250
+        },
+        {
+            "loss": 14.0758,
+            "grad_norm": 2.14182448387146,
+            "learning_rate": 0.0005,
+            "epoch": 0.23551952672003765,
+            "step": 5255
+        },
+        {
+            "loss": 14.0696,
+            "grad_norm": 2.078418731689453,
+            "learning_rate": 0.0005,
+            "epoch": 0.23574361761130314,
+            "step": 5260
+        },
+        {
+            "loss": 14.1406,
+            "grad_norm": 2.12013840675354,
+            "learning_rate": 0.0005,
+            "epoch": 0.23596770850256865,
+            "step": 5265
+        },
+        {
+            "loss": 14.0492,
+            "grad_norm": 1.7853872776031494,
+            "learning_rate": 0.0005,
+            "epoch": 0.23619179939383414,
+            "step": 5270
+        },
+        {
+            "loss": 14.0794,
+            "grad_norm": 1.743823766708374,
+            "learning_rate": 0.0005,
+            "epoch": 0.23641589028509963,
+            "step": 5275
+        },
+        {
+            "loss": 14.2286,
+            "grad_norm": 1.872301459312439,
+            "learning_rate": 0.0005,
+            "epoch": 0.23663998117636514,
+            "step": 5280
+        },
+        {
+            "loss": 14.2119,
+            "grad_norm": 2.0029842853546143,
+            "learning_rate": 0.0005,
+            "epoch": 0.23686407206763063,
+            "step": 5285
+        },
+        {
+            "loss": 14.1721,
+            "grad_norm": 2.0815510749816895,
+            "learning_rate": 0.0005,
+            "epoch": 0.23708816295889612,
+            "step": 5290
+        },
+        {
+            "loss": 14.1907,
+            "grad_norm": 1.9053215980529785,
+            "learning_rate": 0.0005,
+            "epoch": 0.23731225385016164,
+            "step": 5295
+        },
+        {
+            "loss": 14.1224,
+            "grad_norm": 1.9221818447113037,
+            "learning_rate": 0.0005,
+            "epoch": 0.23753634474142712,
+            "step": 5300
+        },
+        {
+            "loss": 14.1437,
+            "grad_norm": 2.031381607055664,
+            "learning_rate": 0.0005,
+            "epoch": 0.2377604356326926,
+            "step": 5305
+        },
+        {
+            "loss": 14.1592,
+            "grad_norm": 1.8390190601348877,
+            "learning_rate": 0.0005,
+            "epoch": 0.23798452652395813,
+            "step": 5310
+        },
+        {
+            "loss": 14.3036,
+            "grad_norm": 1.8330848217010498,
+            "learning_rate": 0.0005,
+            "epoch": 0.23820861741522362,
+            "step": 5315
+        },
+        {
+            "loss": 14.0815,
+            "grad_norm": 1.9881000518798828,
+            "learning_rate": 0.0005,
+            "epoch": 0.2384327083064891,
+            "step": 5320
+        },
+        {
+            "loss": 14.1096,
+            "grad_norm": 2.018603563308716,
+            "learning_rate": 0.0005,
+            "epoch": 0.23865679919775462,
+            "step": 5325
+        },
+        {
+            "loss": 14.2211,
+            "grad_norm": 1.9145399332046509,
+            "learning_rate": 0.0005,
+            "epoch": 0.2388808900890201,
+            "step": 5330
+        },
+        {
+            "loss": 14.2118,
+            "grad_norm": 1.8877010345458984,
+            "learning_rate": 0.0005,
+            "epoch": 0.2391049809802856,
+            "step": 5335
+        },
+        {
+            "loss": 14.1218,
+            "grad_norm": 1.8736507892608643,
+            "learning_rate": 0.0005,
+            "epoch": 0.2393290718715511,
+            "step": 5340
+        },
+        {
+            "loss": 14.2243,
+            "grad_norm": 1.9161192178726196,
+            "learning_rate": 0.0005,
+            "epoch": 0.2395531627628166,
+            "step": 5345
+        },
+        {
+            "loss": 14.0829,
+            "grad_norm": 2.1973557472229004,
+            "learning_rate": 0.0005,
+            "epoch": 0.2397772536540821,
+            "step": 5350
+        },
+        {
+            "loss": 14.1577,
+            "grad_norm": 1.886846661567688,
+            "learning_rate": 0.0005,
+            "epoch": 0.2400013445453476,
+            "step": 5355
+        },
+        {
+            "loss": 14.2135,
+            "grad_norm": 1.8616987466812134,
+            "learning_rate": 0.0005,
+            "epoch": 0.2402254354366131,
+            "step": 5360
+        },
+        {
+            "loss": 14.1345,
+            "grad_norm": 2.0593018531799316,
+            "learning_rate": 0.0005,
+            "epoch": 0.24044952632787858,
+            "step": 5365
+        },
+        {
+            "loss": 14.1975,
+            "grad_norm": 1.9629650115966797,
+            "learning_rate": 0.0005,
+            "epoch": 0.2406736172191441,
+            "step": 5370
+        },
+        {
+            "loss": 14.1527,
+            "grad_norm": 1.9321669340133667,
+            "learning_rate": 0.0005,
+            "epoch": 0.2408977081104096,
+            "step": 5375
+        },
+        {
+            "loss": 14.2791,
+            "grad_norm": 1.8482413291931152,
+            "learning_rate": 0.0005,
+            "epoch": 0.24112179900167507,
+            "step": 5380
+        },
+        {
+            "loss": 14.1751,
+            "grad_norm": 1.824684500694275,
+            "learning_rate": 0.0005,
+            "epoch": 0.2413458898929406,
+            "step": 5385
+        },
+        {
+            "loss": 14.1206,
+            "grad_norm": 1.8810324668884277,
+            "learning_rate": 0.0005,
+            "epoch": 0.24156998078420608,
+            "step": 5390
+        },
+        {
+            "loss": 14.1139,
+            "grad_norm": 2.046349048614502,
+            "learning_rate": 0.0005,
+            "epoch": 0.24179407167547157,
+            "step": 5395
+        },
+        {
+            "loss": 14.2159,
+            "grad_norm": 2.128713369369507,
+            "learning_rate": 0.0005,
+            "epoch": 0.24201816256673706,
+            "step": 5400
+        },
+        {
+            "loss": 14.2577,
+            "grad_norm": 2.089057445526123,
+            "learning_rate": 0.0005,
+            "epoch": 0.24224225345800257,
+            "step": 5405
+        },
+        {
+            "loss": 14.149,
+            "grad_norm": 1.968741536140442,
+            "learning_rate": 0.0005,
+            "epoch": 0.24246634434926806,
+            "step": 5410
+        },
+        {
+            "loss": 14.1253,
+            "grad_norm": 1.9236301183700562,
+            "learning_rate": 0.0005,
+            "epoch": 0.24269043524053355,
+            "step": 5415
+        },
+        {
+            "loss": 14.1906,
+            "grad_norm": 2.0584657192230225,
+            "learning_rate": 0.0005,
+            "epoch": 0.24291452613179906,
+            "step": 5420
+        },
+        {
+            "loss": 14.1517,
+            "grad_norm": 2.031520128250122,
+            "learning_rate": 0.0005,
+            "epoch": 0.24313861702306455,
+            "step": 5425
+        },
+        {
+            "loss": 14.1071,
+            "grad_norm": 2.004542827606201,
+            "learning_rate": 0.0005,
+            "epoch": 0.24336270791433004,
+            "step": 5430
+        },
+        {
+            "loss": 14.1991,
+            "grad_norm": 1.9510637521743774,
+            "learning_rate": 0.0005,
+            "epoch": 0.24358679880559556,
+            "step": 5435
+        },
+        {
+            "loss": 14.1279,
+            "grad_norm": 1.9296494722366333,
+            "learning_rate": 0.0005,
+            "epoch": 0.24381088969686104,
+            "step": 5440
+        },
+        {
+            "loss": 14.1353,
+            "grad_norm": 1.9912152290344238,
+            "learning_rate": 0.0005,
+            "epoch": 0.24403498058812653,
+            "step": 5445
+        },
+        {
+            "loss": 14.0887,
+            "grad_norm": 1.7792458534240723,
+            "learning_rate": 0.0005,
+            "epoch": 0.24425907147939205,
+            "step": 5450
+        },
+        {
+            "loss": 14.1501,
+            "grad_norm": 1.7846481800079346,
+            "learning_rate": 0.0005,
+            "epoch": 0.24448316237065754,
+            "step": 5455
+        },
+        {
+            "loss": 14.3272,
+            "grad_norm": 1.8782302141189575,
+            "learning_rate": 0.0005,
+            "epoch": 0.24470725326192302,
+            "step": 5460
+        },
+        {
+            "loss": 14.1075,
+            "grad_norm": 1.8768726587295532,
+            "learning_rate": 0.0005,
+            "epoch": 0.24493134415318854,
+            "step": 5465
+        },
+        {
+            "loss": 14.1498,
+            "grad_norm": 1.882405400276184,
+            "learning_rate": 0.0005,
+            "epoch": 0.24515543504445403,
+            "step": 5470
+        },
+        {
+            "loss": 14.0709,
+            "grad_norm": 2.045891046524048,
+            "learning_rate": 0.0005,
+            "epoch": 0.24537952593571952,
+            "step": 5475
+        },
+        {
+            "loss": 14.2524,
+            "grad_norm": 1.9826222658157349,
+            "learning_rate": 0.0005,
+            "epoch": 0.24560361682698503,
+            "step": 5480
+        },
+        {
+            "loss": 14.1392,
+            "grad_norm": 2.0543572902679443,
+            "learning_rate": 0.0005,
+            "epoch": 0.24582770771825052,
+            "step": 5485
+        },
+        {
+            "loss": 14.2088,
+            "grad_norm": 1.989733338356018,
+            "learning_rate": 0.0005,
+            "epoch": 0.246051798609516,
+            "step": 5490
+        },
+        {
+            "loss": 14.2357,
+            "grad_norm": 1.967455506324768,
+            "learning_rate": 0.0005,
+            "epoch": 0.24627588950078153,
+            "step": 5495
+        },
+        {
+            "loss": 14.2458,
+            "grad_norm": 2.254347085952759,
+            "learning_rate": 0.0005,
+            "epoch": 0.246499980392047,
+            "step": 5500
+        },
+        {
+            "eval_loss": 1.76084303855896,
+            "eval_runtime": 18.6083,
+            "eval_samples_per_second": 880.468,
+            "eval_steps_per_second": 7.9,
+            "epoch": 0.246499980392047,
+            "step": 5500
+        },
+        {
+            "loss": 14.0829,
+            "grad_norm": 2.1639139652252197,
+            "learning_rate": 0.0005,
+            "epoch": 0.2467240712833125,
+            "step": 5505
+        },
+        {
+            "loss": 14.1356,
+            "grad_norm": 1.9651415348052979,
+            "learning_rate": 0.0005,
+            "epoch": 0.24694816217457802,
+            "step": 5510
+        },
+        {
+            "loss": 14.2047,
+            "grad_norm": 1.8461552858352661,
+            "learning_rate": 0.0005,
+            "epoch": 0.2471722530658435,
+            "step": 5515
+        },
+        {
+            "loss": 14.2647,
+            "grad_norm": 1.8990920782089233,
+            "learning_rate": 0.0005,
+            "epoch": 0.247396343957109,
+            "step": 5520
+        },
+        {
+            "loss": 14.2771,
+            "grad_norm": 1.8398399353027344,
+            "learning_rate": 0.0005,
+            "epoch": 0.2476204348483745,
+            "step": 5525
+        },
+        {
+            "loss": 14.2495,
+            "grad_norm": 1.9693412780761719,
+            "learning_rate": 0.0005,
+            "epoch": 0.24784452573964,
+            "step": 5530
+        },
+        {
+            "loss": 14.2004,
+            "grad_norm": 2.0449378490448,
+            "learning_rate": 0.0005,
+            "epoch": 0.2480686166309055,
+            "step": 5535
+        },
+        {
+            "loss": 14.1347,
+            "grad_norm": 2.1678719520568848,
+            "learning_rate": 0.0005,
+            "epoch": 0.248292707522171,
+            "step": 5540
+        },
+        {
+            "loss": 14.0804,
+            "grad_norm": 1.9390939474105835,
+            "learning_rate": 0.0005,
+            "epoch": 0.2485167984134365,
+            "step": 5545
+        },
+        {
+            "loss": 14.1556,
+            "grad_norm": 2.0248236656188965,
+            "learning_rate": 0.0005,
+            "epoch": 0.24874088930470198,
+            "step": 5550
+        },
+        {
+            "loss": 14.118,
+            "grad_norm": 2.0306644439697266,
+            "learning_rate": 0.0005,
+            "epoch": 0.2489649801959675,
+            "step": 5555
+        },
+        {
+            "loss": 14.1897,
+            "grad_norm": 1.973796010017395,
+            "learning_rate": 0.0005,
+            "epoch": 0.24918907108723298,
+            "step": 5560
+        },
+        {
+            "loss": 14.1298,
+            "grad_norm": 2.283743381500244,
+            "learning_rate": 0.0005,
+            "epoch": 0.24941316197849847,
+            "step": 5565
+        },
+        {
+            "loss": 14.1013,
+            "grad_norm": 1.961428165435791,
+            "learning_rate": 0.0005,
+            "epoch": 0.249637252869764,
+            "step": 5570
+        },
+        {
+            "loss": 14.1541,
+            "grad_norm": 1.8481162786483765,
+            "learning_rate": 0.0005,
+            "epoch": 0.24986134376102948,
+            "step": 5575
+        },
+        {
+            "loss": 14.1204,
+            "grad_norm": 1.7814725637435913,
+            "learning_rate": 0.0005,
+            "epoch": 0.250085434652295,
+            "step": 5580
+        },
+        {
+            "loss": 14.1446,
+            "grad_norm": 1.8474041223526,
+            "learning_rate": 0.0005,
+            "epoch": 0.25030952554356045,
+            "step": 5585
+        },
+        {
+            "loss": 14.1353,
+            "grad_norm": 1.9465347528457642,
+            "learning_rate": 0.0005,
+            "epoch": 0.25053361643482597,
+            "step": 5590
+        },
+        {
+            "loss": 14.1041,
+            "grad_norm": 2.047680616378784,
+            "learning_rate": 0.0005,
+            "epoch": 0.2507577073260915,
+            "step": 5595
+        },
+        {
+            "loss": 14.2028,
+            "grad_norm": 1.9223048686981201,
+            "learning_rate": 0.0005,
+            "epoch": 0.25098179821735694,
+            "step": 5600
+        },
+        {
+            "loss": 14.17,
+            "grad_norm": 2.0348899364471436,
+            "learning_rate": 0.0005,
+            "epoch": 0.25120588910862246,
+            "step": 5605
+        },
+        {
+            "loss": 14.0867,
+            "grad_norm": 2.129124164581299,
+            "learning_rate": 0.0005,
+            "epoch": 0.251429979999888,
+            "step": 5610
+        },
+        {
+            "loss": 14.1084,
+            "grad_norm": 1.8007903099060059,
+            "learning_rate": 0.0005,
+            "epoch": 0.25165407089115344,
+            "step": 5615
+        },
+        {
+            "loss": 14.1026,
+            "grad_norm": 1.8596608638763428,
+            "learning_rate": 0.0005,
+            "epoch": 0.25187816178241895,
+            "step": 5620
+        },
+        {
+            "loss": 14.1257,
+            "grad_norm": 1.837965726852417,
+            "learning_rate": 0.0005,
+            "epoch": 0.25210225267368447,
+            "step": 5625
+        },
+        {
+            "loss": 14.0864,
+            "grad_norm": 2.020282506942749,
+            "learning_rate": 0.0005,
+            "epoch": 0.25232634356494993,
+            "step": 5630
+        },
+        {
+            "loss": 14.2164,
+            "grad_norm": 2.035090446472168,
+            "learning_rate": 0.0005,
+            "epoch": 0.25255043445621544,
+            "step": 5635
+        },
+        {
+            "loss": 14.1225,
+            "grad_norm": 1.9386943578720093,
+            "learning_rate": 0.0005,
+            "epoch": 0.25277452534748096,
+            "step": 5640
+        },
+        {
+            "loss": 14.2029,
+            "grad_norm": 1.9536880254745483,
+            "learning_rate": 0.0005,
+            "epoch": 0.2529986162387464,
+            "step": 5645
+        },
+        {
+            "loss": 14.1768,
+            "grad_norm": 1.7979294061660767,
+            "learning_rate": 0.0005,
+            "epoch": 0.25322270713001194,
+            "step": 5650
+        },
+        {
+            "loss": 14.1672,
+            "grad_norm": 2.0136592388153076,
+            "learning_rate": 0.0005,
+            "epoch": 0.25344679802127745,
+            "step": 5655
+        },
+        {
+            "loss": 14.0883,
+            "grad_norm": 1.9120242595672607,
+            "learning_rate": 0.0005,
+            "epoch": 0.2536708889125429,
+            "step": 5660
+        },
+        {
+            "loss": 14.1071,
+            "grad_norm": 1.9633128643035889,
+            "learning_rate": 0.0005,
+            "epoch": 0.25389497980380843,
+            "step": 5665
+        },
+        {
+            "loss": 14.1785,
+            "grad_norm": 2.052507162094116,
+            "learning_rate": 0.0005,
+            "epoch": 0.25411907069507395,
+            "step": 5670
+        },
+        {
+            "loss": 14.1423,
+            "grad_norm": 1.8795636892318726,
+            "learning_rate": 0.0005,
+            "epoch": 0.2543431615863394,
+            "step": 5675
+        },
+        {
+            "loss": 14.1653,
+            "grad_norm": 1.8552874326705933,
+            "learning_rate": 0.0005,
+            "epoch": 0.2545672524776049,
+            "step": 5680
+        },
+        {
+            "loss": 14.0681,
+            "grad_norm": 1.9075721502304077,
+            "learning_rate": 0.0005,
+            "epoch": 0.25479134336887044,
+            "step": 5685
+        },
+        {
+            "loss": 14.2344,
+            "grad_norm": 2.0484118461608887,
+            "learning_rate": 0.0005,
+            "epoch": 0.2550154342601359,
+            "step": 5690
+        },
+        {
+            "loss": 14.177,
+            "grad_norm": 1.8938542604446411,
+            "learning_rate": 0.0005,
+            "epoch": 0.2552395251514014,
+            "step": 5695
+        },
+        {
+            "loss": 14.1684,
+            "grad_norm": 1.949268102645874,
+            "learning_rate": 0.0005,
+            "epoch": 0.25546361604266693,
+            "step": 5700
+        },
+        {
+            "loss": 14.0736,
+            "grad_norm": 1.8290852308273315,
+            "learning_rate": 0.0005,
+            "epoch": 0.2556877069339324,
+            "step": 5705
+        },
+        {
+            "loss": 14.1068,
+            "grad_norm": 1.9694098234176636,
+            "learning_rate": 0.0005,
+            "epoch": 0.2559117978251979,
+            "step": 5710
+        },
+        {
+            "loss": 14.1696,
+            "grad_norm": 2.0308942794799805,
+            "learning_rate": 0.0005,
+            "epoch": 0.2561358887164634,
+            "step": 5715
+        },
+        {
+            "loss": 14.1764,
+            "grad_norm": 1.825493335723877,
+            "learning_rate": 0.0005,
+            "epoch": 0.2563599796077289,
+            "step": 5720
+        },
+        {
+            "loss": 14.0856,
+            "grad_norm": 1.9709550142288208,
+            "learning_rate": 0.0005,
+            "epoch": 0.2565840704989944,
+            "step": 5725
+        },
+        {
+            "loss": 14.18,
+            "grad_norm": 2.001124620437622,
+            "learning_rate": 0.0005,
+            "epoch": 0.2568081613902599,
+            "step": 5730
+        },
+        {
+            "loss": 14.2103,
+            "grad_norm": 2.108546733856201,
+            "learning_rate": 0.0005,
+            "epoch": 0.2570322522815254,
+            "step": 5735
+        },
+        {
+            "loss": 14.0931,
+            "grad_norm": 2.1841468811035156,
+            "learning_rate": 0.0005,
+            "epoch": 0.2572563431727909,
+            "step": 5740
+        },
+        {
+            "loss": 14.1396,
+            "grad_norm": 2.1639816761016846,
+            "learning_rate": 0.0005,
+            "epoch": 0.2574804340640564,
+            "step": 5745
+        },
+        {
+            "loss": 14.1319,
+            "grad_norm": 2.002530336380005,
+            "learning_rate": 0.0005,
+            "epoch": 0.25770452495532187,
+            "step": 5750
+        },
+        {
+            "loss": 14.1058,
+            "grad_norm": 2.0346779823303223,
+            "learning_rate": 0.0005,
+            "epoch": 0.2579286158465874,
+            "step": 5755
+        },
+        {
+            "loss": 14.1493,
+            "grad_norm": 2.0625033378601074,
+            "learning_rate": 0.0005,
+            "epoch": 0.2581527067378529,
+            "step": 5760
+        },
+        {
+            "loss": 14.1357,
+            "grad_norm": 1.9248889684677124,
+            "learning_rate": 0.0005,
+            "epoch": 0.25837679762911836,
+            "step": 5765
+        },
+        {
+            "loss": 14.098,
+            "grad_norm": 2.009620428085327,
+            "learning_rate": 0.0005,
+            "epoch": 0.2586008885203839,
+            "step": 5770
+        },
+        {
+            "loss": 14.1985,
+            "grad_norm": 2.120296001434326,
+            "learning_rate": 0.0005,
+            "epoch": 0.2588249794116494,
+            "step": 5775
+        },
+        {
+            "loss": 14.1445,
+            "grad_norm": 2.1229355335235596,
+            "learning_rate": 0.0005,
+            "epoch": 0.25904907030291485,
+            "step": 5780
+        },
+        {
+            "loss": 14.1839,
+            "grad_norm": 2.3002853393554688,
+            "learning_rate": 0.0005,
+            "epoch": 0.25927316119418037,
+            "step": 5785
+        },
+        {
+            "loss": 14.0847,
+            "grad_norm": 1.961533784866333,
+            "learning_rate": 0.0005,
+            "epoch": 0.2594972520854459,
+            "step": 5790
+        },
+        {
+            "loss": 14.127,
+            "grad_norm": 2.012505531311035,
+            "learning_rate": 0.0005,
+            "epoch": 0.25972134297671134,
+            "step": 5795
+        },
+        {
+            "loss": 14.1247,
+            "grad_norm": 2.105245590209961,
+            "learning_rate": 0.0005,
+            "epoch": 0.25994543386797686,
+            "step": 5800
+        },
+        {
+            "loss": 14.2372,
+            "grad_norm": 1.9416728019714355,
+            "learning_rate": 0.0005,
+            "epoch": 0.2601695247592423,
+            "step": 5805
+        },
+        {
+            "loss": 14.139,
+            "grad_norm": 2.037787437438965,
+            "learning_rate": 0.0005,
+            "epoch": 0.26039361565050784,
+            "step": 5810
+        },
+        {
+            "loss": 14.0952,
+            "grad_norm": 1.8043665885925293,
+            "learning_rate": 0.0005,
+            "epoch": 0.26061770654177335,
+            "step": 5815
+        },
+        {
+            "loss": 14.0973,
+            "grad_norm": 2.161391496658325,
+            "learning_rate": 0.0005,
+            "epoch": 0.2608417974330388,
+            "step": 5820
+        },
+        {
+            "loss": 14.096,
+            "grad_norm": 2.1336889266967773,
+            "learning_rate": 0.0005,
+            "epoch": 0.26106588832430433,
+            "step": 5825
+        },
+        {
+            "loss": 14.2098,
+            "grad_norm": 1.9870775938034058,
+            "learning_rate": 0.0005,
+            "epoch": 0.26128997921556985,
+            "step": 5830
+        },
+        {
+            "loss": 14.1767,
+            "grad_norm": 1.848604440689087,
+            "learning_rate": 0.0005,
+            "epoch": 0.2615140701068353,
+            "step": 5835
+        },
+        {
+            "loss": 14.1075,
+            "grad_norm": 1.8512866497039795,
+            "learning_rate": 0.0005,
+            "epoch": 0.2617381609981008,
+            "step": 5840
+        },
+        {
+            "loss": 14.2583,
+            "grad_norm": 2.112514019012451,
+            "learning_rate": 0.0005,
+            "epoch": 0.26196225188936634,
+            "step": 5845
+        },
+        {
+            "loss": 14.1369,
+            "grad_norm": 2.0032267570495605,
+            "learning_rate": 0.0005,
+            "epoch": 0.2621863427806318,
+            "step": 5850
+        },
+        {
+            "loss": 14.1011,
+            "grad_norm": 2.060760259628296,
+            "learning_rate": 0.0005,
+            "epoch": 0.2624104336718973,
+            "step": 5855
+        },
+        {
+            "loss": 14.1127,
+            "grad_norm": 1.7968379259109497,
+            "learning_rate": 0.0005,
+            "epoch": 0.26263452456316283,
+            "step": 5860
+        },
+        {
+            "loss": 14.2577,
+            "grad_norm": 1.912384271621704,
+            "learning_rate": 0.0005,
+            "epoch": 0.2628586154544283,
+            "step": 5865
+        },
+        {
+            "loss": 14.0495,
+            "grad_norm": 1.9313682317733765,
+            "learning_rate": 0.0005,
+            "epoch": 0.2630827063456938,
+            "step": 5870
+        },
+        {
+            "loss": 14.232,
+            "grad_norm": 1.9164438247680664,
+            "learning_rate": 0.0005,
+            "epoch": 0.2633067972369593,
+            "step": 5875
+        },
+        {
+            "loss": 14.159,
+            "grad_norm": 2.094454526901245,
+            "learning_rate": 0.0005,
+            "epoch": 0.2635308881282248,
+            "step": 5880
+        },
+        {
+            "loss": 14.0457,
+            "grad_norm": 1.9597326517105103,
+            "learning_rate": 0.0005,
+            "epoch": 0.2637549790194903,
+            "step": 5885
+        },
+        {
+            "loss": 14.0497,
+            "grad_norm": 1.7545689344406128,
+            "learning_rate": 0.0005,
+            "epoch": 0.2639790699107558,
+            "step": 5890
+        },
+        {
+            "loss": 14.1202,
+            "grad_norm": 1.9736093282699585,
+            "learning_rate": 0.0005,
+            "epoch": 0.2642031608020213,
+            "step": 5895
+        },
+        {
+            "loss": 14.1309,
+            "grad_norm": 2.03840970993042,
+            "learning_rate": 0.0005,
+            "epoch": 0.2644272516932868,
+            "step": 5900
+        },
+        {
+            "loss": 14.1956,
+            "grad_norm": 1.9478352069854736,
+            "learning_rate": 0.0005,
+            "epoch": 0.2646513425845523,
+            "step": 5905
+        },
+        {
+            "loss": 14.1178,
+            "grad_norm": 2.0999906063079834,
+            "learning_rate": 0.0005,
+            "epoch": 0.26487543347581777,
+            "step": 5910
+        },
+        {
+            "loss": 14.0856,
+            "grad_norm": 2.1087260246276855,
+            "learning_rate": 0.0005,
+            "epoch": 0.2650995243670833,
+            "step": 5915
+        },
+        {
+            "loss": 14.1338,
+            "grad_norm": 1.7892274856567383,
+            "learning_rate": 0.0005,
+            "epoch": 0.2653236152583488,
+            "step": 5920
+        },
+        {
+            "loss": 14.1355,
+            "grad_norm": 2.0246315002441406,
+            "learning_rate": 0.0005,
+            "epoch": 0.26554770614961426,
+            "step": 5925
+        },
+        {
+            "loss": 14.1132,
+            "grad_norm": 1.98887038230896,
+            "learning_rate": 0.0005,
+            "epoch": 0.2657717970408798,
+            "step": 5930
+        },
+        {
+            "loss": 14.2338,
+            "grad_norm": 2.2294838428497314,
+            "learning_rate": 0.0005,
+            "epoch": 0.2659958879321453,
+            "step": 5935
+        },
+        {
+            "loss": 14.1011,
+            "grad_norm": 1.8237981796264648,
+            "learning_rate": 0.0005,
+            "epoch": 0.26621997882341075,
+            "step": 5940
+        },
+        {
+            "loss": 14.0735,
+            "grad_norm": 2.0049431324005127,
+            "learning_rate": 0.0005,
+            "epoch": 0.26644406971467627,
+            "step": 5945
+        },
+        {
+            "loss": 14.0917,
+            "grad_norm": 2.1553215980529785,
+            "learning_rate": 0.0005,
+            "epoch": 0.2666681606059418,
+            "step": 5950
+        },
+        {
+            "loss": 14.0225,
+            "grad_norm": 2.1619985103607178,
+            "learning_rate": 0.0005,
+            "epoch": 0.26689225149720724,
+            "step": 5955
+        },
+        {
+            "loss": 14.1076,
+            "grad_norm": 1.9701801538467407,
+            "learning_rate": 0.0005,
+            "epoch": 0.26711634238847276,
+            "step": 5960
+        },
+        {
+            "loss": 14.0973,
+            "grad_norm": 2.117084264755249,
+            "learning_rate": 0.0005,
+            "epoch": 0.2673404332797383,
+            "step": 5965
+        },
+        {
+            "loss": 14.1038,
+            "grad_norm": 1.9590649604797363,
+            "learning_rate": 0.0005,
+            "epoch": 0.26756452417100374,
+            "step": 5970
+        },
+        {
+            "loss": 14.1344,
+            "grad_norm": 1.9777534008026123,
+            "learning_rate": 0.0005,
+            "epoch": 0.26778861506226925,
+            "step": 5975
+        },
+        {
+            "loss": 14.0876,
+            "grad_norm": 2.003319025039673,
+            "learning_rate": 0.0005,
+            "epoch": 0.26801270595353477,
+            "step": 5980
+        },
+        {
+            "loss": 14.1786,
+            "grad_norm": 1.934256911277771,
+            "learning_rate": 0.0005,
+            "epoch": 0.26823679684480023,
+            "step": 5985
+        },
+        {
+            "loss": 14.1779,
+            "grad_norm": 1.8747920989990234,
+            "learning_rate": 0.0005,
+            "epoch": 0.26846088773606575,
+            "step": 5990
+        },
+        {
+            "loss": 13.9991,
+            "grad_norm": 2.0487060546875,
+            "learning_rate": 0.0005,
+            "epoch": 0.26868497862733126,
+            "step": 5995
+        },
+        {
+            "loss": 14.0573,
+            "grad_norm": 2.062567949295044,
+            "learning_rate": 0.0005,
+            "epoch": 0.2689090695185967,
+            "step": 6000
+        },
+        {
+            "eval_loss": 1.7624342441558838,
+            "eval_runtime": 18.6015,
+            "eval_samples_per_second": 880.791,
+            "eval_steps_per_second": 7.903,
+            "epoch": 0.2689090695185967,
+            "step": 6000
+        },
+        {
+            "loss": 14.0626,
+            "grad_norm": 1.9452993869781494,
+            "learning_rate": 0.0005,
+            "epoch": 0.26913316040986224,
+            "step": 6005
+        },
+        {
+            "loss": 14.1218,
+            "grad_norm": 1.8791625499725342,
+            "learning_rate": 0.0005,
+            "epoch": 0.26935725130112775,
+            "step": 6010
+        },
+        {
+            "loss": 14.042,
+            "grad_norm": 1.9103970527648926,
+            "learning_rate": 0.0005,
+            "epoch": 0.2695813421923932,
+            "step": 6015
+        },
+        {
+            "loss": 14.1955,
+            "grad_norm": 1.7610692977905273,
+            "learning_rate": 0.0005,
+            "epoch": 0.26980543308365873,
+            "step": 6020
+        },
+        {
+            "loss": 14.2647,
+            "grad_norm": 1.9507147073745728,
+            "learning_rate": 0.0005,
+            "epoch": 0.27002952397492425,
+            "step": 6025
+        },
+        {
+            "loss": 14.1833,
+            "grad_norm": 1.9560884237289429,
+            "learning_rate": 0.0005,
+            "epoch": 0.2702536148661897,
+            "step": 6030
+        },
+        {
+            "loss": 14.2063,
+            "grad_norm": 1.982217788696289,
+            "learning_rate": 0.0005,
+            "epoch": 0.2704777057574552,
+            "step": 6035
+        },
+        {
+            "loss": 14.2817,
+            "grad_norm": 1.8248435258865356,
+            "learning_rate": 0.0005,
+            "epoch": 0.27070179664872074,
+            "step": 6040
+        },
+        {
+            "loss": 14.0953,
+            "grad_norm": 2.075336456298828,
+            "learning_rate": 0.0005,
+            "epoch": 0.2709258875399862,
+            "step": 6045
+        },
+        {
+            "loss": 14.1181,
+            "grad_norm": 1.8482534885406494,
+            "learning_rate": 0.0005,
+            "epoch": 0.2711499784312517,
+            "step": 6050
+        },
+        {
+            "loss": 14.1048,
+            "grad_norm": 1.7765411138534546,
+            "learning_rate": 0.0005,
+            "epoch": 0.27137406932251723,
+            "step": 6055
+        },
+        {
+            "loss": 14.1664,
+            "grad_norm": 2.0339853763580322,
+            "learning_rate": 0.0005,
+            "epoch": 0.2715981602137827,
+            "step": 6060
+        },
+        {
+            "loss": 14.0792,
+            "grad_norm": 2.048095464706421,
+            "learning_rate": 0.0005,
+            "epoch": 0.2718222511050482,
+            "step": 6065
+        },
+        {
+            "loss": 14.1056,
+            "grad_norm": 1.952912449836731,
+            "learning_rate": 0.0005,
+            "epoch": 0.2720463419963137,
+            "step": 6070
+        },
+        {
+            "loss": 14.2421,
+            "grad_norm": 2.003805160522461,
+            "learning_rate": 0.0005,
+            "epoch": 0.2722704328875792,
+            "step": 6075
+        },
+        {
+            "loss": 14.1109,
+            "grad_norm": 1.9438632726669312,
+            "learning_rate": 0.0005,
+            "epoch": 0.2724945237788447,
+            "step": 6080
+        },
+        {
+            "loss": 14.1035,
+            "grad_norm": 2.0806822776794434,
+            "learning_rate": 0.0005,
+            "epoch": 0.2727186146701102,
+            "step": 6085
+        },
+        {
+            "loss": 14.1132,
+            "grad_norm": 1.8243623971939087,
+            "learning_rate": 0.0005,
+            "epoch": 0.2729427055613757,
+            "step": 6090
+        },
+        {
+            "loss": 14.0947,
+            "grad_norm": 2.221346139907837,
+            "learning_rate": 0.0005,
+            "epoch": 0.2731667964526412,
+            "step": 6095
+        },
+        {
+            "loss": 14.1149,
+            "grad_norm": 1.9295768737792969,
+            "learning_rate": 0.0005,
+            "epoch": 0.2733908873439067,
+            "step": 6100
+        },
+        {
+            "loss": 14.0679,
+            "grad_norm": 1.9353907108306885,
+            "learning_rate": 0.0005,
+            "epoch": 0.27361497823517217,
+            "step": 6105
+        },
+        {
+            "loss": 14.1163,
+            "grad_norm": 1.8101950883865356,
+            "learning_rate": 0.0005,
+            "epoch": 0.2738390691264377,
+            "step": 6110
+        },
+        {
+            "loss": 14.184,
+            "grad_norm": 1.8892567157745361,
+            "learning_rate": 0.0005,
+            "epoch": 0.2740631600177032,
+            "step": 6115
+        },
+        {
+            "loss": 14.191,
+            "grad_norm": 1.8542805910110474,
+            "learning_rate": 0.0005,
+            "epoch": 0.27428725090896866,
+            "step": 6120
+        },
+        {
+            "loss": 14.1419,
+            "grad_norm": 1.95559823513031,
+            "learning_rate": 0.0005,
+            "epoch": 0.2745113418002342,
+            "step": 6125
+        },
+        {
+            "loss": 14.1413,
+            "grad_norm": 2.100402355194092,
+            "learning_rate": 0.0005,
+            "epoch": 0.2747354326914997,
+            "step": 6130
+        },
+        {
+            "loss": 14.1712,
+            "grad_norm": 1.8818026781082153,
+            "learning_rate": 0.0005,
+            "epoch": 0.27495952358276515,
+            "step": 6135
+        },
+        {
+            "loss": 14.1133,
+            "grad_norm": 2.1112377643585205,
+            "learning_rate": 0.0005,
+            "epoch": 0.27518361447403067,
+            "step": 6140
+        },
+        {
+            "loss": 14.1174,
+            "grad_norm": 1.9072048664093018,
+            "learning_rate": 0.0005,
+            "epoch": 0.2754077053652962,
+            "step": 6145
+        },
+        {
+            "loss": 14.16,
+            "grad_norm": 1.8370167016983032,
+            "learning_rate": 0.0005,
+            "epoch": 0.27563179625656165,
+            "step": 6150
+        },
+        {
+            "loss": 14.1799,
+            "grad_norm": 1.7252795696258545,
+            "learning_rate": 0.0005,
+            "epoch": 0.27585588714782716,
+            "step": 6155
+        },
+        {
+            "loss": 14.152,
+            "grad_norm": 1.9398436546325684,
+            "learning_rate": 0.0005,
+            "epoch": 0.2760799780390927,
+            "step": 6160
+        },
+        {
+            "loss": 14.0982,
+            "grad_norm": 1.8849056959152222,
+            "learning_rate": 0.0005,
+            "epoch": 0.27630406893035814,
+            "step": 6165
+        },
+        {
+            "loss": 14.2026,
+            "grad_norm": 2.120668888092041,
+            "learning_rate": 0.0005,
+            "epoch": 0.27652815982162365,
+            "step": 6170
+        },
+        {
+            "loss": 14.1554,
+            "grad_norm": 2.057431221008301,
+            "learning_rate": 0.0005,
+            "epoch": 0.27675225071288917,
+            "step": 6175
+        },
+        {
+            "loss": 14.0881,
+            "grad_norm": 1.9021568298339844,
+            "learning_rate": 0.0005,
+            "epoch": 0.27697634160415463,
+            "step": 6180
+        },
+        {
+            "loss": 14.1402,
+            "grad_norm": 1.9208106994628906,
+            "learning_rate": 0.0005,
+            "epoch": 0.27720043249542015,
+            "step": 6185
+        },
+        {
+            "loss": 14.1664,
+            "grad_norm": 2.031012773513794,
+            "learning_rate": 0.0005,
+            "epoch": 0.27742452338668566,
+            "step": 6190
+        },
+        {
+            "loss": 14.0986,
+            "grad_norm": 1.9070667028427124,
+            "learning_rate": 0.0005,
+            "epoch": 0.2776486142779511,
+            "step": 6195
+        },
+        {
+            "loss": 14.1555,
+            "grad_norm": 2.1638131141662598,
+            "learning_rate": 0.0005,
+            "epoch": 0.27787270516921664,
+            "step": 6200
+        },
+        {
+            "loss": 14.0978,
+            "grad_norm": 1.982803225517273,
+            "learning_rate": 0.0005,
+            "epoch": 0.27809679606048215,
+            "step": 6205
+        },
+        {
+            "loss": 14.1707,
+            "grad_norm": 1.9024537801742554,
+            "learning_rate": 0.0005,
+            "epoch": 0.2783208869517476,
+            "step": 6210
+        },
+        {
+            "loss": 14.1495,
+            "grad_norm": 1.9770755767822266,
+            "learning_rate": 0.0005,
+            "epoch": 0.27854497784301313,
+            "step": 6215
+        },
+        {
+            "loss": 14.0563,
+            "grad_norm": 1.8895354270935059,
+            "learning_rate": 0.0005,
+            "epoch": 0.27876906873427865,
+            "step": 6220
+        },
+        {
+            "loss": 13.9898,
+            "grad_norm": 1.9661836624145508,
+            "learning_rate": 0.0005,
+            "epoch": 0.2789931596255441,
+            "step": 6225
+        },
+        {
+            "loss": 14.1714,
+            "grad_norm": 2.0561423301696777,
+            "learning_rate": 0.0005,
+            "epoch": 0.2792172505168096,
+            "step": 6230
+        },
+        {
+            "loss": 14.2114,
+            "grad_norm": 2.0155813694000244,
+            "learning_rate": 0.0005,
+            "epoch": 0.27944134140807514,
+            "step": 6235
+        },
+        {
+            "loss": 14.1845,
+            "grad_norm": 1.8753061294555664,
+            "learning_rate": 0.0005,
+            "epoch": 0.2796654322993406,
+            "step": 6240
+        },
+        {
+            "loss": 14.2542,
+            "grad_norm": 1.7123141288757324,
+            "learning_rate": 0.0005,
+            "epoch": 0.2798895231906061,
+            "step": 6245
+        },
+        {
+            "loss": 14.1507,
+            "grad_norm": 1.9749904870986938,
+            "learning_rate": 0.0005,
+            "epoch": 0.28011361408187163,
+            "step": 6250
+        },
+        {
+            "loss": 14.1013,
+            "grad_norm": 1.9573017358779907,
+            "learning_rate": 0.0005,
+            "epoch": 0.2803377049731371,
+            "step": 6255
+        },
+        {
+            "loss": 14.107,
+            "grad_norm": 1.9311197996139526,
+            "learning_rate": 0.0005,
+            "epoch": 0.2805617958644026,
+            "step": 6260
+        },
+        {
+            "loss": 14.1186,
+            "grad_norm": 1.9432008266448975,
+            "learning_rate": 0.0005,
+            "epoch": 0.2807858867556681,
+            "step": 6265
+        },
+        {
+            "loss": 14.1111,
+            "grad_norm": 2.0310680866241455,
+            "learning_rate": 0.0005,
+            "epoch": 0.2810099776469336,
+            "step": 6270
+        },
+        {
+            "loss": 14.0357,
+            "grad_norm": 1.7939568758010864,
+            "learning_rate": 0.0005,
+            "epoch": 0.2812340685381991,
+            "step": 6275
+        },
+        {
+            "loss": 14.1525,
+            "grad_norm": 1.9812079668045044,
+            "learning_rate": 0.0005,
+            "epoch": 0.2814581594294646,
+            "step": 6280
+        },
+        {
+            "loss": 14.0599,
+            "grad_norm": 1.7790424823760986,
+            "learning_rate": 0.0005,
+            "epoch": 0.2816822503207301,
+            "step": 6285
+        },
+        {
+            "loss": 14.1505,
+            "grad_norm": 1.9458509683609009,
+            "learning_rate": 0.0005,
+            "epoch": 0.2819063412119956,
+            "step": 6290
+        },
+        {
+            "loss": 14.1274,
+            "grad_norm": 1.8281060457229614,
+            "learning_rate": 0.0005,
+            "epoch": 0.2821304321032611,
+            "step": 6295
+        },
+        {
+            "loss": 14.0824,
+            "grad_norm": 2.062748908996582,
+            "learning_rate": 0.0005,
+            "epoch": 0.28235452299452657,
+            "step": 6300
+        },
+        {
+            "loss": 14.1564,
+            "grad_norm": 2.144453763961792,
+            "learning_rate": 0.0005,
+            "epoch": 0.2825786138857921,
+            "step": 6305
+        },
+        {
+            "loss": 14.2729,
+            "grad_norm": 2.056806802749634,
+            "learning_rate": 0.0005,
+            "epoch": 0.2828027047770576,
+            "step": 6310
+        },
+        {
+            "loss": 14.0951,
+            "grad_norm": 1.949935793876648,
+            "learning_rate": 0.0005,
+            "epoch": 0.28302679566832306,
+            "step": 6315
+        },
+        {
+            "loss": 14.2044,
+            "grad_norm": 1.8974496126174927,
+            "learning_rate": 0.0005,
+            "epoch": 0.2832508865595886,
+            "step": 6320
+        },
+        {
+            "loss": 14.0844,
+            "grad_norm": 1.9938383102416992,
+            "learning_rate": 0.0005,
+            "epoch": 0.2834749774508541,
+            "step": 6325
+        },
+        {
+            "loss": 14.0717,
+            "grad_norm": 2.000389575958252,
+            "learning_rate": 0.0005,
+            "epoch": 0.28369906834211955,
+            "step": 6330
+        },
+        {
+            "loss": 14.1252,
+            "grad_norm": 1.9282065629959106,
+            "learning_rate": 0.0005,
+            "epoch": 0.28392315923338507,
+            "step": 6335
+        },
+        {
+            "loss": 14.1557,
+            "grad_norm": 1.9459840059280396,
+            "learning_rate": 0.0005,
+            "epoch": 0.2841472501246506,
+            "step": 6340
+        },
+        {
+            "loss": 14.1244,
+            "grad_norm": 2.0548715591430664,
+            "learning_rate": 0.0005,
+            "epoch": 0.28437134101591605,
+            "step": 6345
+        },
+        {
+            "loss": 14.2648,
+            "grad_norm": 1.9488264322280884,
+            "learning_rate": 0.0005,
+            "epoch": 0.28459543190718156,
+            "step": 6350
+        },
+        {
+            "loss": 14.2416,
+            "grad_norm": 1.9849375486373901,
+            "learning_rate": 0.0005,
+            "epoch": 0.2848195227984471,
+            "step": 6355
+        },
+        {
+            "loss": 14.0532,
+            "grad_norm": 2.030043840408325,
+            "learning_rate": 0.0005,
+            "epoch": 0.28504361368971254,
+            "step": 6360
+        },
+        {
+            "loss": 14.1358,
+            "grad_norm": 1.9630738496780396,
+            "learning_rate": 0.0005,
+            "epoch": 0.28526770458097805,
+            "step": 6365
+        },
+        {
+            "loss": 14.1398,
+            "grad_norm": 1.936179757118225,
+            "learning_rate": 0.0005,
+            "epoch": 0.2854917954722435,
+            "step": 6370
+        },
+        {
+            "loss": 14.147,
+            "grad_norm": 1.8546801805496216,
+            "learning_rate": 0.0005,
+            "epoch": 0.28571588636350903,
+            "step": 6375
+        },
+        {
+            "loss": 14.1357,
+            "grad_norm": 1.7892138957977295,
+            "learning_rate": 0.0005,
+            "epoch": 0.28593997725477455,
+            "step": 6380
+        },
+        {
+            "loss": 14.1072,
+            "grad_norm": 2.2381319999694824,
+            "learning_rate": 0.0005,
+            "epoch": 0.28616406814604,
+            "step": 6385
+        },
+        {
+            "loss": 14.2096,
+            "grad_norm": 1.8750665187835693,
+            "learning_rate": 0.0005,
+            "epoch": 0.2863881590373055,
+            "step": 6390
+        },
+        {
+            "loss": 14.1466,
+            "grad_norm": 1.993774652481079,
+            "learning_rate": 0.0005,
+            "epoch": 0.28661224992857104,
+            "step": 6395
+        },
+        {
+            "loss": 14.1791,
+            "grad_norm": 1.9932512044906616,
+            "learning_rate": 0.0005,
+            "epoch": 0.2868363408198365,
+            "step": 6400
+        },
+        {
+            "loss": 13.9481,
+            "grad_norm": 1.9475047588348389,
+            "learning_rate": 0.0005,
+            "epoch": 0.287060431711102,
+            "step": 6405
+        },
+        {
+            "loss": 14.1085,
+            "grad_norm": 1.8606120347976685,
+            "learning_rate": 0.0005,
+            "epoch": 0.28728452260236753,
+            "step": 6410
+        },
+        {
+            "loss": 14.0986,
+            "grad_norm": 1.98119056224823,
+            "learning_rate": 0.0005,
+            "epoch": 0.287508613493633,
+            "step": 6415
+        },
+        {
+            "loss": 14.0864,
+            "grad_norm": 1.93229341506958,
+            "learning_rate": 0.0005,
+            "epoch": 0.2877327043848985,
+            "step": 6420
+        },
+        {
+            "loss": 14.1505,
+            "grad_norm": 2.0001397132873535,
+            "learning_rate": 0.0005,
+            "epoch": 0.287956795276164,
+            "step": 6425
+        },
+        {
+            "loss": 14.1825,
+            "grad_norm": 2.1840929985046387,
+            "learning_rate": 0.0005,
+            "epoch": 0.2881808861674295,
+            "step": 6430
+        },
+        {
+            "loss": 14.0742,
+            "grad_norm": 1.9578980207443237,
+            "learning_rate": 0.0005,
+            "epoch": 0.288404977058695,
+            "step": 6435
+        },
+        {
+            "loss": 14.0583,
+            "grad_norm": 1.9192719459533691,
+            "learning_rate": 0.0005,
+            "epoch": 0.2886290679499605,
+            "step": 6440
+        },
+        {
+            "loss": 14.1828,
+            "grad_norm": 1.7977638244628906,
+            "learning_rate": 0.0005,
+            "epoch": 0.288853158841226,
+            "step": 6445
+        },
+        {
+            "loss": 14.1839,
+            "grad_norm": 1.9880512952804565,
+            "learning_rate": 0.0005,
+            "epoch": 0.2890772497324915,
+            "step": 6450
+        },
+        {
+            "loss": 14.1521,
+            "grad_norm": 1.9016677141189575,
+            "learning_rate": 0.0005,
+            "epoch": 0.289301340623757,
+            "step": 6455
+        },
+        {
+            "loss": 14.0941,
+            "grad_norm": 2.009845018386841,
+            "learning_rate": 0.0005,
+            "epoch": 0.28952543151502247,
+            "step": 6460
+        },
+        {
+            "loss": 14.0867,
+            "grad_norm": 1.9944508075714111,
+            "learning_rate": 0.0005,
+            "epoch": 0.289749522406288,
+            "step": 6465
+        },
+        {
+            "loss": 14.12,
+            "grad_norm": 1.9572532176971436,
+            "learning_rate": 0.0005,
+            "epoch": 0.2899736132975535,
+            "step": 6470
+        },
+        {
+            "loss": 14.1033,
+            "grad_norm": 1.8286199569702148,
+            "learning_rate": 0.0005,
+            "epoch": 0.29019770418881896,
+            "step": 6475
+        },
+        {
+            "loss": 14.238,
+            "grad_norm": 1.8490030765533447,
+            "learning_rate": 0.0005,
+            "epoch": 0.2904217950800845,
+            "step": 6480
+        },
+        {
+            "loss": 14.0584,
+            "grad_norm": 1.883685827255249,
+            "learning_rate": 0.0005,
+            "epoch": 0.29064588597135,
+            "step": 6485
+        },
+        {
+            "loss": 14.1139,
+            "grad_norm": 1.8208873271942139,
+            "learning_rate": 0.0005,
+            "epoch": 0.29086997686261545,
+            "step": 6490
+        },
+        {
+            "loss": 14.1691,
+            "grad_norm": 1.79149329662323,
+            "learning_rate": 0.0005,
+            "epoch": 0.29109406775388097,
+            "step": 6495
+        },
+        {
+            "loss": 14.131,
+            "grad_norm": 1.8157325983047485,
+            "learning_rate": 0.0005,
+            "epoch": 0.2913181586451465,
+            "step": 6500
+        },
+        {
+            "eval_loss": 1.760825514793396,
+            "eval_runtime": 18.5425,
+            "eval_samples_per_second": 883.592,
+            "eval_steps_per_second": 7.928,
+            "epoch": 0.2913181586451465,
+            "step": 6500
+        },
+        {
+            "loss": 14.1269,
+            "grad_norm": 1.978583812713623,
+            "learning_rate": 0.0005,
+            "epoch": 0.29154224953641195,
+            "step": 6505
+        },
+        {
+            "loss": 14.1458,
+            "grad_norm": 1.9699724912643433,
+            "learning_rate": 0.0005,
+            "epoch": 0.29176634042767746,
+            "step": 6510
+        },
+        {
+            "loss": 14.127,
+            "grad_norm": 1.8420236110687256,
+            "learning_rate": 0.0005,
+            "epoch": 0.291990431318943,
+            "step": 6515
+        },
+        {
+            "loss": 14.1576,
+            "grad_norm": 2.1266531944274902,
+            "learning_rate": 0.0005,
+            "epoch": 0.29221452221020844,
+            "step": 6520
+        },
+        {
+            "loss": 14.057,
+            "grad_norm": 1.9557487964630127,
+            "learning_rate": 0.0005,
+            "epoch": 0.29243861310147395,
+            "step": 6525
+        },
+        {
+            "loss": 14.0618,
+            "grad_norm": 2.1013238430023193,
+            "learning_rate": 0.0005,
+            "epoch": 0.29266270399273947,
+            "step": 6530
+        },
+        {
+            "loss": 14.1422,
+            "grad_norm": 1.920631766319275,
+            "learning_rate": 0.0005,
+            "epoch": 0.29288679488400493,
+            "step": 6535
+        },
+        {
+            "loss": 14.0785,
+            "grad_norm": 1.8109145164489746,
+            "learning_rate": 0.0005,
+            "epoch": 0.29311088577527045,
+            "step": 6540
+        },
+        {
+            "loss": 14.036,
+            "grad_norm": 2.0602829456329346,
+            "learning_rate": 0.0005,
+            "epoch": 0.29333497666653596,
+            "step": 6545
+        },
+        {
+            "loss": 14.0582,
+            "grad_norm": 2.0971171855926514,
+            "learning_rate": 0.0005,
+            "epoch": 0.2935590675578014,
+            "step": 6550
+        },
+        {
+            "loss": 14.1295,
+            "grad_norm": 1.8590223789215088,
+            "learning_rate": 0.0005,
+            "epoch": 0.29378315844906694,
+            "step": 6555
+        },
+        {
+            "loss": 14.0767,
+            "grad_norm": 1.782094120979309,
+            "learning_rate": 0.0005,
+            "epoch": 0.29400724934033245,
+            "step": 6560
+        },
+        {
+            "loss": 14.1499,
+            "grad_norm": 2.0229482650756836,
+            "learning_rate": 0.0005,
+            "epoch": 0.2942313402315979,
+            "step": 6565
+        },
+        {
+            "loss": 14.0854,
+            "grad_norm": 2.0149924755096436,
+            "learning_rate": 0.0005,
+            "epoch": 0.29445543112286343,
+            "step": 6570
+        },
+        {
+            "loss": 14.1079,
+            "grad_norm": 1.915675401687622,
+            "learning_rate": 0.0005,
+            "epoch": 0.29467952201412895,
+            "step": 6575
+        },
+        {
+            "loss": 14.088,
+            "grad_norm": 1.950971245765686,
+            "learning_rate": 0.0005,
+            "epoch": 0.2949036129053944,
+            "step": 6580
+        },
+        {
+            "loss": 14.1089,
+            "grad_norm": 1.8409794569015503,
+            "learning_rate": 0.0005,
+            "epoch": 0.2951277037966599,
+            "step": 6585
+        },
+        {
+            "loss": 14.1737,
+            "grad_norm": 1.8910725116729736,
+            "learning_rate": 0.0005,
+            "epoch": 0.29535179468792544,
+            "step": 6590
+        },
+        {
+            "loss": 14.212,
+            "grad_norm": 2.172785997390747,
+            "learning_rate": 0.0005,
+            "epoch": 0.2955758855791909,
+            "step": 6595
+        },
+        {
+            "loss": 14.1142,
+            "grad_norm": 1.8371375799179077,
+            "learning_rate": 0.0005,
+            "epoch": 0.2957999764704564,
+            "step": 6600
+        },
+        {
+            "loss": 14.1355,
+            "grad_norm": 2.132636308670044,
+            "learning_rate": 0.0005,
+            "epoch": 0.29602406736172193,
+            "step": 6605
+        },
+        {
+            "loss": 14.1586,
+            "grad_norm": 1.8999865055084229,
+            "learning_rate": 0.0005,
+            "epoch": 0.2962481582529874,
+            "step": 6610
+        },
+        {
+            "loss": 14.0566,
+            "grad_norm": 1.8841404914855957,
+            "learning_rate": 0.0005,
+            "epoch": 0.2964722491442529,
+            "step": 6615
+        },
+        {
+            "loss": 14.1065,
+            "grad_norm": 1.9028798341751099,
+            "learning_rate": 0.0005,
+            "epoch": 0.2966963400355184,
+            "step": 6620
+        },
+        {
+            "loss": 14.221,
+            "grad_norm": 1.945740818977356,
+            "learning_rate": 0.0005,
+            "epoch": 0.2969204309267839,
+            "step": 6625
+        },
+        {
+            "loss": 14.1517,
+            "grad_norm": 1.882527232170105,
+            "learning_rate": 0.0005,
+            "epoch": 0.2971445218180494,
+            "step": 6630
+        },
+        {
+            "loss": 14.0765,
+            "grad_norm": 1.7825379371643066,
+            "learning_rate": 0.0005,
+            "epoch": 0.2973686127093149,
+            "step": 6635
+        },
+        {
+            "loss": 14.2537,
+            "grad_norm": 1.8401821851730347,
+            "learning_rate": 0.0005,
+            "epoch": 0.2975927036005804,
+            "step": 6640
+        },
+        {
+            "loss": 14.1541,
+            "grad_norm": 2.0787389278411865,
+            "learning_rate": 0.0005,
+            "epoch": 0.2978167944918459,
+            "step": 6645
+        },
+        {
+            "loss": 14.118,
+            "grad_norm": 1.9669815301895142,
+            "learning_rate": 0.0005,
+            "epoch": 0.2980408853831114,
+            "step": 6650
+        },
+        {
+            "loss": 14.1166,
+            "grad_norm": 1.958560824394226,
+            "learning_rate": 0.0005,
+            "epoch": 0.29826497627437687,
+            "step": 6655
+        },
+        {
+            "loss": 14.0878,
+            "grad_norm": 1.839176893234253,
+            "learning_rate": 0.0005,
+            "epoch": 0.2984890671656424,
+            "step": 6660
+        },
+        {
+            "loss": 14.0742,
+            "grad_norm": 1.9698538780212402,
+            "learning_rate": 0.0005,
+            "epoch": 0.2987131580569079,
+            "step": 6665
+        },
+        {
+            "loss": 14.11,
+            "grad_norm": 1.977521300315857,
+            "learning_rate": 0.0005,
+            "epoch": 0.29893724894817336,
+            "step": 6670
+        },
+        {
+            "loss": 14.0945,
+            "grad_norm": 1.8831443786621094,
+            "learning_rate": 0.0005,
+            "epoch": 0.2991613398394389,
+            "step": 6675
+        },
+        {
+            "loss": 14.1022,
+            "grad_norm": 1.9751718044281006,
+            "learning_rate": 0.0005,
+            "epoch": 0.2993854307307044,
+            "step": 6680
+        },
+        {
+            "loss": 14.0919,
+            "grad_norm": 1.9496080875396729,
+            "learning_rate": 0.0005,
+            "epoch": 0.29960952162196985,
+            "step": 6685
+        },
+        {
+            "loss": 14.1258,
+            "grad_norm": 1.9235432147979736,
+            "learning_rate": 0.0005,
+            "epoch": 0.29983361251323537,
+            "step": 6690
+        },
+        {
+            "loss": 14.2083,
+            "grad_norm": 2.081127166748047,
+            "learning_rate": 0.0005,
+            "epoch": 0.3000577034045009,
+            "step": 6695
+        },
+        {
+            "loss": 14.1192,
+            "grad_norm": 1.8495169878005981,
+            "learning_rate": 0.0005,
+            "epoch": 0.30028179429576635,
+            "step": 6700
+        },
+        {
+            "loss": 14.1334,
+            "grad_norm": 1.945540189743042,
+            "learning_rate": 0.0005,
+            "epoch": 0.30050588518703186,
+            "step": 6705
+        },
+        {
+            "loss": 14.0117,
+            "grad_norm": 1.8660441637039185,
+            "learning_rate": 0.0005,
+            "epoch": 0.3007299760782974,
+            "step": 6710
+        },
+        {
+            "loss": 14.0557,
+            "grad_norm": 2.1192219257354736,
+            "learning_rate": 0.0005,
+            "epoch": 0.30095406696956284,
+            "step": 6715
+        },
+        {
+            "loss": 14.144,
+            "grad_norm": 2.134988307952881,
+            "learning_rate": 0.0005,
+            "epoch": 0.30117815786082835,
+            "step": 6720
+        },
+        {
+            "loss": 14.2177,
+            "grad_norm": 1.8217895030975342,
+            "learning_rate": 0.0005,
+            "epoch": 0.30140224875209387,
+            "step": 6725
+        },
+        {
+            "loss": 14.1873,
+            "grad_norm": 1.897010087966919,
+            "learning_rate": 0.0005,
+            "epoch": 0.30162633964335933,
+            "step": 6730
+        },
+        {
+            "loss": 14.0938,
+            "grad_norm": 1.7802810668945312,
+            "learning_rate": 0.0005,
+            "epoch": 0.30185043053462485,
+            "step": 6735
+        },
+        {
+            "loss": 14.1105,
+            "grad_norm": 1.8884286880493164,
+            "learning_rate": 0.0005,
+            "epoch": 0.30207452142589036,
+            "step": 6740
+        },
+        {
+            "loss": 14.0623,
+            "grad_norm": 1.925511360168457,
+            "learning_rate": 0.0005,
+            "epoch": 0.3022986123171558,
+            "step": 6745
+        },
+        {
+            "loss": 14.078,
+            "grad_norm": 1.9921625852584839,
+            "learning_rate": 0.0005,
+            "epoch": 0.30252270320842134,
+            "step": 6750
+        },
+        {
+            "loss": 14.1964,
+            "grad_norm": 2.1286556720733643,
+            "learning_rate": 0.0005,
+            "epoch": 0.30274679409968686,
+            "step": 6755
+        },
+        {
+            "loss": 14.2371,
+            "grad_norm": 1.959270715713501,
+            "learning_rate": 0.0005,
+            "epoch": 0.3029708849909523,
+            "step": 6760
+        },
+        {
+            "loss": 13.9704,
+            "grad_norm": 2.141359806060791,
+            "learning_rate": 0.0005,
+            "epoch": 0.30319497588221783,
+            "step": 6765
+        },
+        {
+            "loss": 14.1439,
+            "grad_norm": 1.8885819911956787,
+            "learning_rate": 0.0005,
+            "epoch": 0.30341906677348335,
+            "step": 6770
+        },
+        {
+            "loss": 14.1109,
+            "grad_norm": 1.8617545366287231,
+            "learning_rate": 0.0005,
+            "epoch": 0.3036431576647488,
+            "step": 6775
+        },
+        {
+            "loss": 14.101,
+            "grad_norm": 2.027916431427002,
+            "learning_rate": 0.0005,
+            "epoch": 0.3038672485560143,
+            "step": 6780
+        },
+        {
+            "loss": 13.9948,
+            "grad_norm": 1.952297568321228,
+            "learning_rate": 0.0005,
+            "epoch": 0.30409133944727984,
+            "step": 6785
+        },
+        {
+            "loss": 14.0892,
+            "grad_norm": 1.88018000125885,
+            "learning_rate": 0.0005,
+            "epoch": 0.3043154303385453,
+            "step": 6790
+        },
+        {
+            "loss": 14.0721,
+            "grad_norm": 2.2838850021362305,
+            "learning_rate": 0.0005,
+            "epoch": 0.3045395212298108,
+            "step": 6795
+        },
+        {
+            "loss": 14.1025,
+            "grad_norm": 2.2587685585021973,
+            "learning_rate": 0.0005,
+            "epoch": 0.30476361212107633,
+            "step": 6800
+        },
+        {
+            "loss": 14.1222,
+            "grad_norm": 2.1795482635498047,
+            "learning_rate": 0.0005,
+            "epoch": 0.3049877030123418,
+            "step": 6805
+        },
+        {
+            "loss": 14.2536,
+            "grad_norm": 1.8609881401062012,
+            "learning_rate": 0.0005,
+            "epoch": 0.3052117939036073,
+            "step": 6810
+        },
+        {
+            "loss": 14.0862,
+            "grad_norm": 2.1692614555358887,
+            "learning_rate": 0.0005,
+            "epoch": 0.3054358847948728,
+            "step": 6815
+        },
+        {
+            "loss": 14.059,
+            "grad_norm": 2.0320756435394287,
+            "learning_rate": 0.0005,
+            "epoch": 0.3056599756861383,
+            "step": 6820
+        },
+        {
+            "loss": 14.0452,
+            "grad_norm": 2.1237852573394775,
+            "learning_rate": 0.0005,
+            "epoch": 0.3058840665774038,
+            "step": 6825
+        },
+        {
+            "loss": 14.0988,
+            "grad_norm": 1.8583446741104126,
+            "learning_rate": 0.0005,
+            "epoch": 0.3061081574686693,
+            "step": 6830
+        },
+        {
+            "loss": 14.0915,
+            "grad_norm": 2.0603761672973633,
+            "learning_rate": 0.0005,
+            "epoch": 0.3063322483599348,
+            "step": 6835
+        },
+        {
+            "loss": 14.0048,
+            "grad_norm": 2.0896894931793213,
+            "learning_rate": 0.0005,
+            "epoch": 0.3065563392512003,
+            "step": 6840
+        },
+        {
+            "loss": 14.2185,
+            "grad_norm": 2.165130376815796,
+            "learning_rate": 0.0005,
+            "epoch": 0.3067804301424658,
+            "step": 6845
+        },
+        {
+            "loss": 14.157,
+            "grad_norm": 2.0199732780456543,
+            "learning_rate": 0.0005,
+            "epoch": 0.30700452103373127,
+            "step": 6850
+        },
+        {
+            "loss": 14.1059,
+            "grad_norm": 2.059459686279297,
+            "learning_rate": 0.0005,
+            "epoch": 0.3072286119249968,
+            "step": 6855
+        },
+        {
+            "loss": 14.1432,
+            "grad_norm": 2.1732242107391357,
+            "learning_rate": 0.0005,
+            "epoch": 0.3074527028162623,
+            "step": 6860
+        },
+        {
+            "loss": 14.0289,
+            "grad_norm": 1.854026198387146,
+            "learning_rate": 0.0005,
+            "epoch": 0.30767679370752776,
+            "step": 6865
+        },
+        {
+            "loss": 14.0782,
+            "grad_norm": 1.9383268356323242,
+            "learning_rate": 0.0005,
+            "epoch": 0.3079008845987933,
+            "step": 6870
+        },
+        {
+            "loss": 14.0815,
+            "grad_norm": 1.9054582118988037,
+            "learning_rate": 0.0005,
+            "epoch": 0.3081249754900588,
+            "step": 6875
+        },
+        {
+            "loss": 14.1134,
+            "grad_norm": 1.9737333059310913,
+            "learning_rate": 0.0005,
+            "epoch": 0.30834906638132425,
+            "step": 6880
+        },
+        {
+            "loss": 14.0502,
+            "grad_norm": 1.8631350994110107,
+            "learning_rate": 0.0005,
+            "epoch": 0.30857315727258977,
+            "step": 6885
+        },
+        {
+            "loss": 14.1507,
+            "grad_norm": 1.9060652256011963,
+            "learning_rate": 0.0005,
+            "epoch": 0.3087972481638553,
+            "step": 6890
+        },
+        {
+            "loss": 14.0591,
+            "grad_norm": 2.0781631469726562,
+            "learning_rate": 0.0005,
+            "epoch": 0.30902133905512075,
+            "step": 6895
+        },
+        {
+            "loss": 14.012,
+            "grad_norm": 1.8879958391189575,
+            "learning_rate": 0.0005,
+            "epoch": 0.30924542994638626,
+            "step": 6900
+        },
+        {
+            "loss": 14.168,
+            "grad_norm": 1.9677551984786987,
+            "learning_rate": 0.0005,
+            "epoch": 0.3094695208376518,
+            "step": 6905
+        },
+        {
+            "loss": 14.0628,
+            "grad_norm": 2.168494462966919,
+            "learning_rate": 0.0005,
+            "epoch": 0.30969361172891724,
+            "step": 6910
+        },
+        {
+            "loss": 14.1082,
+            "grad_norm": 2.0503041744232178,
+            "learning_rate": 0.0005,
+            "epoch": 0.30991770262018276,
+            "step": 6915
+        },
+        {
+            "loss": 14.1525,
+            "grad_norm": 1.9112447500228882,
+            "learning_rate": 0.0005,
+            "epoch": 0.3101417935114482,
+            "step": 6920
+        },
+        {
+            "loss": 14.0622,
+            "grad_norm": 2.0286989212036133,
+            "learning_rate": 0.0005,
+            "epoch": 0.31036588440271373,
+            "step": 6925
+        },
+        {
+            "loss": 14.0246,
+            "grad_norm": 2.076118230819702,
+            "learning_rate": 0.0005,
+            "epoch": 0.31058997529397925,
+            "step": 6930
+        },
+        {
+            "loss": 14.1809,
+            "grad_norm": 2.0124707221984863,
+            "learning_rate": 0.0005,
+            "epoch": 0.3108140661852447,
+            "step": 6935
+        },
+        {
+            "loss": 14.0884,
+            "grad_norm": 1.8748708963394165,
+            "learning_rate": 0.0005,
+            "epoch": 0.3110381570765102,
+            "step": 6940
+        },
+        {
+            "loss": 14.0756,
+            "grad_norm": 1.9558440446853638,
+            "learning_rate": 0.0005,
+            "epoch": 0.31126224796777574,
+            "step": 6945
+        },
+        {
+            "loss": 14.082,
+            "grad_norm": 1.950158715248108,
+            "learning_rate": 0.0005,
+            "epoch": 0.3114863388590412,
+            "step": 6950
+        },
+        {
+            "loss": 14.0608,
+            "grad_norm": 2.2592644691467285,
+            "learning_rate": 0.0005,
+            "epoch": 0.3117104297503067,
+            "step": 6955
+        },
+        {
+            "loss": 14.1437,
+            "grad_norm": 1.9274147748947144,
+            "learning_rate": 0.0005,
+            "epoch": 0.31193452064157223,
+            "step": 6960
+        },
+        {
+            "loss": 14.1519,
+            "grad_norm": 1.9249687194824219,
+            "learning_rate": 0.0005,
+            "epoch": 0.3121586115328377,
+            "step": 6965
+        },
+        {
+            "loss": 14.1133,
+            "grad_norm": 1.9677976369857788,
+            "learning_rate": 0.0005,
+            "epoch": 0.3123827024241032,
+            "step": 6970
+        },
+        {
+            "loss": 14.027,
+            "grad_norm": 1.7901554107666016,
+            "learning_rate": 0.0005,
+            "epoch": 0.3126067933153687,
+            "step": 6975
+        },
+        {
+            "loss": 14.0051,
+            "grad_norm": 2.0552713871002197,
+            "learning_rate": 0.0005,
+            "epoch": 0.3128308842066342,
+            "step": 6980
+        },
+        {
+            "loss": 14.0906,
+            "grad_norm": 1.7620432376861572,
+            "learning_rate": 0.0005,
+            "epoch": 0.3130549750978997,
+            "step": 6985
+        },
+        {
+            "loss": 14.0703,
+            "grad_norm": 2.0076029300689697,
+            "learning_rate": 0.0005,
+            "epoch": 0.3132790659891652,
+            "step": 6990
+        },
+        {
+            "loss": 14.0815,
+            "grad_norm": 1.863505244255066,
+            "learning_rate": 0.0005,
+            "epoch": 0.3135031568804307,
+            "step": 6995
+        },
+        {
+            "loss": 14.1361,
+            "grad_norm": 1.9563758373260498,
+            "learning_rate": 0.0005,
+            "epoch": 0.3137272477716962,
+            "step": 7000
+        },
+        {
+            "eval_loss": 1.755889892578125,
+            "eval_runtime": 18.4089,
+            "eval_samples_per_second": 890.003,
+            "eval_steps_per_second": 7.985,
+            "epoch": 0.3137272477716962,
+            "step": 7000
+        },
+        {
+            "loss": 14.1046,
+            "grad_norm": 1.793610692024231,
+            "learning_rate": 0.0005,
+            "epoch": 0.3139513386629617,
+            "step": 7005
+        },
+        {
+            "loss": 14.0576,
+            "grad_norm": 1.8872146606445312,
+            "learning_rate": 0.0005,
+            "epoch": 0.31417542955422717,
+            "step": 7010
+        },
+        {
+            "loss": 14.2113,
+            "grad_norm": 2.029392719268799,
+            "learning_rate": 0.0005,
+            "epoch": 0.3143995204454927,
+            "step": 7015
+        },
+        {
+            "loss": 14.1018,
+            "grad_norm": 1.9868321418762207,
+            "learning_rate": 0.0005,
+            "epoch": 0.3146236113367582,
+            "step": 7020
+        },
+        {
+            "loss": 14.2401,
+            "grad_norm": 1.868456244468689,
+            "learning_rate": 0.0005,
+            "epoch": 0.31484770222802366,
+            "step": 7025
+        },
+        {
+            "loss": 14.1273,
+            "grad_norm": 1.926793098449707,
+            "learning_rate": 0.0005,
+            "epoch": 0.3150717931192892,
+            "step": 7030
+        },
+        {
+            "loss": 14.0381,
+            "grad_norm": 2.0602972507476807,
+            "learning_rate": 0.0005,
+            "epoch": 0.3152958840105547,
+            "step": 7035
+        },
+        {
+            "loss": 14.1099,
+            "grad_norm": 1.9188587665557861,
+            "learning_rate": 0.0005,
+            "epoch": 0.31551997490182015,
+            "step": 7040
+        },
+        {
+            "loss": 14.0984,
+            "grad_norm": 1.8581831455230713,
+            "learning_rate": 0.0005,
+            "epoch": 0.31574406579308567,
+            "step": 7045
+        },
+        {
+            "loss": 14.094,
+            "grad_norm": 1.9398704767227173,
+            "learning_rate": 0.0005,
+            "epoch": 0.3159681566843512,
+            "step": 7050
+        },
+        {
+            "loss": 14.1393,
+            "grad_norm": 2.011657953262329,
+            "learning_rate": 0.0005,
+            "epoch": 0.31619224757561665,
+            "step": 7055
+        },
+        {
+            "loss": 14.1263,
+            "grad_norm": 1.9346157312393188,
+            "learning_rate": 0.0005,
+            "epoch": 0.31641633846688216,
+            "step": 7060
+        },
+        {
+            "loss": 14.2194,
+            "grad_norm": 2.0041491985321045,
+            "learning_rate": 0.0005,
+            "epoch": 0.3166404293581477,
+            "step": 7065
+        },
+        {
+            "loss": 14.1546,
+            "grad_norm": 1.8866181373596191,
+            "learning_rate": 0.0005,
+            "epoch": 0.31686452024941314,
+            "step": 7070
+        },
+        {
+            "loss": 14.1474,
+            "grad_norm": 1.9563933610916138,
+            "learning_rate": 0.0005,
+            "epoch": 0.31708861114067866,
+            "step": 7075
+        },
+        {
+            "loss": 14.1298,
+            "grad_norm": 2.010119915008545,
+            "learning_rate": 0.0005,
+            "epoch": 0.31731270203194417,
+            "step": 7080
+        },
+        {
+            "loss": 14.1502,
+            "grad_norm": 2.1588711738586426,
+            "learning_rate": 0.0005,
+            "epoch": 0.31753679292320963,
+            "step": 7085
+        },
+        {
+            "loss": 14.1125,
+            "grad_norm": 2.150607109069824,
+            "learning_rate": 0.0005,
+            "epoch": 0.31776088381447515,
+            "step": 7090
+        },
+        {
+            "loss": 14.1021,
+            "grad_norm": 2.117875099182129,
+            "learning_rate": 0.0005,
+            "epoch": 0.31798497470574066,
+            "step": 7095
+        },
+        {
+            "loss": 14.0199,
+            "grad_norm": 1.9777783155441284,
+            "learning_rate": 0.0005,
+            "epoch": 0.3182090655970061,
+            "step": 7100
+        },
+        {
+            "loss": 14.0376,
+            "grad_norm": 2.0438952445983887,
+            "learning_rate": 0.0005,
+            "epoch": 0.31843315648827164,
+            "step": 7105
+        },
+        {
+            "loss": 14.1883,
+            "grad_norm": 2.117734432220459,
+            "learning_rate": 0.0005,
+            "epoch": 0.31865724737953716,
+            "step": 7110
+        },
+        {
+            "loss": 14.0348,
+            "grad_norm": 1.8688557147979736,
+            "learning_rate": 0.0005,
+            "epoch": 0.3188813382708026,
+            "step": 7115
+        },
+        {
+            "loss": 14.108,
+            "grad_norm": 1.9363583326339722,
+            "learning_rate": 0.0005,
+            "epoch": 0.31910542916206813,
+            "step": 7120
+        },
+        {
+            "loss": 14.11,
+            "grad_norm": 1.835010051727295,
+            "learning_rate": 0.0005,
+            "epoch": 0.31932952005333365,
+            "step": 7125
+        },
+        {
+            "loss": 14.0241,
+            "grad_norm": 1.6924667358398438,
+            "learning_rate": 0.0005,
+            "epoch": 0.3195536109445991,
+            "step": 7130
+        },
+        {
+            "loss": 14.0406,
+            "grad_norm": 1.8923516273498535,
+            "learning_rate": 0.0005,
+            "epoch": 0.3197777018358646,
+            "step": 7135
+        },
+        {
+            "loss": 14.1686,
+            "grad_norm": 2.049652338027954,
+            "learning_rate": 0.0005,
+            "epoch": 0.32000179272713014,
+            "step": 7140
+        },
+        {
+            "loss": 14.0831,
+            "grad_norm": 1.9465097188949585,
+            "learning_rate": 0.0005,
+            "epoch": 0.3202258836183956,
+            "step": 7145
+        },
+        {
+            "loss": 14.1339,
+            "grad_norm": 1.9414552450180054,
+            "learning_rate": 0.0005,
+            "epoch": 0.3204499745096611,
+            "step": 7150
+        },
+        {
+            "loss": 14.0689,
+            "grad_norm": 1.7531272172927856,
+            "learning_rate": 0.0005,
+            "epoch": 0.32067406540092663,
+            "step": 7155
+        },
+        {
+            "loss": 14.1045,
+            "grad_norm": 1.9618552923202515,
+            "learning_rate": 0.0005,
+            "epoch": 0.3208981562921921,
+            "step": 7160
+        },
+        {
+            "loss": 14.0208,
+            "grad_norm": 1.8808432817459106,
+            "learning_rate": 0.0005,
+            "epoch": 0.3211222471834576,
+            "step": 7165
+        },
+        {
+            "loss": 14.0674,
+            "grad_norm": 2.10398530960083,
+            "learning_rate": 0.0005,
+            "epoch": 0.3213463380747231,
+            "step": 7170
+        },
+        {
+            "loss": 14.1006,
+            "grad_norm": 1.930496096611023,
+            "learning_rate": 0.0005,
+            "epoch": 0.3215704289659886,
+            "step": 7175
+        },
+        {
+            "loss": 14.101,
+            "grad_norm": 2.353649139404297,
+            "learning_rate": 0.0005,
+            "epoch": 0.3217945198572541,
+            "step": 7180
+        },
+        {
+            "loss": 14.0608,
+            "grad_norm": 2.090111017227173,
+            "learning_rate": 0.0005,
+            "epoch": 0.3220186107485196,
+            "step": 7185
+        },
+        {
+            "loss": 14.1003,
+            "grad_norm": 1.959304928779602,
+            "learning_rate": 0.0005,
+            "epoch": 0.3222427016397851,
+            "step": 7190
+        },
+        {
+            "loss": 14.0667,
+            "grad_norm": 1.8106657266616821,
+            "learning_rate": 0.0005,
+            "epoch": 0.3224667925310506,
+            "step": 7195
+        },
+        {
+            "loss": 14.1066,
+            "grad_norm": 1.7654136419296265,
+            "learning_rate": 0.0005,
+            "epoch": 0.3226908834223161,
+            "step": 7200
+        },
+        {
+            "loss": 14.1614,
+            "grad_norm": 1.8367727994918823,
+            "learning_rate": 0.0005,
+            "epoch": 0.32291497431358157,
+            "step": 7205
+        },
+        {
+            "loss": 14.1704,
+            "grad_norm": 1.9710397720336914,
+            "learning_rate": 0.0005,
+            "epoch": 0.3231390652048471,
+            "step": 7210
+        },
+        {
+            "loss": 14.1573,
+            "grad_norm": 1.7852445840835571,
+            "learning_rate": 0.0005,
+            "epoch": 0.3233631560961126,
+            "step": 7215
+        },
+        {
+            "loss": 14.2409,
+            "grad_norm": 1.9539028406143188,
+            "learning_rate": 0.0005,
+            "epoch": 0.32358724698737806,
+            "step": 7220
+        },
+        {
+            "loss": 14.0829,
+            "grad_norm": 1.8820106983184814,
+            "learning_rate": 0.0005,
+            "epoch": 0.3238113378786436,
+            "step": 7225
+        },
+        {
+            "loss": 13.9979,
+            "grad_norm": 1.8565396070480347,
+            "learning_rate": 0.0005,
+            "epoch": 0.3240354287699091,
+            "step": 7230
+        },
+        {
+            "loss": 13.9935,
+            "grad_norm": 1.9052975177764893,
+            "learning_rate": 0.0005,
+            "epoch": 0.32425951966117456,
+            "step": 7235
+        },
+        {
+            "loss": 14.1653,
+            "grad_norm": 1.9659631252288818,
+            "learning_rate": 0.0005,
+            "epoch": 0.32448361055244007,
+            "step": 7240
+        },
+        {
+            "loss": 14.1506,
+            "grad_norm": 2.118605375289917,
+            "learning_rate": 0.0005,
+            "epoch": 0.3247077014437056,
+            "step": 7245
+        },
+        {
+            "loss": 14.2135,
+            "grad_norm": 2.056361436843872,
+            "learning_rate": 0.0005,
+            "epoch": 0.32493179233497105,
+            "step": 7250
+        },
+        {
+            "loss": 14.1589,
+            "grad_norm": 2.0376381874084473,
+            "learning_rate": 0.0005,
+            "epoch": 0.32515588322623656,
+            "step": 7255
+        },
+        {
+            "loss": 14.0433,
+            "grad_norm": 2.0109989643096924,
+            "learning_rate": 0.0005,
+            "epoch": 0.3253799741175021,
+            "step": 7260
+        },
+        {
+            "loss": 14.1103,
+            "grad_norm": 1.9517523050308228,
+            "learning_rate": 0.0005,
+            "epoch": 0.32560406500876754,
+            "step": 7265
+        },
+        {
+            "loss": 14.052,
+            "grad_norm": 2.0111052989959717,
+            "learning_rate": 0.0005,
+            "epoch": 0.32582815590003306,
+            "step": 7270
+        },
+        {
+            "loss": 14.0868,
+            "grad_norm": 1.9086047410964966,
+            "learning_rate": 0.0005,
+            "epoch": 0.32605224679129857,
+            "step": 7275
+        },
+        {
+            "loss": 14.1952,
+            "grad_norm": 1.7707585096359253,
+            "learning_rate": 0.0005,
+            "epoch": 0.32627633768256403,
+            "step": 7280
+        },
+        {
+            "loss": 14.149,
+            "grad_norm": 1.7972558736801147,
+            "learning_rate": 0.0005,
+            "epoch": 0.32650042857382955,
+            "step": 7285
+        },
+        {
+            "loss": 14.0481,
+            "grad_norm": 1.899881362915039,
+            "learning_rate": 0.0005,
+            "epoch": 0.32672451946509506,
+            "step": 7290
+        },
+        {
+            "loss": 14.1443,
+            "grad_norm": 1.952229619026184,
+            "learning_rate": 0.0005,
+            "epoch": 0.3269486103563605,
+            "step": 7295
+        },
+        {
+            "loss": 14.1473,
+            "grad_norm": 1.8101708889007568,
+            "learning_rate": 0.0005,
+            "epoch": 0.32717270124762604,
+            "step": 7300
+        },
+        {
+            "loss": 14.0977,
+            "grad_norm": 2.0748109817504883,
+            "learning_rate": 0.0005,
+            "epoch": 0.32739679213889156,
+            "step": 7305
+        },
+        {
+            "loss": 14.1556,
+            "grad_norm": 1.8937422037124634,
+            "learning_rate": 0.0005,
+            "epoch": 0.327620883030157,
+            "step": 7310
+        },
+        {
+            "loss": 13.9949,
+            "grad_norm": 2.060307502746582,
+            "learning_rate": 0.0005,
+            "epoch": 0.32784497392142253,
+            "step": 7315
+        },
+        {
+            "loss": 14.193,
+            "grad_norm": 1.9886871576309204,
+            "learning_rate": 0.0005,
+            "epoch": 0.32806906481268805,
+            "step": 7320
+        },
+        {
+            "loss": 14.1288,
+            "grad_norm": 1.7452255487442017,
+            "learning_rate": 0.0005,
+            "epoch": 0.3282931557039535,
+            "step": 7325
+        },
+        {
+            "loss": 14.0731,
+            "grad_norm": 1.8011245727539062,
+            "learning_rate": 0.0005,
+            "epoch": 0.328517246595219,
+            "step": 7330
+        },
+        {
+            "loss": 14.1108,
+            "grad_norm": 1.8402856588363647,
+            "learning_rate": 0.0005,
+            "epoch": 0.32874133748648454,
+            "step": 7335
+        },
+        {
+            "loss": 14.1586,
+            "grad_norm": 1.8013571500778198,
+            "learning_rate": 0.0005,
+            "epoch": 0.32896542837775,
+            "step": 7340
+        },
+        {
+            "loss": 14.0897,
+            "grad_norm": 1.759786605834961,
+            "learning_rate": 0.0005,
+            "epoch": 0.3291895192690155,
+            "step": 7345
+        },
+        {
+            "loss": 14.1886,
+            "grad_norm": 1.9093493223190308,
+            "learning_rate": 0.0005,
+            "epoch": 0.32941361016028103,
+            "step": 7350
+        },
+        {
+            "loss": 14.0965,
+            "grad_norm": 1.7506341934204102,
+            "learning_rate": 0.0005,
+            "epoch": 0.3296377010515465,
+            "step": 7355
+        },
+        {
+            "loss": 14.007,
+            "grad_norm": 1.922635793685913,
+            "learning_rate": 0.0005,
+            "epoch": 0.329861791942812,
+            "step": 7360
+        },
+        {
+            "loss": 14.127,
+            "grad_norm": 1.9030444622039795,
+            "learning_rate": 0.0005,
+            "epoch": 0.3300858828340775,
+            "step": 7365
+        },
+        {
+            "loss": 14.0548,
+            "grad_norm": 1.9263081550598145,
+            "learning_rate": 0.0005,
+            "epoch": 0.330309973725343,
+            "step": 7370
+        },
+        {
+            "loss": 14.1499,
+            "grad_norm": 1.7502036094665527,
+            "learning_rate": 0.0005,
+            "epoch": 0.3305340646166085,
+            "step": 7375
+        },
+        {
+            "loss": 14.0007,
+            "grad_norm": 2.1209959983825684,
+            "learning_rate": 0.0005,
+            "epoch": 0.330758155507874,
+            "step": 7380
+        },
+        {
+            "loss": 14.0438,
+            "grad_norm": 2.0219545364379883,
+            "learning_rate": 0.0005,
+            "epoch": 0.3309822463991395,
+            "step": 7385
+        },
+        {
+            "loss": 14.1204,
+            "grad_norm": 1.984423041343689,
+            "learning_rate": 0.0005,
+            "epoch": 0.331206337290405,
+            "step": 7390
+        },
+        {
+            "loss": 14.1094,
+            "grad_norm": 2.0163533687591553,
+            "learning_rate": 0.0005,
+            "epoch": 0.3314304281816705,
+            "step": 7395
+        },
+        {
+            "loss": 14.1005,
+            "grad_norm": 1.7359834909439087,
+            "learning_rate": 0.0005,
+            "epoch": 0.33165451907293597,
+            "step": 7400
+        },
+        {
+            "loss": 14.1199,
+            "grad_norm": 1.8637721538543701,
+            "learning_rate": 0.0005,
+            "epoch": 0.3318786099642015,
+            "step": 7405
+        },
+        {
+            "loss": 14.1242,
+            "grad_norm": 1.927276372909546,
+            "learning_rate": 0.0005,
+            "epoch": 0.332102700855467,
+            "step": 7410
+        },
+        {
+            "loss": 14.1951,
+            "grad_norm": 2.265612840652466,
+            "learning_rate": 0.0005,
+            "epoch": 0.33232679174673246,
+            "step": 7415
+        },
+        {
+            "loss": 14.1503,
+            "grad_norm": 1.9447145462036133,
+            "learning_rate": 0.0005,
+            "epoch": 0.332550882637998,
+            "step": 7420
+        },
+        {
+            "loss": 14.0934,
+            "grad_norm": 1.9777686595916748,
+            "learning_rate": 0.0005,
+            "epoch": 0.3327749735292635,
+            "step": 7425
+        },
+        {
+            "loss": 14.0585,
+            "grad_norm": 1.8262908458709717,
+            "learning_rate": 0.0005,
+            "epoch": 0.33299906442052896,
+            "step": 7430
+        },
+        {
+            "loss": 14.0943,
+            "grad_norm": 2.1320831775665283,
+            "learning_rate": 0.0005,
+            "epoch": 0.33322315531179447,
+            "step": 7435
+        },
+        {
+            "loss": 14.1044,
+            "grad_norm": 2.061400890350342,
+            "learning_rate": 0.0005,
+            "epoch": 0.33344724620306,
+            "step": 7440
+        },
+        {
+            "loss": 14.0905,
+            "grad_norm": 1.9482510089874268,
+            "learning_rate": 0.0005,
+            "epoch": 0.33367133709432545,
+            "step": 7445
+        },
+        {
+            "loss": 14.1586,
+            "grad_norm": 1.8687303066253662,
+            "learning_rate": 0.0005,
+            "epoch": 0.33389542798559096,
+            "step": 7450
+        },
+        {
+            "loss": 13.9945,
+            "grad_norm": 1.830513834953308,
+            "learning_rate": 0.0005,
+            "epoch": 0.3341195188768565,
+            "step": 7455
+        },
+        {
+            "loss": 14.1178,
+            "grad_norm": 1.9088151454925537,
+            "learning_rate": 0.0005,
+            "epoch": 0.33434360976812194,
+            "step": 7460
+        },
+        {
+            "loss": 14.0652,
+            "grad_norm": 1.9446756839752197,
+            "learning_rate": 0.0005,
+            "epoch": 0.33456770065938746,
+            "step": 7465
+        },
+        {
+            "loss": 14.0836,
+            "grad_norm": 1.9063694477081299,
+            "learning_rate": 0.0005,
+            "epoch": 0.3347917915506529,
+            "step": 7470
+        },
+        {
+            "loss": 14.1995,
+            "grad_norm": 2.0292277336120605,
+            "learning_rate": 0.0005,
+            "epoch": 0.33501588244191843,
+            "step": 7475
+        },
+        {
+            "loss": 14.0583,
+            "grad_norm": 1.8030997514724731,
+            "learning_rate": 0.0005,
+            "epoch": 0.33523997333318395,
+            "step": 7480
+        },
+        {
+            "loss": 14.0631,
+            "grad_norm": 1.8953999280929565,
+            "learning_rate": 0.0005,
+            "epoch": 0.3354640642244494,
+            "step": 7485
+        },
+        {
+            "loss": 14.0377,
+            "grad_norm": 1.8723889589309692,
+            "learning_rate": 0.0005,
+            "epoch": 0.3356881551157149,
+            "step": 7490
+        },
+        {
+            "loss": 14.1068,
+            "grad_norm": 2.212399959564209,
+            "learning_rate": 0.0005,
+            "epoch": 0.33591224600698044,
+            "step": 7495
+        },
+        {
+            "loss": 14.2251,
+            "grad_norm": 1.9140340089797974,
+            "learning_rate": 0.0005,
+            "epoch": 0.3361363368982459,
+            "step": 7500
+        },
+        {
+            "eval_loss": 1.758628249168396,
+            "eval_runtime": 18.346,
+            "eval_samples_per_second": 893.054,
+            "eval_steps_per_second": 8.013,
+            "epoch": 0.3361363368982459,
+            "step": 7500
+        },
+        {
+            "loss": 14.0754,
+            "grad_norm": 1.972398042678833,
+            "learning_rate": 0.0005,
+            "epoch": 0.3363604277895114,
+            "step": 7505
+        },
+        {
+            "loss": 14.1095,
+            "grad_norm": 2.093316078186035,
+            "learning_rate": 0.0005,
+            "epoch": 0.33658451868077693,
+            "step": 7510
+        },
+        {
+            "loss": 14.1565,
+            "grad_norm": 1.9581537246704102,
+            "learning_rate": 0.0005,
+            "epoch": 0.3368086095720424,
+            "step": 7515
+        },
+        {
+            "loss": 14.0715,
+            "grad_norm": 1.897660732269287,
+            "learning_rate": 0.0005,
+            "epoch": 0.3370327004633079,
+            "step": 7520
+        },
+        {
+            "loss": 14.1641,
+            "grad_norm": 1.992495059967041,
+            "learning_rate": 0.0005,
+            "epoch": 0.3372567913545734,
+            "step": 7525
+        },
+        {
+            "loss": 14.1386,
+            "grad_norm": 1.9933212995529175,
+            "learning_rate": 0.0005,
+            "epoch": 0.3374808822458389,
+            "step": 7530
+        },
+        {
+            "loss": 14.0411,
+            "grad_norm": 2.030996561050415,
+            "learning_rate": 0.0005,
+            "epoch": 0.3377049731371044,
+            "step": 7535
+        },
+        {
+            "loss": 14.1321,
+            "grad_norm": 1.9262065887451172,
+            "learning_rate": 0.0005,
+            "epoch": 0.3379290640283699,
+            "step": 7540
+        },
+        {
+            "loss": 14.1055,
+            "grad_norm": 2.041747808456421,
+            "learning_rate": 0.0005,
+            "epoch": 0.3381531549196354,
+            "step": 7545
+        },
+        {
+            "loss": 14.103,
+            "grad_norm": 1.9697359800338745,
+            "learning_rate": 0.0005,
+            "epoch": 0.3383772458109009,
+            "step": 7550
+        },
+        {
+            "loss": 14.0462,
+            "grad_norm": 1.863049864768982,
+            "learning_rate": 0.0005,
+            "epoch": 0.3386013367021664,
+            "step": 7555
+        },
+        {
+            "loss": 13.9718,
+            "grad_norm": 1.8301945924758911,
+            "learning_rate": 0.0005,
+            "epoch": 0.33882542759343187,
+            "step": 7560
+        },
+        {
+            "loss": 13.9904,
+            "grad_norm": 1.8847441673278809,
+            "learning_rate": 0.0005,
+            "epoch": 0.3390495184846974,
+            "step": 7565
+        },
+        {
+            "loss": 14.0646,
+            "grad_norm": 1.8905744552612305,
+            "learning_rate": 0.0005,
+            "epoch": 0.3392736093759629,
+            "step": 7570
+        },
+        {
+            "loss": 14.0983,
+            "grad_norm": 2.0466771125793457,
+            "learning_rate": 0.0005,
+            "epoch": 0.33949770026722836,
+            "step": 7575
+        },
+        {
+            "loss": 14.0606,
+            "grad_norm": 1.8803582191467285,
+            "learning_rate": 0.0005,
+            "epoch": 0.3397217911584939,
+            "step": 7580
+        },
+        {
+            "loss": 14.078,
+            "grad_norm": 1.9058480262756348,
+            "learning_rate": 0.0005,
+            "epoch": 0.3399458820497594,
+            "step": 7585
+        },
+        {
+            "loss": 13.9842,
+            "grad_norm": 1.855478286743164,
+            "learning_rate": 0.0005,
+            "epoch": 0.34016997294102486,
+            "step": 7590
+        },
+        {
+            "loss": 14.0922,
+            "grad_norm": 1.8349817991256714,
+            "learning_rate": 0.0005,
+            "epoch": 0.34039406383229037,
+            "step": 7595
+        },
+        {
+            "loss": 14.0174,
+            "grad_norm": 2.0688560009002686,
+            "learning_rate": 0.0005,
+            "epoch": 0.3406181547235559,
+            "step": 7600
+        },
+        {
+            "loss": 14.0327,
+            "grad_norm": 2.0723955631256104,
+            "learning_rate": 0.0005,
+            "epoch": 0.34084224561482135,
+            "step": 7605
+        },
+        {
+            "loss": 14.0648,
+            "grad_norm": 1.9300533533096313,
+            "learning_rate": 0.0005,
+            "epoch": 0.34106633650608686,
+            "step": 7610
+        },
+        {
+            "loss": 14.1158,
+            "grad_norm": 1.8659693002700806,
+            "learning_rate": 0.0005,
+            "epoch": 0.3412904273973524,
+            "step": 7615
+        },
+        {
+            "loss": 14.0362,
+            "grad_norm": 2.035356283187866,
+            "learning_rate": 0.0005,
+            "epoch": 0.34151451828861784,
+            "step": 7620
+        },
+        {
+            "loss": 14.1092,
+            "grad_norm": 1.8583754301071167,
+            "learning_rate": 0.0005,
+            "epoch": 0.34173860917988336,
+            "step": 7625
+        },
+        {
+            "loss": 14.0497,
+            "grad_norm": 1.897011637687683,
+            "learning_rate": 0.0005,
+            "epoch": 0.3419627000711489,
+            "step": 7630
+        },
+        {
+            "loss": 14.1952,
+            "grad_norm": 1.9849909543991089,
+            "learning_rate": 0.0005,
+            "epoch": 0.34218679096241433,
+            "step": 7635
+        },
+        {
+            "loss": 14.0395,
+            "grad_norm": 1.7245944738388062,
+            "learning_rate": 0.0005,
+            "epoch": 0.34241088185367985,
+            "step": 7640
+        },
+        {
+            "loss": 14.0689,
+            "grad_norm": 1.7866997718811035,
+            "learning_rate": 0.0005,
+            "epoch": 0.34263497274494537,
+            "step": 7645
+        },
+        {
+            "loss": 14.1168,
+            "grad_norm": 2.009934186935425,
+            "learning_rate": 0.0005,
+            "epoch": 0.3428590636362108,
+            "step": 7650
+        },
+        {
+            "loss": 14.1508,
+            "grad_norm": 1.8619006872177124,
+            "learning_rate": 0.0005,
+            "epoch": 0.34308315452747634,
+            "step": 7655
+        },
+        {
+            "loss": 14.1192,
+            "grad_norm": 1.9013493061065674,
+            "learning_rate": 0.0005,
+            "epoch": 0.34330724541874186,
+            "step": 7660
+        },
+        {
+            "loss": 14.0782,
+            "grad_norm": 1.7894394397735596,
+            "learning_rate": 0.0005,
+            "epoch": 0.3435313363100073,
+            "step": 7665
+        },
+        {
+            "loss": 14.0461,
+            "grad_norm": 1.9507958889007568,
+            "learning_rate": 0.0005,
+            "epoch": 0.34375542720127283,
+            "step": 7670
+        },
+        {
+            "loss": 14.0361,
+            "grad_norm": 1.9026128053665161,
+            "learning_rate": 0.0005,
+            "epoch": 0.34397951809253835,
+            "step": 7675
+        },
+        {
+            "loss": 14.1364,
+            "grad_norm": 2.021289825439453,
+            "learning_rate": 0.0005,
+            "epoch": 0.3442036089838038,
+            "step": 7680
+        },
+        {
+            "loss": 14.15,
+            "grad_norm": 2.0237743854522705,
+            "learning_rate": 0.0005,
+            "epoch": 0.3444276998750693,
+            "step": 7685
+        },
+        {
+            "loss": 14.0614,
+            "grad_norm": 1.9201725721359253,
+            "learning_rate": 0.0005,
+            "epoch": 0.34465179076633484,
+            "step": 7690
+        },
+        {
+            "loss": 14.2131,
+            "grad_norm": 1.8731049299240112,
+            "learning_rate": 0.0005,
+            "epoch": 0.3448758816576003,
+            "step": 7695
+        },
+        {
+            "loss": 14.0291,
+            "grad_norm": 1.8014097213745117,
+            "learning_rate": 0.0005,
+            "epoch": 0.3450999725488658,
+            "step": 7700
+        },
+        {
+            "loss": 14.0754,
+            "grad_norm": 1.9392848014831543,
+            "learning_rate": 0.0005,
+            "epoch": 0.34532406344013133,
+            "step": 7705
+        },
+        {
+            "loss": 14.1579,
+            "grad_norm": 2.0457603931427,
+            "learning_rate": 0.0005,
+            "epoch": 0.3455481543313968,
+            "step": 7710
+        },
+        {
+            "loss": 14.1939,
+            "grad_norm": 1.9583426713943481,
+            "learning_rate": 0.0005,
+            "epoch": 0.3457722452226623,
+            "step": 7715
+        },
+        {
+            "loss": 14.1407,
+            "grad_norm": 2.2009177207946777,
+            "learning_rate": 0.0005,
+            "epoch": 0.3459963361139278,
+            "step": 7720
+        },
+        {
+            "loss": 14.0798,
+            "grad_norm": 1.8904290199279785,
+            "learning_rate": 0.0005,
+            "epoch": 0.3462204270051933,
+            "step": 7725
+        },
+        {
+            "loss": 14.1414,
+            "grad_norm": 1.900882601737976,
+            "learning_rate": 0.0005,
+            "epoch": 0.3464445178964588,
+            "step": 7730
+        },
+        {
+            "loss": 14.0602,
+            "grad_norm": 1.8346198797225952,
+            "learning_rate": 0.0005,
+            "epoch": 0.3466686087877243,
+            "step": 7735
+        },
+        {
+            "loss": 13.9476,
+            "grad_norm": 1.9995858669281006,
+            "learning_rate": 0.0005,
+            "epoch": 0.3468926996789898,
+            "step": 7740
+        },
+        {
+            "loss": 14.1434,
+            "grad_norm": 2.0434787273406982,
+            "learning_rate": 0.0005,
+            "epoch": 0.3471167905702553,
+            "step": 7745
+        },
+        {
+            "loss": 14.1186,
+            "grad_norm": 2.11923885345459,
+            "learning_rate": 0.0005,
+            "epoch": 0.3473408814615208,
+            "step": 7750
+        },
+        {
+            "loss": 14.0013,
+            "grad_norm": 2.0023956298828125,
+            "learning_rate": 0.0005,
+            "epoch": 0.34756497235278627,
+            "step": 7755
+        },
+        {
+            "loss": 14.1206,
+            "grad_norm": 2.100276470184326,
+            "learning_rate": 0.0005,
+            "epoch": 0.3477890632440518,
+            "step": 7760
+        },
+        {
+            "loss": 14.1161,
+            "grad_norm": 1.7902406454086304,
+            "learning_rate": 0.0005,
+            "epoch": 0.3480131541353173,
+            "step": 7765
+        },
+        {
+            "loss": 14.0546,
+            "grad_norm": 2.0395708084106445,
+            "learning_rate": 0.0005,
+            "epoch": 0.34823724502658276,
+            "step": 7770
+        },
+        {
+            "loss": 14.1542,
+            "grad_norm": 2.015235185623169,
+            "learning_rate": 0.0005,
+            "epoch": 0.3484613359178483,
+            "step": 7775
+        },
+        {
+            "loss": 14.0569,
+            "grad_norm": 1.918387770652771,
+            "learning_rate": 0.0005,
+            "epoch": 0.3486854268091138,
+            "step": 7780
+        },
+        {
+            "loss": 14.145,
+            "grad_norm": 1.950944185256958,
+            "learning_rate": 0.0005,
+            "epoch": 0.34890951770037926,
+            "step": 7785
+        },
+        {
+            "loss": 13.9847,
+            "grad_norm": 2.028846263885498,
+            "learning_rate": 0.0005,
+            "epoch": 0.3491336085916448,
+            "step": 7790
+        },
+        {
+            "loss": 14.1951,
+            "grad_norm": 1.9557349681854248,
+            "learning_rate": 0.0005,
+            "epoch": 0.3493576994829103,
+            "step": 7795
+        },
+        {
+            "loss": 14.1204,
+            "grad_norm": 2.0673904418945312,
+            "learning_rate": 0.0005,
+            "epoch": 0.34958179037417575,
+            "step": 7800
+        },
+        {
+            "loss": 14.0529,
+            "grad_norm": 1.918513298034668,
+            "learning_rate": 0.0005,
+            "epoch": 0.34980588126544127,
+            "step": 7805
+        },
+        {
+            "loss": 14.0581,
+            "grad_norm": 1.9200891256332397,
+            "learning_rate": 0.0005,
+            "epoch": 0.3500299721567068,
+            "step": 7810
+        },
+        {
+            "loss": 14.0796,
+            "grad_norm": 1.8107129335403442,
+            "learning_rate": 0.0005,
+            "epoch": 0.35025406304797224,
+            "step": 7815
+        },
+        {
+            "loss": 14.2017,
+            "grad_norm": 1.9424742460250854,
+            "learning_rate": 0.0005,
+            "epoch": 0.35047815393923776,
+            "step": 7820
+        },
+        {
+            "loss": 14.0347,
+            "grad_norm": 1.8949542045593262,
+            "learning_rate": 0.0005,
+            "epoch": 0.3507022448305033,
+            "step": 7825
+        },
+        {
+            "loss": 14.1288,
+            "grad_norm": 1.9397120475769043,
+            "learning_rate": 0.0005,
+            "epoch": 0.35092633572176873,
+            "step": 7830
+        },
+        {
+            "loss": 14.0256,
+            "grad_norm": 1.869429349899292,
+            "learning_rate": 0.0005,
+            "epoch": 0.35115042661303425,
+            "step": 7835
+        },
+        {
+            "loss": 14.1089,
+            "grad_norm": 1.8402559757232666,
+            "learning_rate": 0.0005,
+            "epoch": 0.35137451750429977,
+            "step": 7840
+        },
+        {
+            "loss": 14.1165,
+            "grad_norm": 2.3390116691589355,
+            "learning_rate": 0.0005,
+            "epoch": 0.3515986083955652,
+            "step": 7845
+        },
+        {
+            "loss": 14.0348,
+            "grad_norm": 1.8177026510238647,
+            "learning_rate": 0.0005,
+            "epoch": 0.35182269928683074,
+            "step": 7850
+        },
+        {
+            "loss": 14.0949,
+            "grad_norm": 1.8412022590637207,
+            "learning_rate": 0.0005,
+            "epoch": 0.35204679017809626,
+            "step": 7855
+        },
+        {
+            "loss": 14.1576,
+            "grad_norm": 1.8210394382476807,
+            "learning_rate": 0.0005,
+            "epoch": 0.3522708810693617,
+            "step": 7860
+        },
+        {
+            "loss": 14.027,
+            "grad_norm": 2.008986234664917,
+            "learning_rate": 0.0005,
+            "epoch": 0.35249497196062723,
+            "step": 7865
+        },
+        {
+            "loss": 14.0451,
+            "grad_norm": 1.8268238306045532,
+            "learning_rate": 0.0005,
+            "epoch": 0.35271906285189275,
+            "step": 7870
+        },
+        {
+            "loss": 14.0394,
+            "grad_norm": 1.906351923942566,
+            "learning_rate": 0.0005,
+            "epoch": 0.3529431537431582,
+            "step": 7875
+        },
+        {
+            "loss": 14.0353,
+            "grad_norm": 1.8411533832550049,
+            "learning_rate": 0.0005,
+            "epoch": 0.3531672446344237,
+            "step": 7880
+        },
+        {
+            "loss": 14.0102,
+            "grad_norm": 1.8553107976913452,
+            "learning_rate": 0.0005,
+            "epoch": 0.35339133552568924,
+            "step": 7885
+        },
+        {
+            "loss": 14.1753,
+            "grad_norm": 1.7922219038009644,
+            "learning_rate": 0.0005,
+            "epoch": 0.3536154264169547,
+            "step": 7890
+        },
+        {
+            "loss": 14.2017,
+            "grad_norm": 1.9880884885787964,
+            "learning_rate": 0.0005,
+            "epoch": 0.3538395173082202,
+            "step": 7895
+        },
+        {
+            "loss": 14.1431,
+            "grad_norm": 1.9502806663513184,
+            "learning_rate": 0.0005,
+            "epoch": 0.35406360819948574,
+            "step": 7900
+        },
+        {
+            "loss": 14.1458,
+            "grad_norm": 1.8457231521606445,
+            "learning_rate": 0.0005,
+            "epoch": 0.3542876990907512,
+            "step": 7905
+        },
+        {
+            "loss": 14.2067,
+            "grad_norm": 1.8743624687194824,
+            "learning_rate": 0.0005,
+            "epoch": 0.3545117899820167,
+            "step": 7910
+        },
+        {
+            "loss": 14.0933,
+            "grad_norm": 1.9435484409332275,
+            "learning_rate": 0.0005,
+            "epoch": 0.3547358808732822,
+            "step": 7915
+        },
+        {
+            "loss": 14.1251,
+            "grad_norm": 2.1356041431427,
+            "learning_rate": 0.0005,
+            "epoch": 0.3549599717645477,
+            "step": 7920
+        },
+        {
+            "loss": 14.033,
+            "grad_norm": 2.004192590713501,
+            "learning_rate": 0.0005,
+            "epoch": 0.3551840626558132,
+            "step": 7925
+        },
+        {
+            "loss": 14.0399,
+            "grad_norm": 2.1297152042388916,
+            "learning_rate": 0.0005,
+            "epoch": 0.3554081535470787,
+            "step": 7930
+        },
+        {
+            "loss": 14.1168,
+            "grad_norm": 2.082568407058716,
+            "learning_rate": 0.0005,
+            "epoch": 0.3556322444383442,
+            "step": 7935
+        },
+        {
+            "loss": 14.1034,
+            "grad_norm": 2.28765869140625,
+            "learning_rate": 0.0005,
+            "epoch": 0.3558563353296097,
+            "step": 7940
+        },
+        {
+            "loss": 14.1522,
+            "grad_norm": 2.0880069732666016,
+            "learning_rate": 0.0005,
+            "epoch": 0.3560804262208752,
+            "step": 7945
+        },
+        {
+            "loss": 14.1474,
+            "grad_norm": 1.9852226972579956,
+            "learning_rate": 0.0005,
+            "epoch": 0.3563045171121407,
+            "step": 7950
+        },
+        {
+            "loss": 14.148,
+            "grad_norm": 1.974709391593933,
+            "learning_rate": 0.0005,
+            "epoch": 0.3565286080034062,
+            "step": 7955
+        },
+        {
+            "loss": 14.0528,
+            "grad_norm": 1.8595918416976929,
+            "learning_rate": 0.0005,
+            "epoch": 0.3567526988946717,
+            "step": 7960
+        },
+        {
+            "loss": 13.9739,
+            "grad_norm": 1.963915467262268,
+            "learning_rate": 0.0005,
+            "epoch": 0.35697678978593717,
+            "step": 7965
+        },
+        {
+            "loss": 14.066,
+            "grad_norm": 1.8649743795394897,
+            "learning_rate": 0.0005,
+            "epoch": 0.3572008806772027,
+            "step": 7970
+        },
+        {
+            "loss": 14.0075,
+            "grad_norm": 1.9949859380722046,
+            "learning_rate": 0.0005,
+            "epoch": 0.3574249715684682,
+            "step": 7975
+        },
+        {
+            "loss": 13.947,
+            "grad_norm": 1.871898889541626,
+            "learning_rate": 0.0005,
+            "epoch": 0.35764906245973366,
+            "step": 7980
+        },
+        {
+            "loss": 14.067,
+            "grad_norm": 1.872648000717163,
+            "learning_rate": 0.0005,
+            "epoch": 0.3578731533509992,
+            "step": 7985
+        },
+        {
+            "loss": 14.1521,
+            "grad_norm": 1.8796306848526,
+            "learning_rate": 0.0005,
+            "epoch": 0.3580972442422647,
+            "step": 7990
+        },
+        {
+            "loss": 14.0508,
+            "grad_norm": 2.0444676876068115,
+            "learning_rate": 0.0005,
+            "epoch": 0.35832133513353015,
+            "step": 7995
+        },
+        {
+            "loss": 14.0414,
+            "grad_norm": 1.9715629816055298,
+            "learning_rate": 0.0005,
+            "epoch": 0.35854542602479567,
+            "step": 8000
+        },
+        {
+            "eval_loss": 1.755042314529419,
+            "eval_runtime": 18.6527,
+            "eval_samples_per_second": 878.373,
+            "eval_steps_per_second": 7.881,
+            "epoch": 0.35854542602479567,
+            "step": 8000
+        },
+        {
+            "loss": 14.0417,
+            "grad_norm": 1.8838669061660767,
+            "learning_rate": 0.0005,
+            "epoch": 0.3587695169160612,
+            "step": 8005
+        },
+        {
+            "loss": 14.1389,
+            "grad_norm": 2.0093634128570557,
+            "learning_rate": 0.0005,
+            "epoch": 0.35899360780732664,
+            "step": 8010
+        },
+        {
+            "loss": 14.1017,
+            "grad_norm": 1.9981526136398315,
+            "learning_rate": 0.0005,
+            "epoch": 0.35921769869859216,
+            "step": 8015
+        },
+        {
+            "loss": 14.2044,
+            "grad_norm": 1.8038954734802246,
+            "learning_rate": 0.0005,
+            "epoch": 0.3594417895898576,
+            "step": 8020
+        },
+        {
+            "loss": 14.0573,
+            "grad_norm": 1.8292555809020996,
+            "learning_rate": 0.0005,
+            "epoch": 0.35966588048112313,
+            "step": 8025
+        },
+        {
+            "loss": 14.118,
+            "grad_norm": 2.0507237911224365,
+            "learning_rate": 0.0005,
+            "epoch": 0.35988997137238865,
+            "step": 8030
+        },
+        {
+            "loss": 14.1929,
+            "grad_norm": 1.936213493347168,
+            "learning_rate": 0.0005,
+            "epoch": 0.3601140622636541,
+            "step": 8035
+        },
+        {
+            "loss": 14.0351,
+            "grad_norm": 1.9231542348861694,
+            "learning_rate": 0.0005,
+            "epoch": 0.3603381531549196,
+            "step": 8040
+        },
+        {
+            "loss": 13.9999,
+            "grad_norm": 1.9181580543518066,
+            "learning_rate": 0.0005,
+            "epoch": 0.36056224404618514,
+            "step": 8045
+        },
+        {
+            "loss": 14.1313,
+            "grad_norm": 1.8799428939819336,
+            "learning_rate": 0.0005,
+            "epoch": 0.3607863349374506,
+            "step": 8050
+        },
+        {
+            "loss": 14.1121,
+            "grad_norm": 1.8789302110671997,
+            "learning_rate": 0.0005,
+            "epoch": 0.3610104258287161,
+            "step": 8055
+        },
+        {
+            "loss": 14.1483,
+            "grad_norm": 1.9037946462631226,
+            "learning_rate": 0.0005,
+            "epoch": 0.36123451671998164,
+            "step": 8060
+        },
+        {
+            "loss": 14.0798,
+            "grad_norm": 1.95029878616333,
+            "learning_rate": 0.0005,
+            "epoch": 0.3614586076112471,
+            "step": 8065
+        },
+        {
+            "loss": 14.0892,
+            "grad_norm": 1.9673045873641968,
+            "learning_rate": 0.0005,
+            "epoch": 0.3616826985025126,
+            "step": 8070
+        },
+        {
+            "loss": 14.0908,
+            "grad_norm": 1.9095479249954224,
+            "learning_rate": 0.0005,
+            "epoch": 0.3619067893937781,
+            "step": 8075
+        },
+        {
+            "loss": 14.1214,
+            "grad_norm": 1.9915626049041748,
+            "learning_rate": 0.0005,
+            "epoch": 0.3621308802850436,
+            "step": 8080
+        },
+        {
+            "loss": 14.2213,
+            "grad_norm": 1.9557825326919556,
+            "learning_rate": 0.0005,
+            "epoch": 0.3623549711763091,
+            "step": 8085
+        },
+        {
+            "loss": 14.0364,
+            "grad_norm": 2.075934886932373,
+            "learning_rate": 0.0005,
+            "epoch": 0.3625790620675746,
+            "step": 8090
+        },
+        {
+            "loss": 14.077,
+            "grad_norm": 1.8182425498962402,
+            "learning_rate": 0.0005,
+            "epoch": 0.3628031529588401,
+            "step": 8095
+        },
+        {
+            "loss": 14.0508,
+            "grad_norm": 2.0258545875549316,
+            "learning_rate": 0.0005,
+            "epoch": 0.3630272438501056,
+            "step": 8100
+        },
+        {
+            "loss": 14.0646,
+            "grad_norm": 2.0093271732330322,
+            "learning_rate": 0.0005,
+            "epoch": 0.3632513347413711,
+            "step": 8105
+        },
+        {
+            "loss": 14.105,
+            "grad_norm": 1.894214391708374,
+            "learning_rate": 0.0005,
+            "epoch": 0.3634754256326366,
+            "step": 8110
+        },
+        {
+            "loss": 14.1646,
+            "grad_norm": 2.042280912399292,
+            "learning_rate": 0.0005,
+            "epoch": 0.3636995165239021,
+            "step": 8115
+        },
+        {
+            "loss": 14.0829,
+            "grad_norm": 1.8640018701553345,
+            "learning_rate": 0.0005,
+            "epoch": 0.3639236074151676,
+            "step": 8120
+        },
+        {
+            "loss": 14.1016,
+            "grad_norm": 1.757461428642273,
+            "learning_rate": 0.0005,
+            "epoch": 0.36414769830643307,
+            "step": 8125
+        },
+        {
+            "loss": 14.0598,
+            "grad_norm": 1.9780429601669312,
+            "learning_rate": 0.0005,
+            "epoch": 0.3643717891976986,
+            "step": 8130
+        },
+        {
+            "loss": 14.1351,
+            "grad_norm": 1.7367233037948608,
+            "learning_rate": 0.0005,
+            "epoch": 0.3645958800889641,
+            "step": 8135
+        },
+        {
+            "loss": 14.0752,
+            "grad_norm": 1.8880066871643066,
+            "learning_rate": 0.0005,
+            "epoch": 0.36481997098022956,
+            "step": 8140
+        },
+        {
+            "loss": 14.0078,
+            "grad_norm": 1.8871039152145386,
+            "learning_rate": 0.0005,
+            "epoch": 0.3650440618714951,
+            "step": 8145
+        },
+        {
+            "loss": 14.1339,
+            "grad_norm": 1.9108654260635376,
+            "learning_rate": 0.0005,
+            "epoch": 0.3652681527627606,
+            "step": 8150
+        },
+        {
+            "loss": 14.112,
+            "grad_norm": 1.7433931827545166,
+            "learning_rate": 0.0005,
+            "epoch": 0.36549224365402605,
+            "step": 8155
+        },
+        {
+            "loss": 14.0514,
+            "grad_norm": 1.8739866018295288,
+            "learning_rate": 0.0005,
+            "epoch": 0.36571633454529157,
+            "step": 8160
+        },
+        {
+            "loss": 14.0728,
+            "grad_norm": 1.8400698900222778,
+            "learning_rate": 0.0005,
+            "epoch": 0.3659404254365571,
+            "step": 8165
+        },
+        {
+            "loss": 14.0749,
+            "grad_norm": 1.9645445346832275,
+            "learning_rate": 0.0005,
+            "epoch": 0.36616451632782254,
+            "step": 8170
+        },
+        {
+            "loss": 14.0766,
+            "grad_norm": 1.7623934745788574,
+            "learning_rate": 0.0005,
+            "epoch": 0.36638860721908806,
+            "step": 8175
+        },
+        {
+            "loss": 14.0931,
+            "grad_norm": 1.822167992591858,
+            "learning_rate": 0.0005,
+            "epoch": 0.3666126981103536,
+            "step": 8180
+        },
+        {
+            "loss": 14.112,
+            "grad_norm": 1.8237991333007812,
+            "learning_rate": 0.0005,
+            "epoch": 0.36683678900161903,
+            "step": 8185
+        },
+        {
+            "loss": 14.047,
+            "grad_norm": 1.8531397581100464,
+            "learning_rate": 0.0005,
+            "epoch": 0.36706087989288455,
+            "step": 8190
+        },
+        {
+            "loss": 14.0997,
+            "grad_norm": 1.9457165002822876,
+            "learning_rate": 0.0005,
+            "epoch": 0.36728497078415007,
+            "step": 8195
+        },
+        {
+            "loss": 14.0801,
+            "grad_norm": 2.0252914428710938,
+            "learning_rate": 0.0005,
+            "epoch": 0.3675090616754155,
+            "step": 8200
+        },
+        {
+            "loss": 14.06,
+            "grad_norm": 1.9659444093704224,
+            "learning_rate": 0.0005,
+            "epoch": 0.36773315256668104,
+            "step": 8205
+        },
+        {
+            "loss": 14.033,
+            "grad_norm": 1.795413613319397,
+            "learning_rate": 0.0005,
+            "epoch": 0.36795724345794656,
+            "step": 8210
+        },
+        {
+            "loss": 14.0703,
+            "grad_norm": 1.954249382019043,
+            "learning_rate": 0.0005,
+            "epoch": 0.368181334349212,
+            "step": 8215
+        },
+        {
+            "loss": 14.1219,
+            "grad_norm": 1.9318137168884277,
+            "learning_rate": 0.0005,
+            "epoch": 0.36840542524047754,
+            "step": 8220
+        },
+        {
+            "loss": 14.1956,
+            "grad_norm": 1.9770511388778687,
+            "learning_rate": 0.0005,
+            "epoch": 0.36862951613174305,
+            "step": 8225
+        },
+        {
+            "loss": 13.991,
+            "grad_norm": 2.029613494873047,
+            "learning_rate": 0.0005,
+            "epoch": 0.3688536070230085,
+            "step": 8230
+        },
+        {
+            "loss": 14.076,
+            "grad_norm": 1.9258540868759155,
+            "learning_rate": 0.0005,
+            "epoch": 0.369077697914274,
+            "step": 8235
+        },
+        {
+            "loss": 14.0669,
+            "grad_norm": 1.8921314477920532,
+            "learning_rate": 0.0005,
+            "epoch": 0.36930178880553954,
+            "step": 8240
+        },
+        {
+            "loss": 14.0538,
+            "grad_norm": 2.094562530517578,
+            "learning_rate": 0.0005,
+            "epoch": 0.369525879696805,
+            "step": 8245
+        },
+        {
+            "loss": 14.0573,
+            "grad_norm": 2.0521252155303955,
+            "learning_rate": 0.0005,
+            "epoch": 0.3697499705880705,
+            "step": 8250
+        },
+        {
+            "loss": 14.1545,
+            "grad_norm": 1.9510191679000854,
+            "learning_rate": 0.0005,
+            "epoch": 0.36997406147933604,
+            "step": 8255
+        },
+        {
+            "loss": 14.1197,
+            "grad_norm": 2.046823501586914,
+            "learning_rate": 0.0005,
+            "epoch": 0.3701981523706015,
+            "step": 8260
+        },
+        {
+            "loss": 14.1504,
+            "grad_norm": 1.987096905708313,
+            "learning_rate": 0.0005,
+            "epoch": 0.370422243261867,
+            "step": 8265
+        },
+        {
+            "loss": 14.0392,
+            "grad_norm": 1.924657940864563,
+            "learning_rate": 0.0005,
+            "epoch": 0.37064633415313253,
+            "step": 8270
+        },
+        {
+            "loss": 14.0724,
+            "grad_norm": 2.036562204360962,
+            "learning_rate": 0.0005,
+            "epoch": 0.370870425044398,
+            "step": 8275
+        },
+        {
+            "loss": 14.1124,
+            "grad_norm": 1.9594347476959229,
+            "learning_rate": 0.0005,
+            "epoch": 0.3710945159356635,
+            "step": 8280
+        },
+        {
+            "loss": 14.0489,
+            "grad_norm": 1.9957181215286255,
+            "learning_rate": 0.0005,
+            "epoch": 0.371318606826929,
+            "step": 8285
+        },
+        {
+            "loss": 14.1009,
+            "grad_norm": 2.141080617904663,
+            "learning_rate": 0.0005,
+            "epoch": 0.3715426977181945,
+            "step": 8290
+        },
+        {
+            "loss": 14.1201,
+            "grad_norm": 1.8952361345291138,
+            "learning_rate": 0.0005,
+            "epoch": 0.37176678860946,
+            "step": 8295
+        },
+        {
+            "loss": 14.0303,
+            "grad_norm": 2.0000150203704834,
+            "learning_rate": 0.0005,
+            "epoch": 0.3719908795007255,
+            "step": 8300
+        },
+        {
+            "loss": 14.1823,
+            "grad_norm": 1.8380866050720215,
+            "learning_rate": 0.0005,
+            "epoch": 0.372214970391991,
+            "step": 8305
+        },
+        {
+            "loss": 14.117,
+            "grad_norm": 1.7979339361190796,
+            "learning_rate": 0.0005,
+            "epoch": 0.3724390612832565,
+            "step": 8310
+        },
+        {
+            "loss": 14.0296,
+            "grad_norm": 2.126523494720459,
+            "learning_rate": 0.0005,
+            "epoch": 0.372663152174522,
+            "step": 8315
+        },
+        {
+            "loss": 13.9201,
+            "grad_norm": 1.9141422510147095,
+            "learning_rate": 0.0005,
+            "epoch": 0.37288724306578747,
+            "step": 8320
+        },
+        {
+            "loss": 14.0944,
+            "grad_norm": 1.9338246583938599,
+            "learning_rate": 0.0005,
+            "epoch": 0.373111333957053,
+            "step": 8325
+        },
+        {
+            "loss": 14.0784,
+            "grad_norm": 1.7644654512405396,
+            "learning_rate": 0.0005,
+            "epoch": 0.3733354248483185,
+            "step": 8330
+        },
+        {
+            "loss": 14.0946,
+            "grad_norm": 1.7906842231750488,
+            "learning_rate": 0.0005,
+            "epoch": 0.37355951573958396,
+            "step": 8335
+        },
+        {
+            "loss": 14.0335,
+            "grad_norm": 1.8195035457611084,
+            "learning_rate": 0.0005,
+            "epoch": 0.3737836066308495,
+            "step": 8340
+        },
+        {
+            "loss": 13.9956,
+            "grad_norm": 1.8171489238739014,
+            "learning_rate": 0.0005,
+            "epoch": 0.374007697522115,
+            "step": 8345
+        },
+        {
+            "loss": 14.0929,
+            "grad_norm": 1.9822993278503418,
+            "learning_rate": 0.0005,
+            "epoch": 0.37423178841338045,
+            "step": 8350
+        },
+        {
+            "loss": 14.075,
+            "grad_norm": 2.0512478351593018,
+            "learning_rate": 0.0005,
+            "epoch": 0.37445587930464597,
+            "step": 8355
+        },
+        {
+            "loss": 14.1075,
+            "grad_norm": 2.137077808380127,
+            "learning_rate": 0.0005,
+            "epoch": 0.3746799701959115,
+            "step": 8360
+        },
+        {
+            "loss": 13.9692,
+            "grad_norm": 1.8836326599121094,
+            "learning_rate": 0.0005,
+            "epoch": 0.37490406108717694,
+            "step": 8365
+        },
+        {
+            "loss": 14.0401,
+            "grad_norm": 1.8308993577957153,
+            "learning_rate": 0.0005,
+            "epoch": 0.37512815197844246,
+            "step": 8370
+        },
+        {
+            "loss": 14.0071,
+            "grad_norm": 1.8756014108657837,
+            "learning_rate": 0.0005,
+            "epoch": 0.375352242869708,
+            "step": 8375
+        },
+        {
+            "loss": 14.057,
+            "grad_norm": 1.8783321380615234,
+            "learning_rate": 0.0005,
+            "epoch": 0.37557633376097344,
+            "step": 8380
+        },
+        {
+            "loss": 14.1134,
+            "grad_norm": 1.9342875480651855,
+            "learning_rate": 0.0005,
+            "epoch": 0.37580042465223895,
+            "step": 8385
+        },
+        {
+            "loss": 14.155,
+            "grad_norm": 1.8861297369003296,
+            "learning_rate": 0.0005,
+            "epoch": 0.37602451554350447,
+            "step": 8390
+        },
+        {
+            "loss": 14.0238,
+            "grad_norm": 1.832170844078064,
+            "learning_rate": 0.0005,
+            "epoch": 0.3762486064347699,
+            "step": 8395
+        },
+        {
+            "loss": 14.0867,
+            "grad_norm": 2.1398568153381348,
+            "learning_rate": 0.0005,
+            "epoch": 0.37647269732603544,
+            "step": 8400
+        },
+        {
+            "loss": 14.1837,
+            "grad_norm": 1.9946504831314087,
+            "learning_rate": 0.0005,
+            "epoch": 0.37669678821730096,
+            "step": 8405
+        },
+        {
+            "loss": 13.9595,
+            "grad_norm": 1.8542578220367432,
+            "learning_rate": 0.0005,
+            "epoch": 0.3769208791085664,
+            "step": 8410
+        },
+        {
+            "loss": 14.067,
+            "grad_norm": 2.0427167415618896,
+            "learning_rate": 0.0005,
+            "epoch": 0.37714496999983194,
+            "step": 8415
+        },
+        {
+            "loss": 14.0449,
+            "grad_norm": 1.8909755945205688,
+            "learning_rate": 0.0005,
+            "epoch": 0.37736906089109745,
+            "step": 8420
+        },
+        {
+            "loss": 14.0803,
+            "grad_norm": 1.9262720346450806,
+            "learning_rate": 0.0005,
+            "epoch": 0.3775931517823629,
+            "step": 8425
+        },
+        {
+            "loss": 14.0376,
+            "grad_norm": 1.8407044410705566,
+            "learning_rate": 0.0005,
+            "epoch": 0.37781724267362843,
+            "step": 8430
+        },
+        {
+            "loss": 14.1077,
+            "grad_norm": 1.8292977809906006,
+            "learning_rate": 0.0005,
+            "epoch": 0.37804133356489394,
+            "step": 8435
+        },
+        {
+            "loss": 14.2441,
+            "grad_norm": 2.043469190597534,
+            "learning_rate": 0.0005,
+            "epoch": 0.3782654244561594,
+            "step": 8440
+        },
+        {
+            "loss": 14.019,
+            "grad_norm": 1.8498077392578125,
+            "learning_rate": 0.0005,
+            "epoch": 0.3784895153474249,
+            "step": 8445
+        },
+        {
+            "loss": 14.0124,
+            "grad_norm": 1.902140736579895,
+            "learning_rate": 0.0005,
+            "epoch": 0.37871360623869044,
+            "step": 8450
+        },
+        {
+            "loss": 14.0932,
+            "grad_norm": 2.1382274627685547,
+            "learning_rate": 0.0005,
+            "epoch": 0.3789376971299559,
+            "step": 8455
+        },
+        {
+            "loss": 14.0854,
+            "grad_norm": 2.017334461212158,
+            "learning_rate": 0.0005,
+            "epoch": 0.3791617880212214,
+            "step": 8460
+        },
+        {
+            "loss": 14.1117,
+            "grad_norm": 1.8728015422821045,
+            "learning_rate": 0.0005,
+            "epoch": 0.37938587891248693,
+            "step": 8465
+        },
+        {
+            "loss": 14.0292,
+            "grad_norm": 1.930253028869629,
+            "learning_rate": 0.0005,
+            "epoch": 0.3796099698037524,
+            "step": 8470
+        },
+        {
+            "loss": 14.0795,
+            "grad_norm": 2.0670037269592285,
+            "learning_rate": 0.0005,
+            "epoch": 0.3798340606950179,
+            "step": 8475
+        },
+        {
+            "loss": 14.0816,
+            "grad_norm": 1.9624706506729126,
+            "learning_rate": 0.0005,
+            "epoch": 0.3800581515862834,
+            "step": 8480
+        },
+        {
+            "loss": 14.116,
+            "grad_norm": 1.935685157775879,
+            "learning_rate": 0.0005,
+            "epoch": 0.3802822424775489,
+            "step": 8485
+        },
+        {
+            "loss": 14.1128,
+            "grad_norm": 1.997086763381958,
+            "learning_rate": 0.0005,
+            "epoch": 0.3805063333688144,
+            "step": 8490
+        },
+        {
+            "loss": 14.0283,
+            "grad_norm": 2.0466487407684326,
+            "learning_rate": 0.0005,
+            "epoch": 0.3807304242600799,
+            "step": 8495
+        },
+        {
+            "loss": 14.1002,
+            "grad_norm": 2.157344102859497,
+            "learning_rate": 0.0005,
+            "epoch": 0.3809545151513454,
+            "step": 8500
+        },
+        {
+            "eval_loss": 1.7569822072982788,
+            "eval_runtime": 18.6681,
+            "eval_samples_per_second": 877.645,
+            "eval_steps_per_second": 7.874,
+            "epoch": 0.3809545151513454,
+            "step": 8500
+        },
+        {
+            "loss": 14.0763,
+            "grad_norm": 2.2164812088012695,
+            "learning_rate": 0.0005,
+            "epoch": 0.3811786060426109,
+            "step": 8505
+        },
+        {
+            "loss": 14.1342,
+            "grad_norm": 2.0805397033691406,
+            "learning_rate": 0.0005,
+            "epoch": 0.3814026969338764,
+            "step": 8510
+        },
+        {
+            "loss": 14.1283,
+            "grad_norm": 1.8117724657058716,
+            "learning_rate": 0.0005,
+            "epoch": 0.38162678782514187,
+            "step": 8515
+        },
+        {
+            "loss": 14.0734,
+            "grad_norm": 1.8892253637313843,
+            "learning_rate": 0.0005,
+            "epoch": 0.3818508787164074,
+            "step": 8520
+        },
+        {
+            "loss": 14.0782,
+            "grad_norm": 2.018918752670288,
+            "learning_rate": 0.0005,
+            "epoch": 0.3820749696076729,
+            "step": 8525
+        },
+        {
+            "loss": 14.1138,
+            "grad_norm": 2.0088467597961426,
+            "learning_rate": 0.0005,
+            "epoch": 0.38229906049893836,
+            "step": 8530
+        },
+        {
+            "loss": 14.1419,
+            "grad_norm": 2.0010855197906494,
+            "learning_rate": 0.0005,
+            "epoch": 0.3825231513902039,
+            "step": 8535
+        },
+        {
+            "loss": 14.0559,
+            "grad_norm": 1.973644495010376,
+            "learning_rate": 0.0005,
+            "epoch": 0.3827472422814694,
+            "step": 8540
+        },
+        {
+            "loss": 13.9804,
+            "grad_norm": 1.8364356756210327,
+            "learning_rate": 0.0005,
+            "epoch": 0.38297133317273485,
+            "step": 8545
+        },
+        {
+            "loss": 14.0609,
+            "grad_norm": 1.7789885997772217,
+            "learning_rate": 0.0005,
+            "epoch": 0.38319542406400037,
+            "step": 8550
+        },
+        {
+            "loss": 14.1007,
+            "grad_norm": 1.9528905153274536,
+            "learning_rate": 0.0005,
+            "epoch": 0.3834195149552659,
+            "step": 8555
+        },
+        {
+            "loss": 14.0495,
+            "grad_norm": 1.8014353513717651,
+            "learning_rate": 0.0005,
+            "epoch": 0.38364360584653134,
+            "step": 8560
+        },
+        {
+            "loss": 14.1449,
+            "grad_norm": 1.844429612159729,
+            "learning_rate": 0.0005,
+            "epoch": 0.38386769673779686,
+            "step": 8565
+        },
+        {
+            "loss": 14.0979,
+            "grad_norm": 2.125936508178711,
+            "learning_rate": 0.0005,
+            "epoch": 0.3840917876290623,
+            "step": 8570
+        },
+        {
+            "loss": 14.0512,
+            "grad_norm": 2.0436089038848877,
+            "learning_rate": 0.0005,
+            "epoch": 0.38431587852032784,
+            "step": 8575
+        },
+        {
+            "loss": 14.0902,
+            "grad_norm": 2.1031296253204346,
+            "learning_rate": 0.0005,
+            "epoch": 0.38453996941159335,
+            "step": 8580
+        },
+        {
+            "loss": 14.0436,
+            "grad_norm": 1.9743539094924927,
+            "learning_rate": 0.0005,
+            "epoch": 0.3847640603028588,
+            "step": 8585
+        },
+        {
+            "loss": 14.0761,
+            "grad_norm": 2.10516357421875,
+            "learning_rate": 0.0005,
+            "epoch": 0.38498815119412433,
+            "step": 8590
+        },
+        {
+            "loss": 14.0811,
+            "grad_norm": 2.1132824420928955,
+            "learning_rate": 0.0005,
+            "epoch": 0.38521224208538984,
+            "step": 8595
+        },
+        {
+            "loss": 14.0995,
+            "grad_norm": 2.0395777225494385,
+            "learning_rate": 0.0005,
+            "epoch": 0.3854363329766553,
+            "step": 8600
+        },
+        {
+            "loss": 14.0018,
+            "grad_norm": 1.8657382726669312,
+            "learning_rate": 0.0005,
+            "epoch": 0.3856604238679208,
+            "step": 8605
+        },
+        {
+            "loss": 14.0678,
+            "grad_norm": 1.8442227840423584,
+            "learning_rate": 0.0005,
+            "epoch": 0.38588451475918634,
+            "step": 8610
+        },
+        {
+            "loss": 14.1595,
+            "grad_norm": 2.0256452560424805,
+            "learning_rate": 0.0005,
+            "epoch": 0.3861086056504518,
+            "step": 8615
+        },
+        {
+            "loss": 14.1808,
+            "grad_norm": 2.03298282623291,
+            "learning_rate": 0.0005,
+            "epoch": 0.3863326965417173,
+            "step": 8620
+        },
+        {
+            "loss": 14.0694,
+            "grad_norm": 1.8715609312057495,
+            "learning_rate": 0.0005,
+            "epoch": 0.38655678743298283,
+            "step": 8625
+        },
+        {
+            "loss": 14.106,
+            "grad_norm": 1.9478174448013306,
+            "learning_rate": 0.0005,
+            "epoch": 0.3867808783242483,
+            "step": 8630
+        },
+        {
+            "loss": 13.9328,
+            "grad_norm": 1.817594051361084,
+            "learning_rate": 0.0005,
+            "epoch": 0.3870049692155138,
+            "step": 8635
+        },
+        {
+            "loss": 14.019,
+            "grad_norm": 1.7071633338928223,
+            "learning_rate": 0.0005,
+            "epoch": 0.3872290601067793,
+            "step": 8640
+        },
+        {
+            "loss": 14.0565,
+            "grad_norm": 1.9565168619155884,
+            "learning_rate": 0.0005,
+            "epoch": 0.3874531509980448,
+            "step": 8645
+        },
+        {
+            "loss": 14.1427,
+            "grad_norm": 1.9424870014190674,
+            "learning_rate": 0.0005,
+            "epoch": 0.3876772418893103,
+            "step": 8650
+        },
+        {
+            "loss": 14.1189,
+            "grad_norm": 1.878967046737671,
+            "learning_rate": 0.0005,
+            "epoch": 0.3879013327805758,
+            "step": 8655
+        },
+        {
+            "loss": 14.1579,
+            "grad_norm": 1.932654619216919,
+            "learning_rate": 0.0005,
+            "epoch": 0.3881254236718413,
+            "step": 8660
+        },
+        {
+            "loss": 14.1557,
+            "grad_norm": 1.8046934604644775,
+            "learning_rate": 0.0005,
+            "epoch": 0.3883495145631068,
+            "step": 8665
+        },
+        {
+            "loss": 14.0229,
+            "grad_norm": 1.834702968597412,
+            "learning_rate": 0.0005,
+            "epoch": 0.3885736054543723,
+            "step": 8670
+        },
+        {
+            "loss": 14.1103,
+            "grad_norm": 1.7599685192108154,
+            "learning_rate": 0.0005,
+            "epoch": 0.38879769634563777,
+            "step": 8675
+        },
+        {
+            "loss": 14.1212,
+            "grad_norm": 1.8775702714920044,
+            "learning_rate": 0.0005,
+            "epoch": 0.3890217872369033,
+            "step": 8680
+        },
+        {
+            "loss": 14.1246,
+            "grad_norm": 2.1239051818847656,
+            "learning_rate": 0.0005,
+            "epoch": 0.3892458781281688,
+            "step": 8685
+        },
+        {
+            "loss": 13.9983,
+            "grad_norm": 2.047067880630493,
+            "learning_rate": 0.0005,
+            "epoch": 0.38946996901943426,
+            "step": 8690
+        },
+        {
+            "loss": 14.1257,
+            "grad_norm": 1.868910789489746,
+            "learning_rate": 0.0005,
+            "epoch": 0.3896940599106998,
+            "step": 8695
+        },
+        {
+            "loss": 14.1012,
+            "grad_norm": 2.0451931953430176,
+            "learning_rate": 0.0005,
+            "epoch": 0.3899181508019653,
+            "step": 8700
+        },
+        {
+            "loss": 14.0864,
+            "grad_norm": 1.8716967105865479,
+            "learning_rate": 0.0005,
+            "epoch": 0.39014224169323075,
+            "step": 8705
+        },
+        {
+            "loss": 14.1215,
+            "grad_norm": 1.8944684267044067,
+            "learning_rate": 0.0005,
+            "epoch": 0.39036633258449627,
+            "step": 8710
+        },
+        {
+            "loss": 14.0179,
+            "grad_norm": 1.885197639465332,
+            "learning_rate": 0.0005,
+            "epoch": 0.3905904234757618,
+            "step": 8715
+        },
+        {
+            "loss": 14.1015,
+            "grad_norm": 2.0148558616638184,
+            "learning_rate": 0.0005,
+            "epoch": 0.39081451436702724,
+            "step": 8720
+        },
+        {
+            "loss": 14.1754,
+            "grad_norm": 1.9229111671447754,
+            "learning_rate": 0.0005,
+            "epoch": 0.39103860525829276,
+            "step": 8725
+        },
+        {
+            "loss": 14.0637,
+            "grad_norm": 1.8236292600631714,
+            "learning_rate": 0.0005,
+            "epoch": 0.3912626961495583,
+            "step": 8730
+        },
+        {
+            "loss": 14.0855,
+            "grad_norm": 1.950639247894287,
+            "learning_rate": 0.0005,
+            "epoch": 0.39148678704082374,
+            "step": 8735
+        },
+        {
+            "loss": 14.0255,
+            "grad_norm": 1.8944975137710571,
+            "learning_rate": 0.0005,
+            "epoch": 0.39171087793208925,
+            "step": 8740
+        },
+        {
+            "loss": 14.1443,
+            "grad_norm": 2.0181682109832764,
+            "learning_rate": 0.0005,
+            "epoch": 0.39193496882335477,
+            "step": 8745
+        },
+        {
+            "loss": 13.984,
+            "grad_norm": 1.8844550848007202,
+            "learning_rate": 0.0005,
+            "epoch": 0.39215905971462023,
+            "step": 8750
+        },
+        {
+            "loss": 14.0865,
+            "grad_norm": 2.0160844326019287,
+            "learning_rate": 0.0005,
+            "epoch": 0.39238315060588574,
+            "step": 8755
+        },
+        {
+            "loss": 14.0881,
+            "grad_norm": 1.9414035081863403,
+            "learning_rate": 0.0005,
+            "epoch": 0.39260724149715126,
+            "step": 8760
+        },
+        {
+            "loss": 14.0639,
+            "grad_norm": 1.839568853378296,
+            "learning_rate": 0.0005,
+            "epoch": 0.3928313323884167,
+            "step": 8765
+        },
+        {
+            "loss": 14.1536,
+            "grad_norm": 1.899423599243164,
+            "learning_rate": 0.0005,
+            "epoch": 0.39305542327968224,
+            "step": 8770
+        },
+        {
+            "loss": 14.1575,
+            "grad_norm": 1.8649468421936035,
+            "learning_rate": 0.0005,
+            "epoch": 0.39327951417094775,
+            "step": 8775
+        },
+        {
+            "loss": 14.0814,
+            "grad_norm": 1.937558889389038,
+            "learning_rate": 0.0005,
+            "epoch": 0.3935036050622132,
+            "step": 8780
+        },
+        {
+            "loss": 14.0491,
+            "grad_norm": 1.889802098274231,
+            "learning_rate": 0.0005,
+            "epoch": 0.39372769595347873,
+            "step": 8785
+        },
+        {
+            "loss": 14.1103,
+            "grad_norm": 2.0124385356903076,
+            "learning_rate": 0.0005,
+            "epoch": 0.39395178684474425,
+            "step": 8790
+        },
+        {
+            "loss": 14.109,
+            "grad_norm": 2.07157564163208,
+            "learning_rate": 0.0005,
+            "epoch": 0.3941758777360097,
+            "step": 8795
+        },
+        {
+            "loss": 14.1463,
+            "grad_norm": 1.7840614318847656,
+            "learning_rate": 0.0005,
+            "epoch": 0.3943999686272752,
+            "step": 8800
+        },
+        {
+            "loss": 14.082,
+            "grad_norm": 1.8262500762939453,
+            "learning_rate": 0.0005,
+            "epoch": 0.39462405951854074,
+            "step": 8805
+        },
+        {
+            "loss": 14.1194,
+            "grad_norm": 1.9340875148773193,
+            "learning_rate": 0.0005,
+            "epoch": 0.3948481504098062,
+            "step": 8810
+        },
+        {
+            "loss": 14.0318,
+            "grad_norm": 1.982276201248169,
+            "learning_rate": 0.0005,
+            "epoch": 0.3950722413010717,
+            "step": 8815
+        },
+        {
+            "loss": 13.9853,
+            "grad_norm": 1.7948544025421143,
+            "learning_rate": 0.0005,
+            "epoch": 0.39529633219233723,
+            "step": 8820
+        },
+        {
+            "loss": 14.0622,
+            "grad_norm": 1.7773853540420532,
+            "learning_rate": 0.0005,
+            "epoch": 0.3955204230836027,
+            "step": 8825
+        },
+        {
+            "loss": 14.1185,
+            "grad_norm": 1.7478654384613037,
+            "learning_rate": 0.0005,
+            "epoch": 0.3957445139748682,
+            "step": 8830
+        },
+        {
+            "loss": 14.0453,
+            "grad_norm": 2.0046229362487793,
+            "learning_rate": 0.0005,
+            "epoch": 0.3959686048661337,
+            "step": 8835
+        },
+        {
+            "loss": 14.196,
+            "grad_norm": 1.9458394050598145,
+            "learning_rate": 0.0005,
+            "epoch": 0.3961926957573992,
+            "step": 8840
+        },
+        {
+            "loss": 14.0311,
+            "grad_norm": 1.9534013271331787,
+            "learning_rate": 0.0005,
+            "epoch": 0.3964167866486647,
+            "step": 8845
+        },
+        {
+            "loss": 14.1093,
+            "grad_norm": 1.789678692817688,
+            "learning_rate": 0.0005,
+            "epoch": 0.3966408775399302,
+            "step": 8850
+        },
+        {
+            "loss": 14.0338,
+            "grad_norm": 1.9533801078796387,
+            "learning_rate": 0.0005,
+            "epoch": 0.3968649684311957,
+            "step": 8855
+        },
+        {
+            "loss": 13.9558,
+            "grad_norm": 2.187878370285034,
+            "learning_rate": 0.0005,
+            "epoch": 0.3970890593224612,
+            "step": 8860
+        },
+        {
+            "loss": 14.0059,
+            "grad_norm": 1.9042341709136963,
+            "learning_rate": 0.0005,
+            "epoch": 0.3973131502137267,
+            "step": 8865
+        },
+        {
+            "loss": 14.1696,
+            "grad_norm": 2.0896317958831787,
+            "learning_rate": 0.0005,
+            "epoch": 0.39753724110499217,
+            "step": 8870
+        },
+        {
+            "loss": 14.1228,
+            "grad_norm": 1.904725193977356,
+            "learning_rate": 0.0005,
+            "epoch": 0.3977613319962577,
+            "step": 8875
+        },
+        {
+            "loss": 14.0367,
+            "grad_norm": 1.8315083980560303,
+            "learning_rate": 0.0005,
+            "epoch": 0.3979854228875232,
+            "step": 8880
+        },
+        {
+            "loss": 14.1059,
+            "grad_norm": 1.8545987606048584,
+            "learning_rate": 0.0005,
+            "epoch": 0.39820951377878866,
+            "step": 8885
+        },
+        {
+            "loss": 14.0085,
+            "grad_norm": 2.000612258911133,
+            "learning_rate": 0.0005,
+            "epoch": 0.3984336046700542,
+            "step": 8890
+        },
+        {
+            "loss": 14.0234,
+            "grad_norm": 1.8360822200775146,
+            "learning_rate": 0.0005,
+            "epoch": 0.3986576955613197,
+            "step": 8895
+        },
+        {
+            "loss": 14.0266,
+            "grad_norm": 1.7959775924682617,
+            "learning_rate": 0.0005,
+            "epoch": 0.39888178645258515,
+            "step": 8900
+        },
+        {
+            "loss": 14.179,
+            "grad_norm": 2.0182433128356934,
+            "learning_rate": 0.0005,
+            "epoch": 0.39910587734385067,
+            "step": 8905
+        },
+        {
+            "loss": 14.0248,
+            "grad_norm": 2.0366687774658203,
+            "learning_rate": 0.0005,
+            "epoch": 0.3993299682351162,
+            "step": 8910
+        },
+        {
+            "loss": 14.0431,
+            "grad_norm": 2.001441717147827,
+            "learning_rate": 0.0005,
+            "epoch": 0.39955405912638164,
+            "step": 8915
+        },
+        {
+            "loss": 14.222,
+            "grad_norm": 1.8706127405166626,
+            "learning_rate": 0.0005,
+            "epoch": 0.39977815001764716,
+            "step": 8920
+        },
+        {
+            "loss": 14.1706,
+            "grad_norm": 1.9963351488113403,
+            "learning_rate": 0.0005,
+            "epoch": 0.4000022409089127,
+            "step": 8925
+        },
+        {
+            "loss": 14.1615,
+            "grad_norm": 2.0632803440093994,
+            "learning_rate": 0.0005,
+            "epoch": 0.40022633180017814,
+            "step": 8930
+        },
+        {
+            "loss": 14.0615,
+            "grad_norm": 2.052077293395996,
+            "learning_rate": 0.0005,
+            "epoch": 0.40045042269144365,
+            "step": 8935
+        },
+        {
+            "loss": 14.0403,
+            "grad_norm": 1.9180212020874023,
+            "learning_rate": 0.0005,
+            "epoch": 0.40067451358270917,
+            "step": 8940
+        },
+        {
+            "loss": 14.0491,
+            "grad_norm": 1.9582902193069458,
+            "learning_rate": 0.0005,
+            "epoch": 0.40089860447397463,
+            "step": 8945
+        },
+        {
+            "loss": 14.1251,
+            "grad_norm": 1.6565606594085693,
+            "learning_rate": 0.0005,
+            "epoch": 0.40112269536524015,
+            "step": 8950
+        },
+        {
+            "loss": 14.09,
+            "grad_norm": 2.0036303997039795,
+            "learning_rate": 0.0005,
+            "epoch": 0.40134678625650566,
+            "step": 8955
+        },
+        {
+            "loss": 14.128,
+            "grad_norm": 1.8896582126617432,
+            "learning_rate": 0.0005,
+            "epoch": 0.4015708771477711,
+            "step": 8960
+        },
+        {
+            "loss": 14.0465,
+            "grad_norm": 1.9904286861419678,
+            "learning_rate": 0.0005,
+            "epoch": 0.40179496803903664,
+            "step": 8965
+        },
+        {
+            "loss": 14.0307,
+            "grad_norm": 1.9393749237060547,
+            "learning_rate": 0.0005,
+            "epoch": 0.40201905893030215,
+            "step": 8970
+        },
+        {
+            "loss": 14.0348,
+            "grad_norm": 1.9569603204727173,
+            "learning_rate": 0.0005,
+            "epoch": 0.4022431498215676,
+            "step": 8975
+        },
+        {
+            "loss": 14.1068,
+            "grad_norm": 1.8881086111068726,
+            "learning_rate": 0.0005,
+            "epoch": 0.40246724071283313,
+            "step": 8980
+        },
+        {
+            "loss": 13.9595,
+            "grad_norm": 1.8988546133041382,
+            "learning_rate": 0.0005,
+            "epoch": 0.40269133160409865,
+            "step": 8985
+        },
+        {
+            "loss": 14.0135,
+            "grad_norm": 1.9805768728256226,
+            "learning_rate": 0.0005,
+            "epoch": 0.4029154224953641,
+            "step": 8990
+        },
+        {
+            "loss": 14.0296,
+            "grad_norm": 1.8838372230529785,
+            "learning_rate": 0.0005,
+            "epoch": 0.4031395133866296,
+            "step": 8995
+        },
+        {
+            "loss": 14.1413,
+            "grad_norm": 1.8862353563308716,
+            "learning_rate": 0.0005,
+            "epoch": 0.40336360427789514,
+            "step": 9000
+        },
+        {
+            "eval_loss": 1.750436782836914,
+            "eval_runtime": 18.7449,
+            "eval_samples_per_second": 874.051,
+            "eval_steps_per_second": 7.842,
+            "epoch": 0.40336360427789514,
+            "step": 9000
+        },
+        {
+            "loss": 14.0521,
+            "grad_norm": 1.953079104423523,
+            "learning_rate": 0.0005,
+            "epoch": 0.4035876951691606,
+            "step": 9005
+        },
+        {
+            "loss": 14.1428,
+            "grad_norm": 1.8866537809371948,
+            "learning_rate": 0.0005,
+            "epoch": 0.4038117860604261,
+            "step": 9010
+        },
+        {
+            "loss": 14.0127,
+            "grad_norm": 2.0234899520874023,
+            "learning_rate": 0.0005,
+            "epoch": 0.40403587695169163,
+            "step": 9015
+        },
+        {
+            "loss": 14.0549,
+            "grad_norm": 1.7747480869293213,
+            "learning_rate": 0.0005,
+            "epoch": 0.4042599678429571,
+            "step": 9020
+        },
+        {
+            "loss": 13.9821,
+            "grad_norm": 2.022836208343506,
+            "learning_rate": 0.0005,
+            "epoch": 0.4044840587342226,
+            "step": 9025
+        },
+        {
+            "loss": 14.0746,
+            "grad_norm": 1.9047553539276123,
+            "learning_rate": 0.0005,
+            "epoch": 0.4047081496254881,
+            "step": 9030
+        },
+        {
+            "loss": 13.99,
+            "grad_norm": 1.864039421081543,
+            "learning_rate": 0.0005,
+            "epoch": 0.4049322405167536,
+            "step": 9035
+        },
+        {
+            "loss": 14.0763,
+            "grad_norm": 2.018603563308716,
+            "learning_rate": 0.0005,
+            "epoch": 0.4051563314080191,
+            "step": 9040
+        },
+        {
+            "loss": 14.058,
+            "grad_norm": 2.015758514404297,
+            "learning_rate": 0.0005,
+            "epoch": 0.4053804222992846,
+            "step": 9045
+        },
+        {
+            "loss": 14.0219,
+            "grad_norm": 2.26421856880188,
+            "learning_rate": 0.0005,
+            "epoch": 0.4056045131905501,
+            "step": 9050
+        },
+        {
+            "loss": 14.0153,
+            "grad_norm": 1.898167610168457,
+            "learning_rate": 0.0005,
+            "epoch": 0.4058286040818156,
+            "step": 9055
+        },
+        {
+            "loss": 14.1529,
+            "grad_norm": 1.7895253896713257,
+            "learning_rate": 0.0005,
+            "epoch": 0.4060526949730811,
+            "step": 9060
+        },
+        {
+            "loss": 14.077,
+            "grad_norm": 1.8098982572555542,
+            "learning_rate": 0.0005,
+            "epoch": 0.40627678586434657,
+            "step": 9065
+        },
+        {
+            "loss": 14.0183,
+            "grad_norm": 1.9741629362106323,
+            "learning_rate": 0.0005,
+            "epoch": 0.4065008767556121,
+            "step": 9070
+        },
+        {
+            "loss": 14.076,
+            "grad_norm": 1.8350106477737427,
+            "learning_rate": 0.0005,
+            "epoch": 0.4067249676468776,
+            "step": 9075
+        },
+        {
+            "loss": 14.0936,
+            "grad_norm": 1.9996545314788818,
+            "learning_rate": 0.0005,
+            "epoch": 0.40694905853814306,
+            "step": 9080
+        },
+        {
+            "loss": 14.126,
+            "grad_norm": 1.8608310222625732,
+            "learning_rate": 0.0005,
+            "epoch": 0.4071731494294086,
+            "step": 9085
+        },
+        {
+            "loss": 14.0375,
+            "grad_norm": 1.8878345489501953,
+            "learning_rate": 0.0005,
+            "epoch": 0.4073972403206741,
+            "step": 9090
+        },
+        {
+            "loss": 14.0464,
+            "grad_norm": 1.8385180234909058,
+            "learning_rate": 0.0005,
+            "epoch": 0.40762133121193955,
+            "step": 9095
+        },
+        {
+            "loss": 14.1311,
+            "grad_norm": 1.998307228088379,
+            "learning_rate": 0.0005,
+            "epoch": 0.40784542210320507,
+            "step": 9100
+        },
+        {
+            "loss": 14.1412,
+            "grad_norm": 1.788956642150879,
+            "learning_rate": 0.0005,
+            "epoch": 0.4080695129944706,
+            "step": 9105
+        },
+        {
+            "loss": 14.0221,
+            "grad_norm": 1.7810660600662231,
+            "learning_rate": 0.0005,
+            "epoch": 0.40829360388573605,
+            "step": 9110
+        },
+        {
+            "loss": 14.0332,
+            "grad_norm": 1.9653055667877197,
+            "learning_rate": 0.0005,
+            "epoch": 0.40851769477700156,
+            "step": 9115
+        },
+        {
+            "loss": 14.0218,
+            "grad_norm": 1.846150279045105,
+            "learning_rate": 0.0005,
+            "epoch": 0.408741785668267,
+            "step": 9120
+        },
+        {
+            "loss": 14.065,
+            "grad_norm": 1.925832748413086,
+            "learning_rate": 0.0005,
+            "epoch": 0.40896587655953254,
+            "step": 9125
+        },
+        {
+            "loss": 14.0653,
+            "grad_norm": 1.9436774253845215,
+            "learning_rate": 0.0005,
+            "epoch": 0.40918996745079805,
+            "step": 9130
+        },
+        {
+            "loss": 14.2366,
+            "grad_norm": 1.8719289302825928,
+            "learning_rate": 0.0005,
+            "epoch": 0.4094140583420635,
+            "step": 9135
+        },
+        {
+            "loss": 14.0781,
+            "grad_norm": 2.0129549503326416,
+            "learning_rate": 0.0005,
+            "epoch": 0.40963814923332903,
+            "step": 9140
+        },
+        {
+            "loss": 14.0747,
+            "grad_norm": 2.2035460472106934,
+            "learning_rate": 0.0005,
+            "epoch": 0.40986224012459455,
+            "step": 9145
+        },
+        {
+            "loss": 14.1699,
+            "grad_norm": 2.0379064083099365,
+            "learning_rate": 0.0005,
+            "epoch": 0.41008633101586,
+            "step": 9150
+        },
+        {
+            "loss": 13.9654,
+            "grad_norm": 1.9979286193847656,
+            "learning_rate": 0.0005,
+            "epoch": 0.4103104219071255,
+            "step": 9155
+        },
+        {
+            "loss": 13.9914,
+            "grad_norm": 1.9642233848571777,
+            "learning_rate": 0.0005,
+            "epoch": 0.41053451279839104,
+            "step": 9160
+        },
+        {
+            "loss": 13.9992,
+            "grad_norm": 2.0654077529907227,
+            "learning_rate": 0.0005,
+            "epoch": 0.4107586036896565,
+            "step": 9165
+        },
+        {
+            "loss": 14.094,
+            "grad_norm": 1.8470113277435303,
+            "learning_rate": 0.0005,
+            "epoch": 0.410982694580922,
+            "step": 9170
+        },
+        {
+            "loss": 14.1311,
+            "grad_norm": 1.7958095073699951,
+            "learning_rate": 0.0005,
+            "epoch": 0.41120678547218753,
+            "step": 9175
+        },
+        {
+            "loss": 14.0738,
+            "grad_norm": 1.732292890548706,
+            "learning_rate": 0.0005,
+            "epoch": 0.411430876363453,
+            "step": 9180
+        },
+        {
+            "loss": 14.0696,
+            "grad_norm": 1.9240189790725708,
+            "learning_rate": 0.0005,
+            "epoch": 0.4116549672547185,
+            "step": 9185
+        },
+        {
+            "loss": 14.0463,
+            "grad_norm": 1.9035232067108154,
+            "learning_rate": 0.0005,
+            "epoch": 0.411879058145984,
+            "step": 9190
+        },
+        {
+            "loss": 14.1022,
+            "grad_norm": 1.982646107673645,
+            "learning_rate": 0.0005,
+            "epoch": 0.4121031490372495,
+            "step": 9195
+        },
+        {
+            "loss": 13.9918,
+            "grad_norm": 1.9125901460647583,
+            "learning_rate": 0.0005,
+            "epoch": 0.412327239928515,
+            "step": 9200
+        },
+        {
+            "loss": 14.0336,
+            "grad_norm": 1.8618894815444946,
+            "learning_rate": 0.0005,
+            "epoch": 0.4125513308197805,
+            "step": 9205
+        },
+        {
+            "loss": 13.986,
+            "grad_norm": 2.1027495861053467,
+            "learning_rate": 0.0005,
+            "epoch": 0.412775421711046,
+            "step": 9210
+        },
+        {
+            "loss": 14.0505,
+            "grad_norm": 2.065297842025757,
+            "learning_rate": 0.0005,
+            "epoch": 0.4129995126023115,
+            "step": 9215
+        },
+        {
+            "loss": 14.0786,
+            "grad_norm": 1.935354471206665,
+            "learning_rate": 0.0005,
+            "epoch": 0.413223603493577,
+            "step": 9220
+        },
+        {
+            "loss": 14.1429,
+            "grad_norm": 2.1198596954345703,
+            "learning_rate": 0.0005,
+            "epoch": 0.41344769438484247,
+            "step": 9225
+        },
+        {
+            "loss": 14.1127,
+            "grad_norm": 2.406881093978882,
+            "learning_rate": 0.0005,
+            "epoch": 0.413671785276108,
+            "step": 9230
+        },
+        {
+            "loss": 14.0365,
+            "grad_norm": 2.0867960453033447,
+            "learning_rate": 0.0005,
+            "epoch": 0.4138958761673735,
+            "step": 9235
+        },
+        {
+            "loss": 14.0288,
+            "grad_norm": 1.940619945526123,
+            "learning_rate": 0.0005,
+            "epoch": 0.41411996705863896,
+            "step": 9240
+        },
+        {
+            "loss": 14.0551,
+            "grad_norm": 1.8351531028747559,
+            "learning_rate": 0.0005,
+            "epoch": 0.4143440579499045,
+            "step": 9245
+        },
+        {
+            "loss": 14.0954,
+            "grad_norm": 1.847739815711975,
+            "learning_rate": 0.0005,
+            "epoch": 0.41456814884117,
+            "step": 9250
+        },
+        {
+            "loss": 14.061,
+            "grad_norm": 2.023423910140991,
+            "learning_rate": 0.0005,
+            "epoch": 0.41479223973243545,
+            "step": 9255
+        },
+        {
+            "loss": 13.9158,
+            "grad_norm": 1.890186071395874,
+            "learning_rate": 0.0005,
+            "epoch": 0.41501633062370097,
+            "step": 9260
+        },
+        {
+            "loss": 14.1103,
+            "grad_norm": 1.8148173093795776,
+            "learning_rate": 0.0005,
+            "epoch": 0.4152404215149665,
+            "step": 9265
+        },
+        {
+            "loss": 14.0353,
+            "grad_norm": 1.978704571723938,
+            "learning_rate": 0.0005,
+            "epoch": 0.41546451240623195,
+            "step": 9270
+        },
+        {
+            "loss": 14.0829,
+            "grad_norm": 1.8345826864242554,
+            "learning_rate": 0.0005,
+            "epoch": 0.41568860329749746,
+            "step": 9275
+        },
+        {
+            "loss": 13.973,
+            "grad_norm": 1.8378230333328247,
+            "learning_rate": 0.0005,
+            "epoch": 0.415912694188763,
+            "step": 9280
+        },
+        {
+            "loss": 14.1264,
+            "grad_norm": 1.8736780881881714,
+            "learning_rate": 0.0005,
+            "epoch": 0.41613678508002844,
+            "step": 9285
+        },
+        {
+            "loss": 14.1078,
+            "grad_norm": 1.7875055074691772,
+            "learning_rate": 0.0005,
+            "epoch": 0.41636087597129395,
+            "step": 9290
+        },
+        {
+            "loss": 14.1918,
+            "grad_norm": 1.9262399673461914,
+            "learning_rate": 0.0005,
+            "epoch": 0.41658496686255947,
+            "step": 9295
+        },
+        {
+            "loss": 14.0808,
+            "grad_norm": 1.9088530540466309,
+            "learning_rate": 0.0005,
+            "epoch": 0.41680905775382493,
+            "step": 9300
+        },
+        {
+            "loss": 14.1056,
+            "grad_norm": 1.8706014156341553,
+            "learning_rate": 0.0005,
+            "epoch": 0.41703314864509045,
+            "step": 9305
+        },
+        {
+            "loss": 14.1663,
+            "grad_norm": 1.8975495100021362,
+            "learning_rate": 0.0005,
+            "epoch": 0.41725723953635596,
+            "step": 9310
+        },
+        {
+            "loss": 13.9967,
+            "grad_norm": 1.8865870237350464,
+            "learning_rate": 0.0005,
+            "epoch": 0.4174813304276214,
+            "step": 9315
+        },
+        {
+            "loss": 14.0276,
+            "grad_norm": 1.9450527429580688,
+            "learning_rate": 0.0005,
+            "epoch": 0.41770542131888694,
+            "step": 9320
+        },
+        {
+            "loss": 13.9744,
+            "grad_norm": 1.8535137176513672,
+            "learning_rate": 0.0005,
+            "epoch": 0.41792951221015245,
+            "step": 9325
+        },
+        {
+            "loss": 14.0797,
+            "grad_norm": 1.8707098960876465,
+            "learning_rate": 0.0005,
+            "epoch": 0.4181536031014179,
+            "step": 9330
+        },
+        {
+            "loss": 13.9541,
+            "grad_norm": 1.8629024028778076,
+            "learning_rate": 0.0005,
+            "epoch": 0.41837769399268343,
+            "step": 9335
+        },
+        {
+            "loss": 14.0776,
+            "grad_norm": 2.0101118087768555,
+            "learning_rate": 0.0005,
+            "epoch": 0.41860178488394895,
+            "step": 9340
+        },
+        {
+            "loss": 14.0206,
+            "grad_norm": 1.778620958328247,
+            "learning_rate": 0.0005,
+            "epoch": 0.4188258757752144,
+            "step": 9345
+        },
+        {
+            "loss": 14.0152,
+            "grad_norm": 1.9611141681671143,
+            "learning_rate": 0.0005,
+            "epoch": 0.4190499666664799,
+            "step": 9350
+        },
+        {
+            "loss": 13.9832,
+            "grad_norm": 1.933802604675293,
+            "learning_rate": 0.0005,
+            "epoch": 0.41927405755774544,
+            "step": 9355
+        },
+        {
+            "loss": 14.1467,
+            "grad_norm": 2.054326057434082,
+            "learning_rate": 0.0005,
+            "epoch": 0.4194981484490109,
+            "step": 9360
+        },
+        {
+            "loss": 13.9849,
+            "grad_norm": 1.8363392353057861,
+            "learning_rate": 0.0005,
+            "epoch": 0.4197222393402764,
+            "step": 9365
+        },
+        {
+            "loss": 14.1202,
+            "grad_norm": 1.9941134452819824,
+            "learning_rate": 0.0005,
+            "epoch": 0.41994633023154193,
+            "step": 9370
+        },
+        {
+            "loss": 14.1037,
+            "grad_norm": 2.0817792415618896,
+            "learning_rate": 0.0005,
+            "epoch": 0.4201704211228074,
+            "step": 9375
+        },
+        {
+            "loss": 14.0734,
+            "grad_norm": 1.8130356073379517,
+            "learning_rate": 0.0005,
+            "epoch": 0.4203945120140729,
+            "step": 9380
+        },
+        {
+            "loss": 14.036,
+            "grad_norm": 1.88438880443573,
+            "learning_rate": 0.0005,
+            "epoch": 0.4206186029053384,
+            "step": 9385
+        },
+        {
+            "loss": 14.0426,
+            "grad_norm": 1.95289146900177,
+            "learning_rate": 0.0005,
+            "epoch": 0.4208426937966039,
+            "step": 9390
+        },
+        {
+            "loss": 14.0527,
+            "grad_norm": 2.1133406162261963,
+            "learning_rate": 0.0005,
+            "epoch": 0.4210667846878694,
+            "step": 9395
+        },
+        {
+            "loss": 14.0653,
+            "grad_norm": 2.0192480087280273,
+            "learning_rate": 0.0005,
+            "epoch": 0.4212908755791349,
+            "step": 9400
+        },
+        {
+            "loss": 14.0586,
+            "grad_norm": 2.0362696647644043,
+            "learning_rate": 0.0005,
+            "epoch": 0.4215149664704004,
+            "step": 9405
+        },
+        {
+            "loss": 14.0897,
+            "grad_norm": 2.0646374225616455,
+            "learning_rate": 0.0005,
+            "epoch": 0.4217390573616659,
+            "step": 9410
+        },
+        {
+            "loss": 13.9908,
+            "grad_norm": 1.804474949836731,
+            "learning_rate": 0.0005,
+            "epoch": 0.4219631482529314,
+            "step": 9415
+        },
+        {
+            "loss": 13.9978,
+            "grad_norm": 1.795154333114624,
+            "learning_rate": 0.0005,
+            "epoch": 0.42218723914419687,
+            "step": 9420
+        },
+        {
+            "loss": 13.9795,
+            "grad_norm": 1.8122735023498535,
+            "learning_rate": 0.0005,
+            "epoch": 0.4224113300354624,
+            "step": 9425
+        },
+        {
+            "loss": 14.1127,
+            "grad_norm": 2.0129716396331787,
+            "learning_rate": 0.0005,
+            "epoch": 0.4226354209267279,
+            "step": 9430
+        },
+        {
+            "loss": 14.0821,
+            "grad_norm": 1.7567360401153564,
+            "learning_rate": 0.0005,
+            "epoch": 0.42285951181799336,
+            "step": 9435
+        },
+        {
+            "loss": 14.0475,
+            "grad_norm": 1.889049768447876,
+            "learning_rate": 0.0005,
+            "epoch": 0.4230836027092589,
+            "step": 9440
+        },
+        {
+            "loss": 14.0047,
+            "grad_norm": 1.8952374458312988,
+            "learning_rate": 0.0005,
+            "epoch": 0.4233076936005244,
+            "step": 9445
+        },
+        {
+            "loss": 14.0252,
+            "grad_norm": 2.077031373977661,
+            "learning_rate": 0.0005,
+            "epoch": 0.42353178449178985,
+            "step": 9450
+        },
+        {
+            "loss": 14.0336,
+            "grad_norm": 1.902687907218933,
+            "learning_rate": 0.0005,
+            "epoch": 0.42375587538305537,
+            "step": 9455
+        },
+        {
+            "loss": 14.0308,
+            "grad_norm": 1.8955004215240479,
+            "learning_rate": 0.0005,
+            "epoch": 0.4239799662743209,
+            "step": 9460
+        },
+        {
+            "loss": 14.1439,
+            "grad_norm": 1.7904077768325806,
+            "learning_rate": 0.0005,
+            "epoch": 0.42420405716558635,
+            "step": 9465
+        },
+        {
+            "loss": 14.0121,
+            "grad_norm": 1.9405328035354614,
+            "learning_rate": 0.0005,
+            "epoch": 0.42442814805685186,
+            "step": 9470
+        },
+        {
+            "loss": 14.0604,
+            "grad_norm": 2.001901388168335,
+            "learning_rate": 0.0005,
+            "epoch": 0.4246522389481174,
+            "step": 9475
+        },
+        {
+            "loss": 13.8987,
+            "grad_norm": 2.1401023864746094,
+            "learning_rate": 0.0005,
+            "epoch": 0.42487632983938284,
+            "step": 9480
+        },
+        {
+            "loss": 14.0958,
+            "grad_norm": 1.8659871816635132,
+            "learning_rate": 0.0005,
+            "epoch": 0.42510042073064835,
+            "step": 9485
+        },
+        {
+            "loss": 14.0518,
+            "grad_norm": 1.9418147802352905,
+            "learning_rate": 0.0005,
+            "epoch": 0.42532451162191387,
+            "step": 9490
+        },
+        {
+            "loss": 14.0263,
+            "grad_norm": 1.788070559501648,
+            "learning_rate": 0.0005,
+            "epoch": 0.42554860251317933,
+            "step": 9495
+        },
+        {
+            "loss": 14.0373,
+            "grad_norm": 1.8509469032287598,
+            "learning_rate": 0.0005,
+            "epoch": 0.42577269340444485,
+            "step": 9500
+        },
+        {
+            "eval_loss": 1.7514612674713135,
+            "eval_runtime": 18.448,
+            "eval_samples_per_second": 888.118,
+            "eval_steps_per_second": 7.968,
+            "epoch": 0.42577269340444485,
+            "step": 9500
+        },
+        {
+            "loss": 13.9943,
+            "grad_norm": 1.886093020439148,
+            "learning_rate": 0.0005,
+            "epoch": 0.42599678429571036,
+            "step": 9505
+        },
+        {
+            "loss": 14.1202,
+            "grad_norm": 2.0397934913635254,
+            "learning_rate": 0.0005,
+            "epoch": 0.4262208751869758,
+            "step": 9510
+        },
+        {
+            "loss": 14.0062,
+            "grad_norm": 2.0059173107147217,
+            "learning_rate": 0.0005,
+            "epoch": 0.42644496607824134,
+            "step": 9515
+        },
+        {
+            "loss": 14.1346,
+            "grad_norm": 2.0011186599731445,
+            "learning_rate": 0.0005,
+            "epoch": 0.42666905696950685,
+            "step": 9520
+        },
+        {
+            "loss": 14.1812,
+            "grad_norm": 1.9647142887115479,
+            "learning_rate": 0.0005,
+            "epoch": 0.4268931478607723,
+            "step": 9525
+        },
+        {
+            "loss": 14.2527,
+            "grad_norm": 2.0474014282226562,
+            "learning_rate": 0.0005,
+            "epoch": 0.42711723875203783,
+            "step": 9530
+        },
+        {
+            "loss": 14.0878,
+            "grad_norm": 2.0448601245880127,
+            "learning_rate": 0.0005,
+            "epoch": 0.42734132964330335,
+            "step": 9535
+        },
+        {
+            "loss": 14.0419,
+            "grad_norm": 1.9019169807434082,
+            "learning_rate": 0.0005,
+            "epoch": 0.4275654205345688,
+            "step": 9540
+        },
+        {
+            "loss": 14.0307,
+            "grad_norm": 1.9999176263809204,
+            "learning_rate": 0.0005,
+            "epoch": 0.4277895114258343,
+            "step": 9545
+        },
+        {
+            "loss": 14.1149,
+            "grad_norm": 2.0039052963256836,
+            "learning_rate": 0.0005,
+            "epoch": 0.42801360231709984,
+            "step": 9550
+        },
+        {
+            "loss": 14.0169,
+            "grad_norm": 1.9922994375228882,
+            "learning_rate": 0.0005,
+            "epoch": 0.4282376932083653,
+            "step": 9555
+        },
+        {
+            "loss": 14.1576,
+            "grad_norm": 2.1537740230560303,
+            "learning_rate": 0.0005,
+            "epoch": 0.4284617840996308,
+            "step": 9560
+        },
+        {
+            "loss": 14.0213,
+            "grad_norm": 1.8866078853607178,
+            "learning_rate": 0.0005,
+            "epoch": 0.42868587499089633,
+            "step": 9565
+        },
+        {
+            "loss": 14.1185,
+            "grad_norm": 1.8932602405548096,
+            "learning_rate": 0.0005,
+            "epoch": 0.4289099658821618,
+            "step": 9570
+        },
+        {
+            "loss": 14.0622,
+            "grad_norm": 1.8032782077789307,
+            "learning_rate": 0.0005,
+            "epoch": 0.4291340567734273,
+            "step": 9575
+        },
+        {
+            "loss": 14.0255,
+            "grad_norm": 1.8393731117248535,
+            "learning_rate": 0.0005,
+            "epoch": 0.4293581476646928,
+            "step": 9580
+        },
+        {
+            "loss": 14.048,
+            "grad_norm": 1.914337396621704,
+            "learning_rate": 0.0005,
+            "epoch": 0.4295822385559583,
+            "step": 9585
+        },
+        {
+            "loss": 14.0042,
+            "grad_norm": 1.8856425285339355,
+            "learning_rate": 0.0005,
+            "epoch": 0.4298063294472238,
+            "step": 9590
+        },
+        {
+            "loss": 14.1046,
+            "grad_norm": 1.7746888399124146,
+            "learning_rate": 0.0005,
+            "epoch": 0.4300304203384893,
+            "step": 9595
+        },
+        {
+            "loss": 14.0323,
+            "grad_norm": 1.9397006034851074,
+            "learning_rate": 0.0005,
+            "epoch": 0.4302545112297548,
+            "step": 9600
+        },
+        {
+            "loss": 14.0858,
+            "grad_norm": 1.9266762733459473,
+            "learning_rate": 0.0005,
+            "epoch": 0.4304786021210203,
+            "step": 9605
+        },
+        {
+            "loss": 14.0142,
+            "grad_norm": 1.9198224544525146,
+            "learning_rate": 0.0005,
+            "epoch": 0.4307026930122858,
+            "step": 9610
+        },
+        {
+            "loss": 14.0901,
+            "grad_norm": 1.9672712087631226,
+            "learning_rate": 0.0005,
+            "epoch": 0.43092678390355127,
+            "step": 9615
+        },
+        {
+            "loss": 13.9971,
+            "grad_norm": 1.9564368724822998,
+            "learning_rate": 0.0005,
+            "epoch": 0.4311508747948168,
+            "step": 9620
+        },
+        {
+            "loss": 14.0168,
+            "grad_norm": 1.7687296867370605,
+            "learning_rate": 0.0005,
+            "epoch": 0.4313749656860823,
+            "step": 9625
+        },
+        {
+            "loss": 14.0594,
+            "grad_norm": 1.8282158374786377,
+            "learning_rate": 0.0005,
+            "epoch": 0.43159905657734776,
+            "step": 9630
+        },
+        {
+            "loss": 13.99,
+            "grad_norm": 1.986424207687378,
+            "learning_rate": 0.0005,
+            "epoch": 0.4318231474686133,
+            "step": 9635
+        },
+        {
+            "loss": 14.0392,
+            "grad_norm": 1.755864143371582,
+            "learning_rate": 0.0005,
+            "epoch": 0.4320472383598788,
+            "step": 9640
+        },
+        {
+            "loss": 14.1816,
+            "grad_norm": 1.9651696681976318,
+            "learning_rate": 0.0005,
+            "epoch": 0.43227132925114425,
+            "step": 9645
+        },
+        {
+            "loss": 14.1242,
+            "grad_norm": 1.9259588718414307,
+            "learning_rate": 0.0005,
+            "epoch": 0.43249542014240977,
+            "step": 9650
+        },
+        {
+            "loss": 14.0321,
+            "grad_norm": 2.0099666118621826,
+            "learning_rate": 0.0005,
+            "epoch": 0.4327195110336753,
+            "step": 9655
+        },
+        {
+            "loss": 14.1886,
+            "grad_norm": 2.332737922668457,
+            "learning_rate": 0.0005,
+            "epoch": 0.43294360192494075,
+            "step": 9660
+        },
+        {
+            "loss": 14.0657,
+            "grad_norm": 2.220191240310669,
+            "learning_rate": 0.0005,
+            "epoch": 0.43316769281620626,
+            "step": 9665
+        },
+        {
+            "loss": 14.046,
+            "grad_norm": 1.9909117221832275,
+            "learning_rate": 0.0005,
+            "epoch": 0.4333917837074718,
+            "step": 9670
+        },
+        {
+            "loss": 14.0047,
+            "grad_norm": 1.9637538194656372,
+            "learning_rate": 0.0005,
+            "epoch": 0.43361587459873724,
+            "step": 9675
+        },
+        {
+            "loss": 14.104,
+            "grad_norm": 2.0495548248291016,
+            "learning_rate": 0.0005,
+            "epoch": 0.43383996549000275,
+            "step": 9680
+        },
+        {
+            "loss": 14.0572,
+            "grad_norm": 2.065443277359009,
+            "learning_rate": 0.0005,
+            "epoch": 0.4340640563812682,
+            "step": 9685
+        },
+        {
+            "loss": 13.9782,
+            "grad_norm": 1.896341323852539,
+            "learning_rate": 0.0005,
+            "epoch": 0.43428814727253373,
+            "step": 9690
+        },
+        {
+            "loss": 14.0158,
+            "grad_norm": 2.0682294368743896,
+            "learning_rate": 0.0005,
+            "epoch": 0.43451223816379925,
+            "step": 9695
+        },
+        {
+            "loss": 14.0686,
+            "grad_norm": 2.1541178226470947,
+            "learning_rate": 0.0005,
+            "epoch": 0.4347363290550647,
+            "step": 9700
+        },
+        {
+            "loss": 14.111,
+            "grad_norm": 2.1676082611083984,
+            "learning_rate": 0.0005,
+            "epoch": 0.4349604199463302,
+            "step": 9705
+        },
+        {
+            "loss": 14.0887,
+            "grad_norm": 2.198476552963257,
+            "learning_rate": 0.0005,
+            "epoch": 0.43518451083759574,
+            "step": 9710
+        },
+        {
+            "loss": 14.0078,
+            "grad_norm": 1.7735761404037476,
+            "learning_rate": 0.0005,
+            "epoch": 0.4354086017288612,
+            "step": 9715
+        },
+        {
+            "loss": 14.0178,
+            "grad_norm": 2.0045340061187744,
+            "learning_rate": 0.0005,
+            "epoch": 0.4356326926201267,
+            "step": 9720
+        },
+        {
+            "loss": 14.0358,
+            "grad_norm": 1.8827794790267944,
+            "learning_rate": 0.0005,
+            "epoch": 0.43585678351139223,
+            "step": 9725
+        },
+        {
+            "loss": 14.0842,
+            "grad_norm": 1.7653541564941406,
+            "learning_rate": 0.0005,
+            "epoch": 0.4360808744026577,
+            "step": 9730
+        },
+        {
+            "loss": 14.129,
+            "grad_norm": 1.8109867572784424,
+            "learning_rate": 0.0005,
+            "epoch": 0.4363049652939232,
+            "step": 9735
+        },
+        {
+            "loss": 14.0635,
+            "grad_norm": 1.8981984853744507,
+            "learning_rate": 0.0005,
+            "epoch": 0.4365290561851887,
+            "step": 9740
+        },
+        {
+            "loss": 13.9951,
+            "grad_norm": 1.8760943412780762,
+            "learning_rate": 0.0005,
+            "epoch": 0.4367531470764542,
+            "step": 9745
+        },
+        {
+            "loss": 14.0403,
+            "grad_norm": 1.8407485485076904,
+            "learning_rate": 0.0005,
+            "epoch": 0.4369772379677197,
+            "step": 9750
+        },
+        {
+            "loss": 13.9382,
+            "grad_norm": 1.9359339475631714,
+            "learning_rate": 0.0005,
+            "epoch": 0.4372013288589852,
+            "step": 9755
+        },
+        {
+            "loss": 14.0912,
+            "grad_norm": 1.8769389390945435,
+            "learning_rate": 0.0005,
+            "epoch": 0.4374254197502507,
+            "step": 9760
+        },
+        {
+            "loss": 13.9935,
+            "grad_norm": 1.6882269382476807,
+            "learning_rate": 0.0005,
+            "epoch": 0.4376495106415162,
+            "step": 9765
+        },
+        {
+            "loss": 14.0368,
+            "grad_norm": 1.8080883026123047,
+            "learning_rate": 0.0005,
+            "epoch": 0.4378736015327817,
+            "step": 9770
+        },
+        {
+            "loss": 13.9275,
+            "grad_norm": 1.9042779207229614,
+            "learning_rate": 0.0005,
+            "epoch": 0.43809769242404717,
+            "step": 9775
+        },
+        {
+            "loss": 14.1232,
+            "grad_norm": 1.9694958925247192,
+            "learning_rate": 0.0005,
+            "epoch": 0.4383217833153127,
+            "step": 9780
+        },
+        {
+            "loss": 14.0069,
+            "grad_norm": 1.853636384010315,
+            "learning_rate": 0.0005,
+            "epoch": 0.4385458742065782,
+            "step": 9785
+        },
+        {
+            "loss": 14.0662,
+            "grad_norm": 1.9663853645324707,
+            "learning_rate": 0.0005,
+            "epoch": 0.43876996509784366,
+            "step": 9790
+        },
+        {
+            "loss": 14.0328,
+            "grad_norm": 2.089175224304199,
+            "learning_rate": 0.0005,
+            "epoch": 0.4389940559891092,
+            "step": 9795
+        },
+        {
+            "loss": 14.0414,
+            "grad_norm": 1.8818589448928833,
+            "learning_rate": 0.0005,
+            "epoch": 0.4392181468803747,
+            "step": 9800
+        },
+        {
+            "loss": 14.0658,
+            "grad_norm": 1.8281285762786865,
+            "learning_rate": 0.0005,
+            "epoch": 0.43944223777164015,
+            "step": 9805
+        },
+        {
+            "loss": 14.1294,
+            "grad_norm": 2.1331112384796143,
+            "learning_rate": 0.0005,
+            "epoch": 0.43966632866290567,
+            "step": 9810
+        },
+        {
+            "loss": 14.1253,
+            "grad_norm": 2.412768602371216,
+            "learning_rate": 0.0005,
+            "epoch": 0.4398904195541712,
+            "step": 9815
+        },
+        {
+            "loss": 14.0673,
+            "grad_norm": 1.8944461345672607,
+            "learning_rate": 0.0005,
+            "epoch": 0.44011451044543665,
+            "step": 9820
+        },
+        {
+            "loss": 14.0721,
+            "grad_norm": 1.772596836090088,
+            "learning_rate": 0.0005,
+            "epoch": 0.44033860133670216,
+            "step": 9825
+        },
+        {
+            "loss": 13.9508,
+            "grad_norm": 1.7824463844299316,
+            "learning_rate": 0.0005,
+            "epoch": 0.4405626922279677,
+            "step": 9830
+        },
+        {
+            "loss": 14.0619,
+            "grad_norm": 1.9199808835983276,
+            "learning_rate": 0.0005,
+            "epoch": 0.44078678311923314,
+            "step": 9835
+        },
+        {
+            "loss": 13.9435,
+            "grad_norm": 1.7367388010025024,
+            "learning_rate": 0.0005,
+            "epoch": 0.44101087401049865,
+            "step": 9840
+        },
+        {
+            "loss": 14.0238,
+            "grad_norm": 1.9348160028457642,
+            "learning_rate": 0.0005,
+            "epoch": 0.44123496490176417,
+            "step": 9845
+        },
+        {
+            "loss": 14.1016,
+            "grad_norm": 1.9653925895690918,
+            "learning_rate": 0.0005,
+            "epoch": 0.44145905579302963,
+            "step": 9850
+        },
+        {
+            "loss": 14.0995,
+            "grad_norm": 1.7807902097702026,
+            "learning_rate": 0.0005,
+            "epoch": 0.44168314668429515,
+            "step": 9855
+        },
+        {
+            "loss": 14.0309,
+            "grad_norm": 1.9768279790878296,
+            "learning_rate": 0.0005,
+            "epoch": 0.44190723757556066,
+            "step": 9860
+        },
+        {
+            "loss": 14.026,
+            "grad_norm": 1.8459382057189941,
+            "learning_rate": 0.0005,
+            "epoch": 0.4421313284668261,
+            "step": 9865
+        },
+        {
+            "loss": 13.9491,
+            "grad_norm": 1.765913963317871,
+            "learning_rate": 0.0005,
+            "epoch": 0.44235541935809164,
+            "step": 9870
+        },
+        {
+            "loss": 14.0261,
+            "grad_norm": 1.9289934635162354,
+            "learning_rate": 0.0005,
+            "epoch": 0.44257951024935716,
+            "step": 9875
+        },
+        {
+            "loss": 14.0231,
+            "grad_norm": 2.0702064037323,
+            "learning_rate": 0.0005,
+            "epoch": 0.4428036011406226,
+            "step": 9880
+        },
+        {
+            "loss": 14.0428,
+            "grad_norm": 1.9474433660507202,
+            "learning_rate": 0.0005,
+            "epoch": 0.44302769203188813,
+            "step": 9885
+        },
+        {
+            "loss": 14.0147,
+            "grad_norm": 1.930172085762024,
+            "learning_rate": 0.0005,
+            "epoch": 0.44325178292315365,
+            "step": 9890
+        },
+        {
+            "loss": 14.0694,
+            "grad_norm": 2.0727486610412598,
+            "learning_rate": 0.0005,
+            "epoch": 0.4434758738144191,
+            "step": 9895
+        },
+        {
+            "loss": 14.1825,
+            "grad_norm": 1.8144179582595825,
+            "learning_rate": 0.0005,
+            "epoch": 0.4436999647056846,
+            "step": 9900
+        },
+        {
+            "loss": 13.9797,
+            "grad_norm": 1.8270831108093262,
+            "learning_rate": 0.0005,
+            "epoch": 0.44392405559695014,
+            "step": 9905
+        },
+        {
+            "loss": 14.0102,
+            "grad_norm": 1.9958640336990356,
+            "learning_rate": 0.0005,
+            "epoch": 0.4441481464882156,
+            "step": 9910
+        },
+        {
+            "loss": 14.0815,
+            "grad_norm": 1.8873703479766846,
+            "learning_rate": 0.0005,
+            "epoch": 0.4443722373794811,
+            "step": 9915
+        },
+        {
+            "loss": 13.9492,
+            "grad_norm": 2.3722341060638428,
+            "learning_rate": 0.0005,
+            "epoch": 0.44459632827074663,
+            "step": 9920
+        },
+        {
+            "loss": 14.0492,
+            "grad_norm": 1.8521634340286255,
+            "learning_rate": 0.0005,
+            "epoch": 0.4448204191620121,
+            "step": 9925
+        },
+        {
+            "loss": 14.1128,
+            "grad_norm": 1.9003937244415283,
+            "learning_rate": 0.0005,
+            "epoch": 0.4450445100532776,
+            "step": 9930
+        },
+        {
+            "loss": 14.0259,
+            "grad_norm": 1.7331500053405762,
+            "learning_rate": 0.0005,
+            "epoch": 0.4452686009445431,
+            "step": 9935
+        },
+        {
+            "loss": 13.979,
+            "grad_norm": 1.9712514877319336,
+            "learning_rate": 0.0005,
+            "epoch": 0.4454926918358086,
+            "step": 9940
+        },
+        {
+            "loss": 14.0766,
+            "grad_norm": 1.9684165716171265,
+            "learning_rate": 0.0005,
+            "epoch": 0.4457167827270741,
+            "step": 9945
+        },
+        {
+            "loss": 14.0004,
+            "grad_norm": 1.961761474609375,
+            "learning_rate": 0.0005,
+            "epoch": 0.4459408736183396,
+            "step": 9950
+        },
+        {
+            "loss": 14.1178,
+            "grad_norm": 2.1732451915740967,
+            "learning_rate": 0.0005,
+            "epoch": 0.4461649645096051,
+            "step": 9955
+        },
+        {
+            "loss": 14.0609,
+            "grad_norm": 1.9633890390396118,
+            "learning_rate": 0.0005,
+            "epoch": 0.4463890554008706,
+            "step": 9960
+        },
+        {
+            "loss": 14.0079,
+            "grad_norm": 1.9380450248718262,
+            "learning_rate": 0.0005,
+            "epoch": 0.4466131462921361,
+            "step": 9965
+        },
+        {
+            "loss": 14.1057,
+            "grad_norm": 2.074018955230713,
+            "learning_rate": 0.0005,
+            "epoch": 0.44683723718340157,
+            "step": 9970
+        },
+        {
+            "loss": 14.1039,
+            "grad_norm": 1.8502141237258911,
+            "learning_rate": 0.0005,
+            "epoch": 0.4470613280746671,
+            "step": 9975
+        },
+        {
+            "loss": 14.0487,
+            "grad_norm": 2.16658616065979,
+            "learning_rate": 0.0005,
+            "epoch": 0.4472854189659326,
+            "step": 9980
+        },
+        {
+            "loss": 13.9841,
+            "grad_norm": 1.9512484073638916,
+            "learning_rate": 0.0005,
+            "epoch": 0.44750950985719806,
+            "step": 9985
+        },
+        {
+            "loss": 14.0446,
+            "grad_norm": 1.728674292564392,
+            "learning_rate": 0.0005,
+            "epoch": 0.4477336007484636,
+            "step": 9990
+        },
+        {
+            "loss": 13.9403,
+            "grad_norm": 1.8797376155853271,
+            "learning_rate": 0.0005,
+            "epoch": 0.4479576916397291,
+            "step": 9995
+        },
+        {
+            "loss": 13.9627,
+            "grad_norm": 1.7768317461013794,
+            "learning_rate": 0.0005,
+            "epoch": 0.44818178253099455,
+            "step": 10000
+        },
+        {
+            "eval_loss": 1.7535099983215332,
+            "eval_runtime": 18.6103,
+            "eval_samples_per_second": 880.37,
+            "eval_steps_per_second": 7.899,
+            "epoch": 0.44818178253099455,
+            "step": 10000
+        },
+        {
+            "loss": 14.0345,
+            "grad_norm": 1.8116319179534912,
+            "learning_rate": 0.0005,
+            "epoch": 0.44840587342226007,
+            "step": 10005
+        },
+        {
+            "loss": 13.9768,
+            "grad_norm": 1.941550612449646,
+            "learning_rate": 0.0005,
+            "epoch": 0.4486299643135256,
+            "step": 10010
+        },
+        {
+            "loss": 13.9213,
+            "grad_norm": 2.029679298400879,
+            "learning_rate": 0.0005,
+            "epoch": 0.44885405520479105,
+            "step": 10015
+        },
+        {
+            "loss": 14.0488,
+            "grad_norm": 1.9416868686676025,
+            "learning_rate": 0.0005,
+            "epoch": 0.44907814609605656,
+            "step": 10020
+        },
+        {
+            "loss": 13.9741,
+            "grad_norm": 1.7495111227035522,
+            "learning_rate": 0.0005,
+            "epoch": 0.4493022369873221,
+            "step": 10025
+        },
+        {
+            "loss": 14.0472,
+            "grad_norm": 1.8708152770996094,
+            "learning_rate": 0.0005,
+            "epoch": 0.44952632787858754,
+            "step": 10030
+        },
+        {
+            "loss": 14.0744,
+            "grad_norm": 1.8809341192245483,
+            "learning_rate": 0.0005,
+            "epoch": 0.44975041876985306,
+            "step": 10035
+        },
+        {
+            "loss": 14.1143,
+            "grad_norm": 1.8372942209243774,
+            "learning_rate": 0.0005,
+            "epoch": 0.44997450966111857,
+            "step": 10040
+        },
+        {
+            "loss": 13.9523,
+            "grad_norm": 1.7713371515274048,
+            "learning_rate": 0.0005,
+            "epoch": 0.45019860055238403,
+            "step": 10045
+        },
+        {
+            "loss": 14.0756,
+            "grad_norm": 1.8108537197113037,
+            "learning_rate": 0.0005,
+            "epoch": 0.45042269144364955,
+            "step": 10050
+        },
+        {
+            "loss": 13.9517,
+            "grad_norm": 1.839207410812378,
+            "learning_rate": 0.0005,
+            "epoch": 0.45064678233491506,
+            "step": 10055
+        },
+        {
+            "loss": 14.0217,
+            "grad_norm": 2.060572385787964,
+            "learning_rate": 0.0005,
+            "epoch": 0.4508708732261805,
+            "step": 10060
+        },
+        {
+            "loss": 14.0306,
+            "grad_norm": 1.7858555316925049,
+            "learning_rate": 0.0005,
+            "epoch": 0.45109496411744604,
+            "step": 10065
+        },
+        {
+            "loss": 14.0025,
+            "grad_norm": 1.8673533201217651,
+            "learning_rate": 0.0005,
+            "epoch": 0.45131905500871156,
+            "step": 10070
+        },
+        {
+            "loss": 14.0783,
+            "grad_norm": 2.0727386474609375,
+            "learning_rate": 0.0005,
+            "epoch": 0.451543145899977,
+            "step": 10075
+        },
+        {
+            "loss": 14.042,
+            "grad_norm": 1.91311514377594,
+            "learning_rate": 0.0005,
+            "epoch": 0.45176723679124253,
+            "step": 10080
+        },
+        {
+            "loss": 14.0383,
+            "grad_norm": 1.808239221572876,
+            "learning_rate": 0.0005,
+            "epoch": 0.45199132768250805,
+            "step": 10085
+        },
+        {
+            "loss": 14.0227,
+            "grad_norm": 1.7587231397628784,
+            "learning_rate": 0.0005,
+            "epoch": 0.4522154185737735,
+            "step": 10090
+        },
+        {
+            "loss": 14.0389,
+            "grad_norm": 1.8921979665756226,
+            "learning_rate": 0.0005,
+            "epoch": 0.452439509465039,
+            "step": 10095
+        },
+        {
+            "loss": 14.0514,
+            "grad_norm": 4.310143947601318,
+            "learning_rate": 0.0005,
+            "epoch": 0.45266360035630454,
+            "step": 10100
+        },
+        {
+            "loss": 14.6969,
+            "grad_norm": 64.53416442871094,
+            "learning_rate": 0.0005,
+            "epoch": 0.45288769124757,
+            "step": 10105
+        },
+        {
+            "loss": 14.4036,
+            "grad_norm": 2.269564628601074,
+            "learning_rate": 0.0005,
+            "epoch": 0.4531117821388355,
+            "step": 10110
+        },
+        {
+            "loss": 14.1395,
+            "grad_norm": 29.50452995300293,
+            "learning_rate": 0.0005,
+            "epoch": 0.45333587303010103,
+            "step": 10115
+        },
+        {
+            "loss": 14.1476,
+            "grad_norm": 2.0817313194274902,
+            "learning_rate": 0.0005,
+            "epoch": 0.4535599639213665,
+            "step": 10120
+        },
+        {
+            "loss": 13.938,
+            "grad_norm": 2.068413257598877,
+            "learning_rate": 0.0005,
+            "epoch": 0.453784054812632,
+            "step": 10125
+        },
+        {
+            "loss": 14.0095,
+            "grad_norm": 1.9635816812515259,
+            "learning_rate": 0.0005,
+            "epoch": 0.4540081457038975,
+            "step": 10130
+        },
+        {
+            "loss": 13.9837,
+            "grad_norm": 1.942551612854004,
+            "learning_rate": 0.0005,
+            "epoch": 0.454232236595163,
+            "step": 10135
+        },
+        {
+            "loss": 14.0523,
+            "grad_norm": 84.11277770996094,
+            "learning_rate": 0.0005,
+            "epoch": 0.4544563274864285,
+            "step": 10140
+        },
+        {
+            "loss": 13.989,
+            "grad_norm": 2.137803792953491,
+            "learning_rate": 0.0005,
+            "epoch": 0.454680418377694,
+            "step": 10145
+        },
+        {
+            "loss": 14.1064,
+            "grad_norm": 2.055637836456299,
+            "learning_rate": 0.0005,
+            "epoch": 0.4549045092689595,
+            "step": 10150
+        },
+        {
+            "loss": 14.0542,
+            "grad_norm": 1.9040488004684448,
+            "learning_rate": 0.0005,
+            "epoch": 0.455128600160225,
+            "step": 10155
+        },
+        {
+            "loss": 14.0577,
+            "grad_norm": 10.268532752990723,
+            "learning_rate": 0.0005,
+            "epoch": 0.4553526910514905,
+            "step": 10160
+        },
+        {
+            "loss": 14.1002,
+            "grad_norm": 1.9174935817718506,
+            "learning_rate": 0.0005,
+            "epoch": 0.45557678194275597,
+            "step": 10165
+        },
+        {
+            "loss": 14.0308,
+            "grad_norm": 1.8861088752746582,
+            "learning_rate": 0.0005,
+            "epoch": 0.4558008728340215,
+            "step": 10170
+        },
+        {
+            "loss": 14.0439,
+            "grad_norm": 2.4005439281463623,
+            "learning_rate": 0.0005,
+            "epoch": 0.456024963725287,
+            "step": 10175
+        },
+        {
+            "loss": 14.038,
+            "grad_norm": 1.8772950172424316,
+            "learning_rate": 0.0005,
+            "epoch": 0.45624905461655246,
+            "step": 10180
+        },
+        {
+            "loss": 13.9871,
+            "grad_norm": 1.927895426750183,
+            "learning_rate": 0.0005,
+            "epoch": 0.456473145507818,
+            "step": 10185
+        },
+        {
+            "loss": 13.98,
+            "grad_norm": 1.8765352964401245,
+            "learning_rate": 0.0005,
+            "epoch": 0.4566972363990835,
+            "step": 10190
+        },
+        {
+            "loss": 14.1258,
+            "grad_norm": 1.8601654767990112,
+            "learning_rate": 0.0005,
+            "epoch": 0.45692132729034896,
+            "step": 10195
+        },
+        {
+            "loss": 13.9915,
+            "grad_norm": 1.9006295204162598,
+            "learning_rate": 0.0005,
+            "epoch": 0.45714541818161447,
+            "step": 10200
+        },
+        {
+            "loss": 14.0611,
+            "grad_norm": 1.9740972518920898,
+            "learning_rate": 0.0005,
+            "epoch": 0.45736950907288,
+            "step": 10205
+        },
+        {
+            "loss": 14.055,
+            "grad_norm": 1.8645578622817993,
+            "learning_rate": 0.0005,
+            "epoch": 0.45759359996414545,
+            "step": 10210
+        },
+        {
+            "loss": 14.0846,
+            "grad_norm": 1.9724942445755005,
+            "learning_rate": 0.0005,
+            "epoch": 0.45781769085541096,
+            "step": 10215
+        },
+        {
+            "loss": 14.0726,
+            "grad_norm": 2.019284725189209,
+            "learning_rate": 0.0005,
+            "epoch": 0.4580417817466765,
+            "step": 10220
+        },
+        {
+            "loss": 14.0259,
+            "grad_norm": 2.0942375659942627,
+            "learning_rate": 0.0005,
+            "epoch": 0.45826587263794194,
+            "step": 10225
+        },
+        {
+            "loss": 14.0304,
+            "grad_norm": 2.0027451515197754,
+            "learning_rate": 0.0005,
+            "epoch": 0.45848996352920746,
+            "step": 10230
+        },
+        {
+            "loss": 13.9557,
+            "grad_norm": 1.8633819818496704,
+            "learning_rate": 0.0005,
+            "epoch": 0.4587140544204729,
+            "step": 10235
+        },
+        {
+            "loss": 14.113,
+            "grad_norm": 1.8201923370361328,
+            "learning_rate": 0.0005,
+            "epoch": 0.45893814531173843,
+            "step": 10240
+        },
+        {
+            "loss": 14.0944,
+            "grad_norm": 1.6903871297836304,
+            "learning_rate": 0.0005,
+            "epoch": 0.45916223620300395,
+            "step": 10245
+        },
+        {
+            "loss": 14.0911,
+            "grad_norm": 2.0811848640441895,
+            "learning_rate": 0.0005,
+            "epoch": 0.4593863270942694,
+            "step": 10250
+        },
+        {
+            "loss": 14.0192,
+            "grad_norm": 2.03033447265625,
+            "learning_rate": 0.0005,
+            "epoch": 0.4596104179855349,
+            "step": 10255
+        },
+        {
+            "loss": 14.0824,
+            "grad_norm": 2.0053560733795166,
+            "learning_rate": 0.0005,
+            "epoch": 0.45983450887680044,
+            "step": 10260
+        },
+        {
+            "loss": 14.1325,
+            "grad_norm": 1.9983246326446533,
+            "learning_rate": 0.0005,
+            "epoch": 0.4600585997680659,
+            "step": 10265
+        },
+        {
+            "loss": 14.0628,
+            "grad_norm": 1.8627090454101562,
+            "learning_rate": 0.0005,
+            "epoch": 0.4602826906593314,
+            "step": 10270
+        },
+        {
+            "loss": 14.0698,
+            "grad_norm": 1.819238305091858,
+            "learning_rate": 0.0005,
+            "epoch": 0.46050678155059693,
+            "step": 10275
+        },
+        {
+            "loss": 14.087,
+            "grad_norm": 1.8443467617034912,
+            "learning_rate": 0.0005,
+            "epoch": 0.4607308724418624,
+            "step": 10280
+        },
+        {
+            "loss": 14.0054,
+            "grad_norm": 1.9102563858032227,
+            "learning_rate": 0.0005,
+            "epoch": 0.4609549633331279,
+            "step": 10285
+        },
+        {
+            "loss": 13.9632,
+            "grad_norm": 1.9677178859710693,
+            "learning_rate": 0.0005,
+            "epoch": 0.4611790542243934,
+            "step": 10290
+        },
+        {
+            "loss": 14.105,
+            "grad_norm": 1.943468451499939,
+            "learning_rate": 0.0005,
+            "epoch": 0.4614031451156589,
+            "step": 10295
+        },
+        {
+            "loss": 14.0248,
+            "grad_norm": 1.8622480630874634,
+            "learning_rate": 0.0005,
+            "epoch": 0.4616272360069244,
+            "step": 10300
+        },
+        {
+            "loss": 14.0241,
+            "grad_norm": 1.9953628778457642,
+            "learning_rate": 0.0005,
+            "epoch": 0.4618513268981899,
+            "step": 10305
+        },
+        {
+            "loss": 14.1536,
+            "grad_norm": 2.2701895236968994,
+            "learning_rate": 0.0005,
+            "epoch": 0.4620754177894554,
+            "step": 10310
+        },
+        {
+            "loss": 14.0628,
+            "grad_norm": 2.028715133666992,
+            "learning_rate": 0.0005,
+            "epoch": 0.4622995086807209,
+            "step": 10315
+        },
+        {
+            "loss": 14.0113,
+            "grad_norm": 1.8300288915634155,
+            "learning_rate": 0.0005,
+            "epoch": 0.4625235995719864,
+            "step": 10320
+        },
+        {
+            "loss": 14.1851,
+            "grad_norm": 1.9454532861709595,
+            "learning_rate": 0.0005,
+            "epoch": 0.46274769046325187,
+            "step": 10325
+        },
+        {
+            "loss": 14.098,
+            "grad_norm": 2.0312724113464355,
+            "learning_rate": 0.0005,
+            "epoch": 0.4629717813545174,
+            "step": 10330
+        },
+        {
+            "loss": 14.0318,
+            "grad_norm": 1.91213858127594,
+            "learning_rate": 0.0005,
+            "epoch": 0.4631958722457829,
+            "step": 10335
+        },
+        {
+            "loss": 14.0608,
+            "grad_norm": 1.9075995683670044,
+            "learning_rate": 0.0005,
+            "epoch": 0.46341996313704836,
+            "step": 10340
+        },
+        {
+            "loss": 13.9737,
+            "grad_norm": 1.7976598739624023,
+            "learning_rate": 0.0005,
+            "epoch": 0.4636440540283139,
+            "step": 10345
+        },
+        {
+            "loss": 14.0832,
+            "grad_norm": 1.8574029207229614,
+            "learning_rate": 0.0005,
+            "epoch": 0.4638681449195794,
+            "step": 10350
+        },
+        {
+            "loss": 13.9781,
+            "grad_norm": 1.830602765083313,
+            "learning_rate": 0.0005,
+            "epoch": 0.46409223581084486,
+            "step": 10355
+        },
+        {
+            "loss": 14.0166,
+            "grad_norm": 1.8759225606918335,
+            "learning_rate": 0.0005,
+            "epoch": 0.46431632670211037,
+            "step": 10360
+        },
+        {
+            "loss": 14.0674,
+            "grad_norm": 2.1599314212799072,
+            "learning_rate": 0.0005,
+            "epoch": 0.4645404175933759,
+            "step": 10365
+        },
+        {
+            "loss": 14.0889,
+            "grad_norm": 2.089174270629883,
+            "learning_rate": 0.0005,
+            "epoch": 0.46476450848464135,
+            "step": 10370
+        },
+        {
+            "loss": 13.9817,
+            "grad_norm": 1.879372239112854,
+            "learning_rate": 0.0005,
+            "epoch": 0.46498859937590686,
+            "step": 10375
+        },
+        {
+            "loss": 13.9842,
+            "grad_norm": 1.8365589380264282,
+            "learning_rate": 0.0005,
+            "epoch": 0.4652126902671724,
+            "step": 10380
+        },
+        {
+            "loss": 14.1293,
+            "grad_norm": 1.8881949186325073,
+            "learning_rate": 0.0005,
+            "epoch": 0.46543678115843784,
+            "step": 10385
+        },
+        {
+            "loss": 14.0341,
+            "grad_norm": 1.8966625928878784,
+            "learning_rate": 0.0005,
+            "epoch": 0.46566087204970336,
+            "step": 10390
+        },
+        {
+            "loss": 14.1341,
+            "grad_norm": 1.953307032585144,
+            "learning_rate": 0.0005,
+            "epoch": 0.46588496294096887,
+            "step": 10395
+        },
+        {
+            "loss": 14.0498,
+            "grad_norm": 1.8317679166793823,
+            "learning_rate": 0.0005,
+            "epoch": 0.46610905383223433,
+            "step": 10400
+        },
+        {
+            "loss": 14.0098,
+            "grad_norm": 1.9041969776153564,
+            "learning_rate": 0.0005,
+            "epoch": 0.46633314472349985,
+            "step": 10405
+        },
+        {
+            "loss": 14.0836,
+            "grad_norm": 2.216858386993408,
+            "learning_rate": 0.0005,
+            "epoch": 0.46655723561476536,
+            "step": 10410
+        },
+        {
+            "loss": 14.1337,
+            "grad_norm": 1.7740803956985474,
+            "learning_rate": 0.0005,
+            "epoch": 0.4667813265060308,
+            "step": 10415
+        },
+        {
+            "loss": 14.1292,
+            "grad_norm": 1.901526927947998,
+            "learning_rate": 0.0005,
+            "epoch": 0.46700541739729634,
+            "step": 10420
+        },
+        {
+            "loss": 14.0508,
+            "grad_norm": 2.048635244369507,
+            "learning_rate": 0.0005,
+            "epoch": 0.46722950828856186,
+            "step": 10425
+        },
+        {
+            "loss": 14.1055,
+            "grad_norm": 1.9151564836502075,
+            "learning_rate": 0.0005,
+            "epoch": 0.4674535991798273,
+            "step": 10430
+        },
+        {
+            "loss": 14.0799,
+            "grad_norm": 1.8743534088134766,
+            "learning_rate": 0.0005,
+            "epoch": 0.46767769007109283,
+            "step": 10435
+        },
+        {
+            "loss": 14.0249,
+            "grad_norm": 2.037383794784546,
+            "learning_rate": 0.0005,
+            "epoch": 0.46790178096235835,
+            "step": 10440
+        },
+        {
+            "loss": 14.1092,
+            "grad_norm": 1.9318444728851318,
+            "learning_rate": 0.0005,
+            "epoch": 0.4681258718536238,
+            "step": 10445
+        },
+        {
+            "loss": 14.1141,
+            "grad_norm": 1.904248595237732,
+            "learning_rate": 0.0005,
+            "epoch": 0.4683499627448893,
+            "step": 10450
+        },
+        {
+            "loss": 13.9967,
+            "grad_norm": 1.9173495769500732,
+            "learning_rate": 0.0005,
+            "epoch": 0.46857405363615484,
+            "step": 10455
+        },
+        {
+            "loss": 14.0076,
+            "grad_norm": 2.2218124866485596,
+            "learning_rate": 0.0005,
+            "epoch": 0.4687981445274203,
+            "step": 10460
+        },
+        {
+            "loss": 14.0625,
+            "grad_norm": 2.1360721588134766,
+            "learning_rate": 0.0005,
+            "epoch": 0.4690222354186858,
+            "step": 10465
+        },
+        {
+            "loss": 14.1023,
+            "grad_norm": 1.7907522916793823,
+            "learning_rate": 0.0005,
+            "epoch": 0.46924632630995133,
+            "step": 10470
+        },
+        {
+            "loss": 14.0633,
+            "grad_norm": 1.9924049377441406,
+            "learning_rate": 0.0005,
+            "epoch": 0.4694704172012168,
+            "step": 10475
+        },
+        {
+            "loss": 14.0495,
+            "grad_norm": 1.9566833972930908,
+            "learning_rate": 0.0005,
+            "epoch": 0.4696945080924823,
+            "step": 10480
+        },
+        {
+            "loss": 14.0408,
+            "grad_norm": 2.0456693172454834,
+            "learning_rate": 0.0005,
+            "epoch": 0.4699185989837478,
+            "step": 10485
+        },
+        {
+            "loss": 14.0525,
+            "grad_norm": 1.915246844291687,
+            "learning_rate": 0.0005,
+            "epoch": 0.4701426898750133,
+            "step": 10490
+        },
+        {
+            "loss": 14.0103,
+            "grad_norm": 1.8662196397781372,
+            "learning_rate": 0.0005,
+            "epoch": 0.4703667807662788,
+            "step": 10495
+        },
+        {
+            "loss": 14.0615,
+            "grad_norm": 2.1604061126708984,
+            "learning_rate": 0.0005,
+            "epoch": 0.4705908716575443,
+            "step": 10500
+        },
+        {
+            "eval_loss": 1.7521030902862549,
+            "eval_runtime": 18.4061,
+            "eval_samples_per_second": 890.139,
+            "eval_steps_per_second": 7.986,
+            "epoch": 0.4705908716575443,
+            "step": 10500
+        },
+        {
+            "loss": 14.1009,
+            "grad_norm": 2.012334108352661,
+            "learning_rate": 0.0005,
+            "epoch": 0.4708149625488098,
+            "step": 10505
+        },
+        {
+            "loss": 14.0878,
+            "grad_norm": 2.1435182094573975,
+            "learning_rate": 0.0005,
+            "epoch": 0.4710390534400753,
+            "step": 10510
+        },
+        {
+            "loss": 13.9449,
+            "grad_norm": 2.0218379497528076,
+            "learning_rate": 0.0005,
+            "epoch": 0.4712631443313408,
+            "step": 10515
+        },
+        {
+            "loss": 14.1722,
+            "grad_norm": 1.9891680479049683,
+            "learning_rate": 0.0005,
+            "epoch": 0.47148723522260627,
+            "step": 10520
+        },
+        {
+            "loss": 13.9856,
+            "grad_norm": 1.8381295204162598,
+            "learning_rate": 0.0005,
+            "epoch": 0.4717113261138718,
+            "step": 10525
+        },
+        {
+            "loss": 14.0003,
+            "grad_norm": 1.849729299545288,
+            "learning_rate": 0.0005,
+            "epoch": 0.4719354170051373,
+            "step": 10530
+        },
+        {
+            "loss": 13.9588,
+            "grad_norm": 1.8582322597503662,
+            "learning_rate": 0.0005,
+            "epoch": 0.47215950789640276,
+            "step": 10535
+        },
+        {
+            "loss": 13.9515,
+            "grad_norm": 1.9138835668563843,
+            "learning_rate": 0.0005,
+            "epoch": 0.4723835987876683,
+            "step": 10540
+        },
+        {
+            "loss": 14.0562,
+            "grad_norm": 1.8926070928573608,
+            "learning_rate": 0.0005,
+            "epoch": 0.4726076896789338,
+            "step": 10545
+        },
+        {
+            "loss": 13.957,
+            "grad_norm": 1.9087103605270386,
+            "learning_rate": 0.0005,
+            "epoch": 0.47283178057019926,
+            "step": 10550
+        },
+        {
+            "loss": 13.9708,
+            "grad_norm": 2.0126662254333496,
+            "learning_rate": 0.0005,
+            "epoch": 0.47305587146146477,
+            "step": 10555
+        },
+        {
+            "loss": 13.9551,
+            "grad_norm": 1.951674461364746,
+            "learning_rate": 0.0005,
+            "epoch": 0.4732799623527303,
+            "step": 10560
+        },
+        {
+            "loss": 14.1009,
+            "grad_norm": 1.8163164854049683,
+            "learning_rate": 0.0005,
+            "epoch": 0.47350405324399575,
+            "step": 10565
+        },
+        {
+            "loss": 14.1147,
+            "grad_norm": 2.0408079624176025,
+            "learning_rate": 0.0005,
+            "epoch": 0.47372814413526126,
+            "step": 10570
+        },
+        {
+            "loss": 14.0755,
+            "grad_norm": 2.0262444019317627,
+            "learning_rate": 0.0005,
+            "epoch": 0.4739522350265268,
+            "step": 10575
+        },
+        {
+            "loss": 14.1181,
+            "grad_norm": 1.8418506383895874,
+            "learning_rate": 0.0005,
+            "epoch": 0.47417632591779224,
+            "step": 10580
+        },
+        {
+            "loss": 13.9996,
+            "grad_norm": 1.8340831995010376,
+            "learning_rate": 0.0005,
+            "epoch": 0.47440041680905776,
+            "step": 10585
+        },
+        {
+            "loss": 14.0614,
+            "grad_norm": 1.894936203956604,
+            "learning_rate": 0.0005,
+            "epoch": 0.4746245077003233,
+            "step": 10590
+        },
+        {
+            "loss": 13.9815,
+            "grad_norm": 1.7638943195343018,
+            "learning_rate": 0.0005,
+            "epoch": 0.47484859859158873,
+            "step": 10595
+        },
+        {
+            "loss": 14.115,
+            "grad_norm": 2.0073249340057373,
+            "learning_rate": 0.0005,
+            "epoch": 0.47507268948285425,
+            "step": 10600
+        },
+        {
+            "loss": 14.0861,
+            "grad_norm": 1.953412413597107,
+            "learning_rate": 0.0005,
+            "epoch": 0.47529678037411976,
+            "step": 10605
+        },
+        {
+            "loss": 13.9858,
+            "grad_norm": 1.7932475805282593,
+            "learning_rate": 0.0005,
+            "epoch": 0.4755208712653852,
+            "step": 10610
+        },
+        {
+            "loss": 13.9297,
+            "grad_norm": 1.9420661926269531,
+            "learning_rate": 0.0005,
+            "epoch": 0.47574496215665074,
+            "step": 10615
+        },
+        {
+            "loss": 14.0705,
+            "grad_norm": 1.8492196798324585,
+            "learning_rate": 0.0005,
+            "epoch": 0.47596905304791626,
+            "step": 10620
+        },
+        {
+            "loss": 14.06,
+            "grad_norm": 1.8225876092910767,
+            "learning_rate": 0.0005,
+            "epoch": 0.4761931439391817,
+            "step": 10625
+        },
+        {
+            "loss": 14.0748,
+            "grad_norm": 1.7609186172485352,
+            "learning_rate": 0.0005,
+            "epoch": 0.47641723483044723,
+            "step": 10630
+        },
+        {
+            "loss": 13.9309,
+            "grad_norm": 1.9755157232284546,
+            "learning_rate": 0.0005,
+            "epoch": 0.47664132572171275,
+            "step": 10635
+        },
+        {
+            "loss": 14.0347,
+            "grad_norm": 1.896437644958496,
+            "learning_rate": 0.0005,
+            "epoch": 0.4768654166129782,
+            "step": 10640
+        },
+        {
+            "loss": 14.029,
+            "grad_norm": 1.8733117580413818,
+            "learning_rate": 0.0005,
+            "epoch": 0.4770895075042437,
+            "step": 10645
+        },
+        {
+            "loss": 14.0766,
+            "grad_norm": 2.0224342346191406,
+            "learning_rate": 0.0005,
+            "epoch": 0.47731359839550924,
+            "step": 10650
+        },
+        {
+            "loss": 13.886,
+            "grad_norm": 1.963172197341919,
+            "learning_rate": 0.0005,
+            "epoch": 0.4775376892867747,
+            "step": 10655
+        },
+        {
+            "loss": 14.0283,
+            "grad_norm": 1.968991994857788,
+            "learning_rate": 0.0005,
+            "epoch": 0.4777617801780402,
+            "step": 10660
+        },
+        {
+            "loss": 14.0249,
+            "grad_norm": 1.9693257808685303,
+            "learning_rate": 0.0005,
+            "epoch": 0.47798587106930573,
+            "step": 10665
+        },
+        {
+            "loss": 13.9854,
+            "grad_norm": 2.002882242202759,
+            "learning_rate": 0.0005,
+            "epoch": 0.4782099619605712,
+            "step": 10670
+        },
+        {
+            "loss": 14.0813,
+            "grad_norm": 1.8276903629302979,
+            "learning_rate": 0.0005,
+            "epoch": 0.4784340528518367,
+            "step": 10675
+        },
+        {
+            "loss": 14.0866,
+            "grad_norm": 1.9171091318130493,
+            "learning_rate": 0.0005,
+            "epoch": 0.4786581437431022,
+            "step": 10680
+        },
+        {
+            "loss": 14.0089,
+            "grad_norm": 1.8800194263458252,
+            "learning_rate": 0.0005,
+            "epoch": 0.4788822346343677,
+            "step": 10685
+        },
+        {
+            "loss": 13.999,
+            "grad_norm": 1.8443156480789185,
+            "learning_rate": 0.0005,
+            "epoch": 0.4791063255256332,
+            "step": 10690
+        },
+        {
+            "loss": 14.0472,
+            "grad_norm": 1.7933176755905151,
+            "learning_rate": 0.0005,
+            "epoch": 0.4793304164168987,
+            "step": 10695
+        },
+        {
+            "loss": 14.0846,
+            "grad_norm": 2.079235315322876,
+            "learning_rate": 0.0005,
+            "epoch": 0.4795545073081642,
+            "step": 10700
+        },
+        {
+            "loss": 14.0334,
+            "grad_norm": 1.8366498947143555,
+            "learning_rate": 0.0005,
+            "epoch": 0.4797785981994297,
+            "step": 10705
+        },
+        {
+            "loss": 14.0646,
+            "grad_norm": 1.997281789779663,
+            "learning_rate": 0.0005,
+            "epoch": 0.4800026890906952,
+            "step": 10710
+        },
+        {
+            "loss": 13.9765,
+            "grad_norm": 2.008910655975342,
+            "learning_rate": 0.0005,
+            "epoch": 0.48022677998196067,
+            "step": 10715
+        },
+        {
+            "loss": 14.0477,
+            "grad_norm": 1.8822402954101562,
+            "learning_rate": 0.0005,
+            "epoch": 0.4804508708732262,
+            "step": 10720
+        },
+        {
+            "loss": 14.0811,
+            "grad_norm": 2.069028854370117,
+            "learning_rate": 0.0005,
+            "epoch": 0.4806749617644917,
+            "step": 10725
+        },
+        {
+            "loss": 13.9759,
+            "grad_norm": 2.019068479537964,
+            "learning_rate": 0.0005,
+            "epoch": 0.48089905265575716,
+            "step": 10730
+        },
+        {
+            "loss": 14.0791,
+            "grad_norm": 1.9277989864349365,
+            "learning_rate": 0.0005,
+            "epoch": 0.4811231435470227,
+            "step": 10735
+        },
+        {
+            "loss": 14.0211,
+            "grad_norm": 2.0322346687316895,
+            "learning_rate": 0.0005,
+            "epoch": 0.4813472344382882,
+            "step": 10740
+        },
+        {
+            "loss": 14.1959,
+            "grad_norm": 2.0111358165740967,
+            "learning_rate": 0.0005,
+            "epoch": 0.48157132532955366,
+            "step": 10745
+        },
+        {
+            "loss": 14.0045,
+            "grad_norm": 2.2163989543914795,
+            "learning_rate": 0.0005,
+            "epoch": 0.4817954162208192,
+            "step": 10750
+        },
+        {
+            "loss": 13.9323,
+            "grad_norm": 2.0738158226013184,
+            "learning_rate": 0.0005,
+            "epoch": 0.4820195071120847,
+            "step": 10755
+        },
+        {
+            "loss": 14.0603,
+            "grad_norm": 1.9725066423416138,
+            "learning_rate": 0.0005,
+            "epoch": 0.48224359800335015,
+            "step": 10760
+        },
+        {
+            "loss": 14.0725,
+            "grad_norm": 2.052971601486206,
+            "learning_rate": 0.0005,
+            "epoch": 0.48246768889461566,
+            "step": 10765
+        },
+        {
+            "loss": 14.1246,
+            "grad_norm": 2.0623252391815186,
+            "learning_rate": 0.0005,
+            "epoch": 0.4826917797858812,
+            "step": 10770
+        },
+        {
+            "loss": 14.0091,
+            "grad_norm": 1.806399941444397,
+            "learning_rate": 0.0005,
+            "epoch": 0.48291587067714664,
+            "step": 10775
+        },
+        {
+            "loss": 14.011,
+            "grad_norm": 1.7548433542251587,
+            "learning_rate": 0.0005,
+            "epoch": 0.48313996156841216,
+            "step": 10780
+        },
+        {
+            "loss": 14.0414,
+            "grad_norm": 1.7872982025146484,
+            "learning_rate": 0.0005,
+            "epoch": 0.4833640524596776,
+            "step": 10785
+        },
+        {
+            "loss": 14.0229,
+            "grad_norm": 1.8104687929153442,
+            "learning_rate": 0.0005,
+            "epoch": 0.48358814335094313,
+            "step": 10790
+        },
+        {
+            "loss": 13.9876,
+            "grad_norm": 1.7846254110336304,
+            "learning_rate": 0.0005,
+            "epoch": 0.48381223424220865,
+            "step": 10795
+        },
+        {
+            "loss": 14.0102,
+            "grad_norm": 1.9332680702209473,
+            "learning_rate": 0.0005,
+            "epoch": 0.4840363251334741,
+            "step": 10800
+        },
+        {
+            "loss": 14.1051,
+            "grad_norm": 1.975170612335205,
+            "learning_rate": 0.0005,
+            "epoch": 0.4842604160247396,
+            "step": 10805
+        },
+        {
+            "loss": 13.994,
+            "grad_norm": 1.882921576499939,
+            "learning_rate": 0.0005,
+            "epoch": 0.48448450691600514,
+            "step": 10810
+        },
+        {
+            "loss": 14.0343,
+            "grad_norm": 1.848868727684021,
+            "learning_rate": 0.0005,
+            "epoch": 0.4847085978072706,
+            "step": 10815
+        },
+        {
+            "loss": 13.9719,
+            "grad_norm": 1.8909270763397217,
+            "learning_rate": 0.0005,
+            "epoch": 0.4849326886985361,
+            "step": 10820
+        },
+        {
+            "loss": 14.0679,
+            "grad_norm": 1.6264833211898804,
+            "learning_rate": 0.0005,
+            "epoch": 0.48515677958980163,
+            "step": 10825
+        },
+        {
+            "loss": 14.073,
+            "grad_norm": 1.9281812906265259,
+            "learning_rate": 0.0005,
+            "epoch": 0.4853808704810671,
+            "step": 10830
+        },
+        {
+            "loss": 14.03,
+            "grad_norm": 2.016878128051758,
+            "learning_rate": 0.0005,
+            "epoch": 0.4856049613723326,
+            "step": 10835
+        },
+        {
+            "loss": 14.0051,
+            "grad_norm": 1.7301148176193237,
+            "learning_rate": 0.0005,
+            "epoch": 0.4858290522635981,
+            "step": 10840
+        },
+        {
+            "loss": 14.0598,
+            "grad_norm": 1.7245323657989502,
+            "learning_rate": 0.0005,
+            "epoch": 0.4860531431548636,
+            "step": 10845
+        },
+        {
+            "loss": 14.0583,
+            "grad_norm": 1.7455264329910278,
+            "learning_rate": 0.0005,
+            "epoch": 0.4862772340461291,
+            "step": 10850
+        },
+        {
+            "loss": 14.0273,
+            "grad_norm": 1.8355683088302612,
+            "learning_rate": 0.0005,
+            "epoch": 0.4865013249373946,
+            "step": 10855
+        },
+        {
+            "loss": 13.9864,
+            "grad_norm": 1.8264118432998657,
+            "learning_rate": 0.0005,
+            "epoch": 0.4867254158286601,
+            "step": 10860
+        },
+        {
+            "loss": 14.049,
+            "grad_norm": 2.033604145050049,
+            "learning_rate": 0.0005,
+            "epoch": 0.4869495067199256,
+            "step": 10865
+        },
+        {
+            "loss": 14.0507,
+            "grad_norm": 2.197317600250244,
+            "learning_rate": 0.0005,
+            "epoch": 0.4871735976111911,
+            "step": 10870
+        },
+        {
+            "loss": 13.9427,
+            "grad_norm": 1.9294264316558838,
+            "learning_rate": 0.0005,
+            "epoch": 0.48739768850245657,
+            "step": 10875
+        },
+        {
+            "loss": 14.109,
+            "grad_norm": 1.965006709098816,
+            "learning_rate": 0.0005,
+            "epoch": 0.4876217793937221,
+            "step": 10880
+        },
+        {
+            "loss": 14.1184,
+            "grad_norm": 1.8745315074920654,
+            "learning_rate": 0.0005,
+            "epoch": 0.4878458702849876,
+            "step": 10885
+        },
+        {
+            "loss": 14.0252,
+            "grad_norm": 1.86996328830719,
+            "learning_rate": 0.0005,
+            "epoch": 0.48806996117625306,
+            "step": 10890
+        },
+        {
+            "loss": 14.0483,
+            "grad_norm": 1.8305915594100952,
+            "learning_rate": 0.0005,
+            "epoch": 0.4882940520675186,
+            "step": 10895
+        },
+        {
+            "loss": 13.9539,
+            "grad_norm": 1.8650505542755127,
+            "learning_rate": 0.0005,
+            "epoch": 0.4885181429587841,
+            "step": 10900
+        },
+        {
+            "loss": 14.1072,
+            "grad_norm": 1.9065357446670532,
+            "learning_rate": 0.0005,
+            "epoch": 0.48874223385004956,
+            "step": 10905
+        },
+        {
+            "loss": 14.0088,
+            "grad_norm": 1.9280742406845093,
+            "learning_rate": 0.0005,
+            "epoch": 0.4889663247413151,
+            "step": 10910
+        },
+        {
+            "loss": 14.0887,
+            "grad_norm": 1.852734923362732,
+            "learning_rate": 0.0005,
+            "epoch": 0.4891904156325806,
+            "step": 10915
+        },
+        {
+            "loss": 14.0088,
+            "grad_norm": 1.8719218969345093,
+            "learning_rate": 0.0005,
+            "epoch": 0.48941450652384605,
+            "step": 10920
+        },
+        {
+            "loss": 14.0579,
+            "grad_norm": 1.9777988195419312,
+            "learning_rate": 0.0005,
+            "epoch": 0.48963859741511156,
+            "step": 10925
+        },
+        {
+            "loss": 13.9555,
+            "grad_norm": 1.9106028079986572,
+            "learning_rate": 0.0005,
+            "epoch": 0.4898626883063771,
+            "step": 10930
+        },
+        {
+            "loss": 13.9221,
+            "grad_norm": 1.9049543142318726,
+            "learning_rate": 0.0005,
+            "epoch": 0.49008677919764254,
+            "step": 10935
+        },
+        {
+            "loss": 14.0693,
+            "grad_norm": 1.785038948059082,
+            "learning_rate": 0.0005,
+            "epoch": 0.49031087008890806,
+            "step": 10940
+        },
+        {
+            "loss": 13.9304,
+            "grad_norm": 1.9151285886764526,
+            "learning_rate": 0.0005,
+            "epoch": 0.4905349609801736,
+            "step": 10945
+        },
+        {
+            "loss": 13.9709,
+            "grad_norm": 1.9431225061416626,
+            "learning_rate": 0.0005,
+            "epoch": 0.49075905187143903,
+            "step": 10950
+        },
+        {
+            "loss": 14.0511,
+            "grad_norm": 1.8583786487579346,
+            "learning_rate": 0.0005,
+            "epoch": 0.49098314276270455,
+            "step": 10955
+        },
+        {
+            "loss": 14.0531,
+            "grad_norm": 1.806581974029541,
+            "learning_rate": 0.0005,
+            "epoch": 0.49120723365397007,
+            "step": 10960
+        },
+        {
+            "loss": 14.0427,
+            "grad_norm": 1.8097478151321411,
+            "learning_rate": 0.0005,
+            "epoch": 0.4914313245452355,
+            "step": 10965
+        },
+        {
+            "loss": 13.9989,
+            "grad_norm": 1.8856887817382812,
+            "learning_rate": 0.0005,
+            "epoch": 0.49165541543650104,
+            "step": 10970
+        },
+        {
+            "loss": 14.0249,
+            "grad_norm": 1.7090424299240112,
+            "learning_rate": 0.0005,
+            "epoch": 0.49187950632776656,
+            "step": 10975
+        },
+        {
+            "loss": 13.9924,
+            "grad_norm": 1.8393125534057617,
+            "learning_rate": 0.0005,
+            "epoch": 0.492103597219032,
+            "step": 10980
+        },
+        {
+            "loss": 14.094,
+            "grad_norm": 1.9828705787658691,
+            "learning_rate": 0.0005,
+            "epoch": 0.49232768811029753,
+            "step": 10985
+        },
+        {
+            "loss": 13.9907,
+            "grad_norm": 1.8064554929733276,
+            "learning_rate": 0.0005,
+            "epoch": 0.49255177900156305,
+            "step": 10990
+        },
+        {
+            "loss": 14.0491,
+            "grad_norm": 1.8088335990905762,
+            "learning_rate": 0.0005,
+            "epoch": 0.4927758698928285,
+            "step": 10995
+        },
+        {
+            "loss": 14.003,
+            "grad_norm": 1.7703953981399536,
+            "learning_rate": 0.0005,
+            "epoch": 0.492999960784094,
+            "step": 11000
+        },
+        {
+            "eval_loss": 1.7476418018341064,
+            "eval_runtime": 18.8201,
+            "eval_samples_per_second": 870.558,
+            "eval_steps_per_second": 7.811,
+            "epoch": 0.492999960784094,
+            "step": 11000
+        },
+        {
+            "loss": 13.9324,
+            "grad_norm": 1.9696674346923828,
+            "learning_rate": 0.0005,
+            "epoch": 0.49322405167535954,
+            "step": 11005
+        },
+        {
+            "loss": 13.99,
+            "grad_norm": 1.7030773162841797,
+            "learning_rate": 0.0005,
+            "epoch": 0.493448142566625,
+            "step": 11010
+        },
+        {
+            "loss": 13.991,
+            "grad_norm": 1.8188631534576416,
+            "learning_rate": 0.0005,
+            "epoch": 0.4936722334578905,
+            "step": 11015
+        },
+        {
+            "loss": 14.0508,
+            "grad_norm": 1.8433138132095337,
+            "learning_rate": 0.0005,
+            "epoch": 0.49389632434915604,
+            "step": 11020
+        },
+        {
+            "loss": 13.9965,
+            "grad_norm": 1.8021408319473267,
+            "learning_rate": 0.0005,
+            "epoch": 0.4941204152404215,
+            "step": 11025
+        },
+        {
+            "loss": 14.0115,
+            "grad_norm": 2.0757226943969727,
+            "learning_rate": 0.0005,
+            "epoch": 0.494344506131687,
+            "step": 11030
+        },
+        {
+            "loss": 14.0115,
+            "grad_norm": 1.926236629486084,
+            "learning_rate": 0.0005,
+            "epoch": 0.4945685970229525,
+            "step": 11035
+        },
+        {
+            "loss": 14.1292,
+            "grad_norm": 1.8639720678329468,
+            "learning_rate": 0.0005,
+            "epoch": 0.494792687914218,
+            "step": 11040
+        },
+        {
+            "loss": 14.1444,
+            "grad_norm": 1.8459327220916748,
+            "learning_rate": 0.0005,
+            "epoch": 0.4950167788054835,
+            "step": 11045
+        },
+        {
+            "loss": 13.9817,
+            "grad_norm": 1.8979179859161377,
+            "learning_rate": 0.0005,
+            "epoch": 0.495240869696749,
+            "step": 11050
+        },
+        {
+            "loss": 13.9854,
+            "grad_norm": 1.7826581001281738,
+            "learning_rate": 0.0005,
+            "epoch": 0.4954649605880145,
+            "step": 11055
+        },
+        {
+            "loss": 13.9973,
+            "grad_norm": 1.806075930595398,
+            "learning_rate": 0.0005,
+            "epoch": 0.49568905147928,
+            "step": 11060
+        },
+        {
+            "loss": 14.0814,
+            "grad_norm": 1.7833302021026611,
+            "learning_rate": 0.0005,
+            "epoch": 0.4959131423705455,
+            "step": 11065
+        },
+        {
+            "loss": 14.0437,
+            "grad_norm": 1.7272964715957642,
+            "learning_rate": 0.0005,
+            "epoch": 0.496137233261811,
+            "step": 11070
+        },
+        {
+            "loss": 14.0318,
+            "grad_norm": 1.983668565750122,
+            "learning_rate": 0.0005,
+            "epoch": 0.4963613241530765,
+            "step": 11075
+        },
+        {
+            "loss": 13.9409,
+            "grad_norm": 1.959157943725586,
+            "learning_rate": 0.0005,
+            "epoch": 0.496585415044342,
+            "step": 11080
+        },
+        {
+            "loss": 14.0618,
+            "grad_norm": 2.06475567817688,
+            "learning_rate": 0.0005,
+            "epoch": 0.49680950593560746,
+            "step": 11085
+        },
+        {
+            "loss": 14.0494,
+            "grad_norm": 2.235586643218994,
+            "learning_rate": 0.0005,
+            "epoch": 0.497033596826873,
+            "step": 11090
+        },
+        {
+            "loss": 13.9587,
+            "grad_norm": 2.183415412902832,
+            "learning_rate": 0.0005,
+            "epoch": 0.4972576877181385,
+            "step": 11095
+        },
+        {
+            "loss": 13.8766,
+            "grad_norm": 1.8275450468063354,
+            "learning_rate": 0.0005,
+            "epoch": 0.49748177860940396,
+            "step": 11100
+        },
+        {
+            "loss": 14.0381,
+            "grad_norm": 1.713709831237793,
+            "learning_rate": 0.0005,
+            "epoch": 0.4977058695006695,
+            "step": 11105
+        },
+        {
+            "loss": 14.0614,
+            "grad_norm": 1.8228458166122437,
+            "learning_rate": 0.0005,
+            "epoch": 0.497929960391935,
+            "step": 11110
+        },
+        {
+            "loss": 14.1712,
+            "grad_norm": 2.0068023204803467,
+            "learning_rate": 0.0005,
+            "epoch": 0.49815405128320045,
+            "step": 11115
+        },
+        {
+            "loss": 14.012,
+            "grad_norm": 1.9167864322662354,
+            "learning_rate": 0.0005,
+            "epoch": 0.49837814217446597,
+            "step": 11120
+        },
+        {
+            "loss": 14.0244,
+            "grad_norm": 1.7285057306289673,
+            "learning_rate": 0.0005,
+            "epoch": 0.4986022330657315,
+            "step": 11125
+        },
+        {
+            "loss": 14.0137,
+            "grad_norm": 1.9569995403289795,
+            "learning_rate": 0.0005,
+            "epoch": 0.49882632395699694,
+            "step": 11130
+        },
+        {
+            "loss": 14.0519,
+            "grad_norm": 1.9410467147827148,
+            "learning_rate": 0.0005,
+            "epoch": 0.49905041484826246,
+            "step": 11135
+        },
+        {
+            "loss": 14.0137,
+            "grad_norm": 1.8874870538711548,
+            "learning_rate": 0.0005,
+            "epoch": 0.499274505739528,
+            "step": 11140
+        },
+        {
+            "loss": 13.9772,
+            "grad_norm": 1.7927515506744385,
+            "learning_rate": 0.0005,
+            "epoch": 0.49949859663079343,
+            "step": 11145
+        },
+        {
+            "loss": 14.0682,
+            "grad_norm": 1.7767921686172485,
+            "learning_rate": 0.0005,
+            "epoch": 0.49972268752205895,
+            "step": 11150
+        },
+        {
+            "loss": 14.0448,
+            "grad_norm": 1.9172736406326294,
+            "learning_rate": 0.0005,
+            "epoch": 0.49994677841332447,
+            "step": 11155
+        },
+        {
+            "loss": 14.024,
+            "grad_norm": 1.8923771381378174,
+            "learning_rate": 0.0005,
+            "epoch": 0.50017086930459,
+            "step": 11160
+        },
+        {
+            "loss": 14.0349,
+            "grad_norm": 1.8693808317184448,
+            "learning_rate": 0.0005,
+            "epoch": 0.5003949601958554,
+            "step": 11165
+        },
+        {
+            "loss": 14.1001,
+            "grad_norm": 1.885910987854004,
+            "learning_rate": 0.0005,
+            "epoch": 0.5006190510871209,
+            "step": 11170
+        },
+        {
+            "loss": 14.0163,
+            "grad_norm": 1.7689731121063232,
+            "learning_rate": 0.0005,
+            "epoch": 0.5008431419783864,
+            "step": 11175
+        },
+        {
+            "loss": 14.1057,
+            "grad_norm": 1.9417201280593872,
+            "learning_rate": 0.0005,
+            "epoch": 0.5010672328696519,
+            "step": 11180
+        },
+        {
+            "loss": 13.9506,
+            "grad_norm": 1.7507840394973755,
+            "learning_rate": 0.0005,
+            "epoch": 0.5012913237609175,
+            "step": 11185
+        },
+        {
+            "loss": 14.1209,
+            "grad_norm": 1.9742436408996582,
+            "learning_rate": 0.0005,
+            "epoch": 0.501515414652183,
+            "step": 11190
+        },
+        {
+            "loss": 14.116,
+            "grad_norm": 2.121680736541748,
+            "learning_rate": 0.0005,
+            "epoch": 0.5017395055434484,
+            "step": 11195
+        },
+        {
+            "loss": 14.0399,
+            "grad_norm": 2.2621843814849854,
+            "learning_rate": 0.0005,
+            "epoch": 0.5019635964347139,
+            "step": 11200
+        },
+        {
+            "loss": 14.0067,
+            "grad_norm": 2.0274646282196045,
+            "learning_rate": 0.0005,
+            "epoch": 0.5021876873259794,
+            "step": 11205
+        },
+        {
+            "loss": 13.9607,
+            "grad_norm": 1.958794355392456,
+            "learning_rate": 0.0005,
+            "epoch": 0.5024117782172449,
+            "step": 11210
+        },
+        {
+            "loss": 14.0712,
+            "grad_norm": 1.909808874130249,
+            "learning_rate": 0.0005,
+            "epoch": 0.5026358691085104,
+            "step": 11215
+        },
+        {
+            "loss": 14.0126,
+            "grad_norm": 1.8823260068893433,
+            "learning_rate": 0.0005,
+            "epoch": 0.502859959999776,
+            "step": 11220
+        },
+        {
+            "loss": 14.1237,
+            "grad_norm": 1.9062442779541016,
+            "learning_rate": 0.0005,
+            "epoch": 0.5030840508910414,
+            "step": 11225
+        },
+        {
+            "loss": 14.0226,
+            "grad_norm": 2.011108875274658,
+            "learning_rate": 0.0005,
+            "epoch": 0.5033081417823069,
+            "step": 11230
+        },
+        {
+            "loss": 13.9678,
+            "grad_norm": 1.982003927230835,
+            "learning_rate": 0.0005,
+            "epoch": 0.5035322326735724,
+            "step": 11235
+        },
+        {
+            "loss": 14.0562,
+            "grad_norm": 2.152905225753784,
+            "learning_rate": 0.0005,
+            "epoch": 0.5037563235648379,
+            "step": 11240
+        },
+        {
+            "loss": 14.0348,
+            "grad_norm": 1.9049335718154907,
+            "learning_rate": 0.0005,
+            "epoch": 0.5039804144561034,
+            "step": 11245
+        },
+        {
+            "loss": 14.095,
+            "grad_norm": 1.7218148708343506,
+            "learning_rate": 0.0005,
+            "epoch": 0.5042045053473689,
+            "step": 11250
+        },
+        {
+            "loss": 14.1642,
+            "grad_norm": 1.7122304439544678,
+            "learning_rate": 0.0005,
+            "epoch": 0.5044285962386343,
+            "step": 11255
+        },
+        {
+            "loss": 14.1851,
+            "grad_norm": 1.7507672309875488,
+            "learning_rate": 0.0005,
+            "epoch": 0.5046526871298999,
+            "step": 11260
+        },
+        {
+            "loss": 14.0357,
+            "grad_norm": 1.7015080451965332,
+            "learning_rate": 0.0005,
+            "epoch": 0.5048767780211654,
+            "step": 11265
+        },
+        {
+            "loss": 13.9302,
+            "grad_norm": 1.9490917921066284,
+            "learning_rate": 0.0005,
+            "epoch": 0.5051008689124309,
+            "step": 11270
+        },
+        {
+            "loss": 14.1446,
+            "grad_norm": 2.090062379837036,
+            "learning_rate": 0.0005,
+            "epoch": 0.5053249598036964,
+            "step": 11275
+        },
+        {
+            "loss": 14.0398,
+            "grad_norm": 2.121561288833618,
+            "learning_rate": 0.0005,
+            "epoch": 0.5055490506949619,
+            "step": 11280
+        },
+        {
+            "loss": 14.0797,
+            "grad_norm": 2.3040390014648438,
+            "learning_rate": 0.0005,
+            "epoch": 0.5057731415862273,
+            "step": 11285
+        },
+        {
+            "loss": 13.9851,
+            "grad_norm": 2.0667917728424072,
+            "learning_rate": 0.0005,
+            "epoch": 0.5059972324774928,
+            "step": 11290
+        },
+        {
+            "loss": 13.9989,
+            "grad_norm": 1.7419085502624512,
+            "learning_rate": 0.0005,
+            "epoch": 0.5062213233687584,
+            "step": 11295
+        },
+        {
+            "loss": 14.0088,
+            "grad_norm": 1.9543952941894531,
+            "learning_rate": 0.0005,
+            "epoch": 0.5064454142600239,
+            "step": 11300
+        },
+        {
+            "loss": 13.9647,
+            "grad_norm": 1.8733443021774292,
+            "learning_rate": 0.0005,
+            "epoch": 0.5066695051512894,
+            "step": 11305
+        },
+        {
+            "loss": 13.911,
+            "grad_norm": 1.7971140146255493,
+            "learning_rate": 0.0005,
+            "epoch": 0.5068935960425549,
+            "step": 11310
+        },
+        {
+            "loss": 14.0001,
+            "grad_norm": 1.9756284952163696,
+            "learning_rate": 0.0005,
+            "epoch": 0.5071176869338203,
+            "step": 11315
+        },
+        {
+            "loss": 13.9781,
+            "grad_norm": 1.8535903692245483,
+            "learning_rate": 0.0005,
+            "epoch": 0.5073417778250858,
+            "step": 11320
+        },
+        {
+            "loss": 14.1203,
+            "grad_norm": 1.8332593441009521,
+            "learning_rate": 0.0005,
+            "epoch": 0.5075658687163513,
+            "step": 11325
+        },
+        {
+            "loss": 14.0118,
+            "grad_norm": 2.1774582862854004,
+            "learning_rate": 0.0005,
+            "epoch": 0.5077899596076169,
+            "step": 11330
+        },
+        {
+            "loss": 14.049,
+            "grad_norm": 2.049069881439209,
+            "learning_rate": 0.0005,
+            "epoch": 0.5080140504988824,
+            "step": 11335
+        },
+        {
+            "loss": 14.1635,
+            "grad_norm": 2.0563483238220215,
+            "learning_rate": 0.0005,
+            "epoch": 0.5082381413901479,
+            "step": 11340
+        },
+        {
+            "loss": 14.012,
+            "grad_norm": 1.9674501419067383,
+            "learning_rate": 0.0005,
+            "epoch": 0.5084622322814133,
+            "step": 11345
+        },
+        {
+            "loss": 13.9944,
+            "grad_norm": 2.106797218322754,
+            "learning_rate": 0.0005,
+            "epoch": 0.5086863231726788,
+            "step": 11350
+        },
+        {
+            "loss": 14.0308,
+            "grad_norm": 1.7767261266708374,
+            "learning_rate": 0.0005,
+            "epoch": 0.5089104140639443,
+            "step": 11355
+        },
+        {
+            "loss": 14.122,
+            "grad_norm": 1.788737177848816,
+            "learning_rate": 0.0005,
+            "epoch": 0.5091345049552098,
+            "step": 11360
+        },
+        {
+            "loss": 14.06,
+            "grad_norm": 1.8600432872772217,
+            "learning_rate": 0.0005,
+            "epoch": 0.5093585958464754,
+            "step": 11365
+        },
+        {
+            "loss": 14.0932,
+            "grad_norm": 1.8921267986297607,
+            "learning_rate": 0.0005,
+            "epoch": 0.5095826867377409,
+            "step": 11370
+        },
+        {
+            "loss": 13.9985,
+            "grad_norm": 1.851744532585144,
+            "learning_rate": 0.0005,
+            "epoch": 0.5098067776290063,
+            "step": 11375
+        },
+        {
+            "loss": 14.0089,
+            "grad_norm": 1.879716396331787,
+            "learning_rate": 0.0005,
+            "epoch": 0.5100308685202718,
+            "step": 11380
+        },
+        {
+            "loss": 14.16,
+            "grad_norm": 2.2035632133483887,
+            "learning_rate": 0.0005,
+            "epoch": 0.5102549594115373,
+            "step": 11385
+        },
+        {
+            "loss": 14.0877,
+            "grad_norm": 2.0263664722442627,
+            "learning_rate": 0.0005,
+            "epoch": 0.5104790503028028,
+            "step": 11390
+        },
+        {
+            "loss": 14.0885,
+            "grad_norm": 1.9041746854782104,
+            "learning_rate": 0.0005,
+            "epoch": 0.5107031411940683,
+            "step": 11395
+        },
+        {
+            "loss": 14.042,
+            "grad_norm": 1.671951413154602,
+            "learning_rate": 0.0005,
+            "epoch": 0.5109272320853339,
+            "step": 11400
+        },
+        {
+            "loss": 13.9809,
+            "grad_norm": 1.9111623764038086,
+            "learning_rate": 0.0005,
+            "epoch": 0.5111513229765993,
+            "step": 11405
+        },
+        {
+            "loss": 14.1244,
+            "grad_norm": 1.9198267459869385,
+            "learning_rate": 0.0005,
+            "epoch": 0.5113754138678648,
+            "step": 11410
+        },
+        {
+            "loss": 13.9623,
+            "grad_norm": 1.9086799621582031,
+            "learning_rate": 0.0005,
+            "epoch": 0.5115995047591303,
+            "step": 11415
+        },
+        {
+            "loss": 14.0608,
+            "grad_norm": 1.8582472801208496,
+            "learning_rate": 0.0005,
+            "epoch": 0.5118235956503958,
+            "step": 11420
+        },
+        {
+            "loss": 14.0039,
+            "grad_norm": 2.1490015983581543,
+            "learning_rate": 0.0005,
+            "epoch": 0.5120476865416613,
+            "step": 11425
+        },
+        {
+            "loss": 14.0627,
+            "grad_norm": 1.8919426202774048,
+            "learning_rate": 0.0005,
+            "epoch": 0.5122717774329268,
+            "step": 11430
+        },
+        {
+            "loss": 13.9811,
+            "grad_norm": 1.8844960927963257,
+            "learning_rate": 0.0005,
+            "epoch": 0.5124958683241922,
+            "step": 11435
+        },
+        {
+            "loss": 13.9331,
+            "grad_norm": 1.825016975402832,
+            "learning_rate": 0.0005,
+            "epoch": 0.5127199592154578,
+            "step": 11440
+        },
+        {
+            "loss": 13.9853,
+            "grad_norm": 1.7920079231262207,
+            "learning_rate": 0.0005,
+            "epoch": 0.5129440501067233,
+            "step": 11445
+        },
+        {
+            "loss": 13.9268,
+            "grad_norm": 1.7441658973693848,
+            "learning_rate": 0.0005,
+            "epoch": 0.5131681409979888,
+            "step": 11450
+        },
+        {
+            "loss": 14.1053,
+            "grad_norm": 1.8774532079696655,
+            "learning_rate": 0.0005,
+            "epoch": 0.5133922318892543,
+            "step": 11455
+        },
+        {
+            "loss": 14.0352,
+            "grad_norm": 1.9215784072875977,
+            "learning_rate": 0.0005,
+            "epoch": 0.5136163227805198,
+            "step": 11460
+        },
+        {
+            "loss": 14.113,
+            "grad_norm": 1.8764196634292603,
+            "learning_rate": 0.0005,
+            "epoch": 0.5138404136717852,
+            "step": 11465
+        },
+        {
+            "loss": 14.0281,
+            "grad_norm": 1.7915804386138916,
+            "learning_rate": 0.0005,
+            "epoch": 0.5140645045630508,
+            "step": 11470
+        },
+        {
+            "loss": 14.0374,
+            "grad_norm": 1.8837881088256836,
+            "learning_rate": 0.0005,
+            "epoch": 0.5142885954543163,
+            "step": 11475
+        },
+        {
+            "loss": 14.0387,
+            "grad_norm": 1.8066962957382202,
+            "learning_rate": 0.0005,
+            "epoch": 0.5145126863455818,
+            "step": 11480
+        },
+        {
+            "loss": 14.0537,
+            "grad_norm": 1.827216625213623,
+            "learning_rate": 0.0005,
+            "epoch": 0.5147367772368473,
+            "step": 11485
+        },
+        {
+            "loss": 14.0311,
+            "grad_norm": 1.8255128860473633,
+            "learning_rate": 0.0005,
+            "epoch": 0.5149608681281128,
+            "step": 11490
+        },
+        {
+            "loss": 13.9886,
+            "grad_norm": 1.8707716464996338,
+            "learning_rate": 0.0005,
+            "epoch": 0.5151849590193782,
+            "step": 11495
+        },
+        {
+            "loss": 14.1195,
+            "grad_norm": 2.0188121795654297,
+            "learning_rate": 0.0005,
+            "epoch": 0.5154090499106437,
+            "step": 11500
+        },
+        {
+            "eval_loss": 1.7547571659088135,
+            "eval_runtime": 18.5348,
+            "eval_samples_per_second": 883.96,
+            "eval_steps_per_second": 7.931,
+            "epoch": 0.5154090499106437,
+            "step": 11500
+        },
+        {
+            "loss": 14.0556,
+            "grad_norm": 1.9607020616531372,
+            "learning_rate": 0.0005,
+            "epoch": 0.5156331408019093,
+            "step": 11505
+        },
+        {
+            "loss": 14.0172,
+            "grad_norm": 1.7551989555358887,
+            "learning_rate": 0.0005,
+            "epoch": 0.5158572316931748,
+            "step": 11510
+        },
+        {
+            "loss": 14.0558,
+            "grad_norm": 1.7272204160690308,
+            "learning_rate": 0.0005,
+            "epoch": 0.5160813225844403,
+            "step": 11515
+        },
+        {
+            "loss": 14.035,
+            "grad_norm": 1.8389956951141357,
+            "learning_rate": 0.0005,
+            "epoch": 0.5163054134757058,
+            "step": 11520
+        },
+        {
+            "loss": 13.9865,
+            "grad_norm": 1.8806477785110474,
+            "learning_rate": 0.0005,
+            "epoch": 0.5165295043669712,
+            "step": 11525
+        },
+        {
+            "loss": 14.0794,
+            "grad_norm": 1.872300386428833,
+            "learning_rate": 0.0005,
+            "epoch": 0.5167535952582367,
+            "step": 11530
+        },
+        {
+            "loss": 14.0977,
+            "grad_norm": 1.9794795513153076,
+            "learning_rate": 0.0005,
+            "epoch": 0.5169776861495022,
+            "step": 11535
+        },
+        {
+            "loss": 13.9651,
+            "grad_norm": 1.8795273303985596,
+            "learning_rate": 0.0005,
+            "epoch": 0.5172017770407678,
+            "step": 11540
+        },
+        {
+            "loss": 14.0109,
+            "grad_norm": 1.9061027765274048,
+            "learning_rate": 0.0005,
+            "epoch": 0.5174258679320333,
+            "step": 11545
+        },
+        {
+            "loss": 14.0307,
+            "grad_norm": 1.7679489850997925,
+            "learning_rate": 0.0005,
+            "epoch": 0.5176499588232988,
+            "step": 11550
+        },
+        {
+            "loss": 14.0569,
+            "grad_norm": 1.9940497875213623,
+            "learning_rate": 0.0005,
+            "epoch": 0.5178740497145642,
+            "step": 11555
+        },
+        {
+            "loss": 13.9736,
+            "grad_norm": 1.8977187871932983,
+            "learning_rate": 0.0005,
+            "epoch": 0.5180981406058297,
+            "step": 11560
+        },
+        {
+            "loss": 14.0435,
+            "grad_norm": 1.9633815288543701,
+            "learning_rate": 0.0005,
+            "epoch": 0.5183222314970952,
+            "step": 11565
+        },
+        {
+            "loss": 14.0761,
+            "grad_norm": 1.9254745244979858,
+            "learning_rate": 0.0005,
+            "epoch": 0.5185463223883607,
+            "step": 11570
+        },
+        {
+            "loss": 13.9783,
+            "grad_norm": 1.8136787414550781,
+            "learning_rate": 0.0005,
+            "epoch": 0.5187704132796263,
+            "step": 11575
+        },
+        {
+            "loss": 14.0783,
+            "grad_norm": 1.8030261993408203,
+            "learning_rate": 0.0005,
+            "epoch": 0.5189945041708918,
+            "step": 11580
+        },
+        {
+            "loss": 13.9301,
+            "grad_norm": 1.7944817543029785,
+            "learning_rate": 0.0005,
+            "epoch": 0.5192185950621572,
+            "step": 11585
+        },
+        {
+            "loss": 14.003,
+            "grad_norm": 1.730994701385498,
+            "learning_rate": 0.0005,
+            "epoch": 0.5194426859534227,
+            "step": 11590
+        },
+        {
+            "loss": 13.936,
+            "grad_norm": 1.8809159994125366,
+            "learning_rate": 0.0005,
+            "epoch": 0.5196667768446882,
+            "step": 11595
+        },
+        {
+            "loss": 14.13,
+            "grad_norm": 1.805044174194336,
+            "learning_rate": 0.0005,
+            "epoch": 0.5198908677359537,
+            "step": 11600
+        },
+        {
+            "loss": 14.0864,
+            "grad_norm": 2.2456016540527344,
+            "learning_rate": 0.0005,
+            "epoch": 0.5201149586272192,
+            "step": 11605
+        },
+        {
+            "loss": 14.0473,
+            "grad_norm": 2.230912446975708,
+            "learning_rate": 0.0005,
+            "epoch": 0.5203390495184846,
+            "step": 11610
+        },
+        {
+            "loss": 14.0396,
+            "grad_norm": 2.18398380279541,
+            "learning_rate": 0.0005,
+            "epoch": 0.5205631404097502,
+            "step": 11615
+        },
+        {
+            "loss": 14.0842,
+            "grad_norm": 1.7159156799316406,
+            "learning_rate": 0.0005,
+            "epoch": 0.5207872313010157,
+            "step": 11620
+        },
+        {
+            "loss": 14.0063,
+            "grad_norm": 1.8257774114608765,
+            "learning_rate": 0.0005,
+            "epoch": 0.5210113221922812,
+            "step": 11625
+        },
+        {
+            "loss": 14.0822,
+            "grad_norm": 1.8890337944030762,
+            "learning_rate": 0.0005,
+            "epoch": 0.5212354130835467,
+            "step": 11630
+        },
+        {
+            "loss": 14.042,
+            "grad_norm": 1.9716954231262207,
+            "learning_rate": 0.0005,
+            "epoch": 0.5214595039748122,
+            "step": 11635
+        },
+        {
+            "loss": 14.055,
+            "grad_norm": 1.842434287071228,
+            "learning_rate": 0.0005,
+            "epoch": 0.5216835948660776,
+            "step": 11640
+        },
+        {
+            "loss": 13.9468,
+            "grad_norm": 1.8771275281906128,
+            "learning_rate": 0.0005,
+            "epoch": 0.5219076857573431,
+            "step": 11645
+        },
+        {
+            "loss": 14.0647,
+            "grad_norm": 1.7564702033996582,
+            "learning_rate": 0.0005,
+            "epoch": 0.5221317766486087,
+            "step": 11650
+        },
+        {
+            "loss": 14.0589,
+            "grad_norm": 1.9361367225646973,
+            "learning_rate": 0.0005,
+            "epoch": 0.5223558675398742,
+            "step": 11655
+        },
+        {
+            "loss": 14.0295,
+            "grad_norm": 1.8580527305603027,
+            "learning_rate": 0.0005,
+            "epoch": 0.5225799584311397,
+            "step": 11660
+        },
+        {
+            "loss": 14.1143,
+            "grad_norm": 1.9104259014129639,
+            "learning_rate": 0.0005,
+            "epoch": 0.5228040493224052,
+            "step": 11665
+        },
+        {
+            "loss": 14.1412,
+            "grad_norm": 1.8740001916885376,
+            "learning_rate": 0.0005,
+            "epoch": 0.5230281402136706,
+            "step": 11670
+        },
+        {
+            "loss": 13.9737,
+            "grad_norm": 1.9475232362747192,
+            "learning_rate": 0.0005,
+            "epoch": 0.5232522311049361,
+            "step": 11675
+        },
+        {
+            "loss": 14.0187,
+            "grad_norm": 1.9687209129333496,
+            "learning_rate": 0.0005,
+            "epoch": 0.5234763219962016,
+            "step": 11680
+        },
+        {
+            "loss": 14.0587,
+            "grad_norm": 1.7069392204284668,
+            "learning_rate": 0.0005,
+            "epoch": 0.5237004128874672,
+            "step": 11685
+        },
+        {
+            "loss": 14.1287,
+            "grad_norm": 1.883009672164917,
+            "learning_rate": 0.0005,
+            "epoch": 0.5239245037787327,
+            "step": 11690
+        },
+        {
+            "loss": 13.9363,
+            "grad_norm": 1.826299786567688,
+            "learning_rate": 0.0005,
+            "epoch": 0.5241485946699982,
+            "step": 11695
+        },
+        {
+            "loss": 14.1063,
+            "grad_norm": 3.701735019683838,
+            "learning_rate": 0.0005,
+            "epoch": 0.5243726855612636,
+            "step": 11700
+        },
+        {
+            "loss": 14.1886,
+            "grad_norm": 9.759357452392578,
+            "learning_rate": 0.0005,
+            "epoch": 0.5245967764525291,
+            "step": 11705
+        },
+        {
+            "loss": 14.0431,
+            "grad_norm": 3.50545334815979,
+            "learning_rate": 0.0005,
+            "epoch": 0.5248208673437946,
+            "step": 11710
+        },
+        {
+            "loss": 14.2454,
+            "grad_norm": 13.252138137817383,
+            "learning_rate": 0.0005,
+            "epoch": 0.5250449582350601,
+            "step": 11715
+        },
+        {
+            "loss": 14.1642,
+            "grad_norm": 3.8615827560424805,
+            "learning_rate": 0.0005,
+            "epoch": 0.5252690491263257,
+            "step": 11720
+        },
+        {
+            "loss": 14.1872,
+            "grad_norm": 3.313523530960083,
+            "learning_rate": 0.0005,
+            "epoch": 0.5254931400175912,
+            "step": 11725
+        },
+        {
+            "loss": 14.1133,
+            "grad_norm": 1.8248555660247803,
+            "learning_rate": 0.0005,
+            "epoch": 0.5257172309088566,
+            "step": 11730
+        },
+        {
+            "loss": 13.9341,
+            "grad_norm": 1.7867991924285889,
+            "learning_rate": 0.0005,
+            "epoch": 0.5259413218001221,
+            "step": 11735
+        },
+        {
+            "loss": 14.0033,
+            "grad_norm": 1.8857210874557495,
+            "learning_rate": 0.0005,
+            "epoch": 0.5261654126913876,
+            "step": 11740
+        },
+        {
+            "loss": 13.9379,
+            "grad_norm": 1.817320466041565,
+            "learning_rate": 0.0005,
+            "epoch": 0.5263895035826531,
+            "step": 11745
+        },
+        {
+            "loss": 14.0188,
+            "grad_norm": 1.8590130805969238,
+            "learning_rate": 0.0005,
+            "epoch": 0.5266135944739186,
+            "step": 11750
+        },
+        {
+            "loss": 13.9976,
+            "grad_norm": 1.9085701704025269,
+            "learning_rate": 0.0005,
+            "epoch": 0.5268376853651842,
+            "step": 11755
+        },
+        {
+            "loss": 14.0711,
+            "grad_norm": 2.036893367767334,
+            "learning_rate": 0.0005,
+            "epoch": 0.5270617762564496,
+            "step": 11760
+        },
+        {
+            "loss": 13.9832,
+            "grad_norm": 1.9735054969787598,
+            "learning_rate": 0.0005,
+            "epoch": 0.5272858671477151,
+            "step": 11765
+        },
+        {
+            "loss": 14.0891,
+            "grad_norm": 1.9286948442459106,
+            "learning_rate": 0.0005,
+            "epoch": 0.5275099580389806,
+            "step": 11770
+        },
+        {
+            "loss": 14.0208,
+            "grad_norm": 1.8106321096420288,
+            "learning_rate": 0.0005,
+            "epoch": 0.5277340489302461,
+            "step": 11775
+        },
+        {
+            "loss": 14.1278,
+            "grad_norm": 1.8342370986938477,
+            "learning_rate": 0.0005,
+            "epoch": 0.5279581398215116,
+            "step": 11780
+        },
+        {
+            "loss": 13.9797,
+            "grad_norm": 1.8223285675048828,
+            "learning_rate": 0.0005,
+            "epoch": 0.5281822307127771,
+            "step": 11785
+        },
+        {
+            "loss": 13.9417,
+            "grad_norm": 1.8281211853027344,
+            "learning_rate": 0.0005,
+            "epoch": 0.5284063216040426,
+            "step": 11790
+        },
+        {
+            "loss": 14.0578,
+            "grad_norm": 2.035158395767212,
+            "learning_rate": 0.0005,
+            "epoch": 0.5286304124953081,
+            "step": 11795
+        },
+        {
+            "loss": 14.0402,
+            "grad_norm": 1.8742728233337402,
+            "learning_rate": 0.0005,
+            "epoch": 0.5288545033865736,
+            "step": 11800
+        },
+        {
+            "loss": 13.9907,
+            "grad_norm": 2.1572251319885254,
+            "learning_rate": 0.0005,
+            "epoch": 0.5290785942778391,
+            "step": 11805
+        },
+        {
+            "loss": 14.012,
+            "grad_norm": 1.802402377128601,
+            "learning_rate": 0.0005,
+            "epoch": 0.5293026851691046,
+            "step": 11810
+        },
+        {
+            "loss": 14.0929,
+            "grad_norm": 1.780339241027832,
+            "learning_rate": 0.0005,
+            "epoch": 0.5295267760603701,
+            "step": 11815
+        },
+        {
+            "loss": 14.1459,
+            "grad_norm": 1.8276687860488892,
+            "learning_rate": 0.0005,
+            "epoch": 0.5297508669516355,
+            "step": 11820
+        },
+        {
+            "loss": 13.9739,
+            "grad_norm": 1.9510287046432495,
+            "learning_rate": 0.0005,
+            "epoch": 0.529974957842901,
+            "step": 11825
+        },
+        {
+            "loss": 14.044,
+            "grad_norm": 1.7390533685684204,
+            "learning_rate": 0.0005,
+            "epoch": 0.5301990487341666,
+            "step": 11830
+        },
+        {
+            "loss": 14.0421,
+            "grad_norm": 1.803524374961853,
+            "learning_rate": 0.0005,
+            "epoch": 0.5304231396254321,
+            "step": 11835
+        },
+        {
+            "loss": 13.8787,
+            "grad_norm": 2.0901153087615967,
+            "learning_rate": 0.0005,
+            "epoch": 0.5306472305166976,
+            "step": 11840
+        },
+        {
+            "loss": 14.0694,
+            "grad_norm": 1.9154999256134033,
+            "learning_rate": 0.0005,
+            "epoch": 0.5308713214079631,
+            "step": 11845
+        },
+        {
+            "loss": 13.9105,
+            "grad_norm": 1.770522117614746,
+            "learning_rate": 0.0005,
+            "epoch": 0.5310954122992285,
+            "step": 11850
+        },
+        {
+            "loss": 14.1578,
+            "grad_norm": 2.0752506256103516,
+            "learning_rate": 0.0005,
+            "epoch": 0.531319503190494,
+            "step": 11855
+        },
+        {
+            "loss": 14.0751,
+            "grad_norm": 2.0580227375030518,
+            "learning_rate": 0.0005,
+            "epoch": 0.5315435940817596,
+            "step": 11860
+        },
+        {
+            "loss": 14.0805,
+            "grad_norm": 1.8285835981369019,
+            "learning_rate": 0.0005,
+            "epoch": 0.5317676849730251,
+            "step": 11865
+        },
+        {
+            "loss": 14.0998,
+            "grad_norm": 1.7315351963043213,
+            "learning_rate": 0.0005,
+            "epoch": 0.5319917758642906,
+            "step": 11870
+        },
+        {
+            "loss": 14.052,
+            "grad_norm": 1.9043768644332886,
+            "learning_rate": 0.0005,
+            "epoch": 0.5322158667555561,
+            "step": 11875
+        },
+        {
+            "loss": 13.9807,
+            "grad_norm": 1.7757564783096313,
+            "learning_rate": 0.0005,
+            "epoch": 0.5324399576468215,
+            "step": 11880
+        },
+        {
+            "loss": 14.0045,
+            "grad_norm": 1.8369570970535278,
+            "learning_rate": 0.0005,
+            "epoch": 0.532664048538087,
+            "step": 11885
+        },
+        {
+            "loss": 14.0585,
+            "grad_norm": 1.9545503854751587,
+            "learning_rate": 0.0005,
+            "epoch": 0.5328881394293525,
+            "step": 11890
+        },
+        {
+            "loss": 14.0333,
+            "grad_norm": 2.004823923110962,
+            "learning_rate": 0.0005,
+            "epoch": 0.533112230320618,
+            "step": 11895
+        },
+        {
+            "loss": 14.0833,
+            "grad_norm": 2.157543659210205,
+            "learning_rate": 0.0005,
+            "epoch": 0.5333363212118836,
+            "step": 11900
+        },
+        {
+            "loss": 14.0551,
+            "grad_norm": 2.024017810821533,
+            "learning_rate": 0.0005,
+            "epoch": 0.5335604121031491,
+            "step": 11905
+        },
+        {
+            "loss": 14.0604,
+            "grad_norm": 1.8759499788284302,
+            "learning_rate": 0.0005,
+            "epoch": 0.5337845029944145,
+            "step": 11910
+        },
+        {
+            "loss": 13.9394,
+            "grad_norm": 1.8008873462677002,
+            "learning_rate": 0.0005,
+            "epoch": 0.53400859388568,
+            "step": 11915
+        },
+        {
+            "loss": 13.9449,
+            "grad_norm": 1.9857897758483887,
+            "learning_rate": 0.0005,
+            "epoch": 0.5342326847769455,
+            "step": 11920
+        },
+        {
+            "loss": 14.0237,
+            "grad_norm": 1.904971957206726,
+            "learning_rate": 0.0005,
+            "epoch": 0.534456775668211,
+            "step": 11925
+        },
+        {
+            "loss": 14.024,
+            "grad_norm": 1.889944314956665,
+            "learning_rate": 0.0005,
+            "epoch": 0.5346808665594766,
+            "step": 11930
+        },
+        {
+            "loss": 14.0295,
+            "grad_norm": 1.821826696395874,
+            "learning_rate": 0.0005,
+            "epoch": 0.5349049574507421,
+            "step": 11935
+        },
+        {
+            "loss": 14.0456,
+            "grad_norm": 1.8112114667892456,
+            "learning_rate": 0.0005,
+            "epoch": 0.5351290483420075,
+            "step": 11940
+        },
+        {
+            "loss": 14.0316,
+            "grad_norm": 1.7718886137008667,
+            "learning_rate": 0.0005,
+            "epoch": 0.535353139233273,
+            "step": 11945
+        },
+        {
+            "loss": 14.0575,
+            "grad_norm": 1.8424601554870605,
+            "learning_rate": 0.0005,
+            "epoch": 0.5355772301245385,
+            "step": 11950
+        },
+        {
+            "loss": 13.9971,
+            "grad_norm": 1.7665033340454102,
+            "learning_rate": 0.0005,
+            "epoch": 0.535801321015804,
+            "step": 11955
+        },
+        {
+            "loss": 13.8843,
+            "grad_norm": 1.8982579708099365,
+            "learning_rate": 0.0005,
+            "epoch": 0.5360254119070695,
+            "step": 11960
+        },
+        {
+            "loss": 14.069,
+            "grad_norm": 1.9286915063858032,
+            "learning_rate": 0.0005,
+            "epoch": 0.536249502798335,
+            "step": 11965
+        },
+        {
+            "loss": 14.0386,
+            "grad_norm": 1.8651976585388184,
+            "learning_rate": 0.0005,
+            "epoch": 0.5364735936896005,
+            "step": 11970
+        },
+        {
+            "loss": 14.0344,
+            "grad_norm": 1.8424943685531616,
+            "learning_rate": 0.0005,
+            "epoch": 0.536697684580866,
+            "step": 11975
+        },
+        {
+            "loss": 14.0157,
+            "grad_norm": 1.9398298263549805,
+            "learning_rate": 0.0005,
+            "epoch": 0.5369217754721315,
+            "step": 11980
+        },
+        {
+            "loss": 14.0076,
+            "grad_norm": 1.9768520593643188,
+            "learning_rate": 0.0005,
+            "epoch": 0.537145866363397,
+            "step": 11985
+        },
+        {
+            "loss": 14.0315,
+            "grad_norm": 1.8067823648452759,
+            "learning_rate": 0.0005,
+            "epoch": 0.5373699572546625,
+            "step": 11990
+        },
+        {
+            "loss": 13.976,
+            "grad_norm": 1.9012209177017212,
+            "learning_rate": 0.0005,
+            "epoch": 0.537594048145928,
+            "step": 11995
+        },
+        {
+            "loss": 13.9989,
+            "grad_norm": 1.8578073978424072,
+            "learning_rate": 0.0005,
+            "epoch": 0.5378181390371934,
+            "step": 12000
+        },
+        {
+            "eval_loss": 1.751168131828308,
+            "eval_runtime": 18.48,
+            "eval_samples_per_second": 886.579,
+            "eval_steps_per_second": 7.955,
+            "epoch": 0.5378181390371934,
+            "step": 12000
+        },
+        {
+            "loss": 13.9335,
+            "grad_norm": 1.706904649734497,
+            "learning_rate": 0.0005,
+            "epoch": 0.538042229928459,
+            "step": 12005
+        },
+        {
+            "loss": 14.1102,
+            "grad_norm": 1.930979609489441,
+            "learning_rate": 0.0005,
+            "epoch": 0.5382663208197245,
+            "step": 12010
+        },
+        {
+            "loss": 14.0288,
+            "grad_norm": 1.8886891603469849,
+            "learning_rate": 0.0005,
+            "epoch": 0.53849041171099,
+            "step": 12015
+        },
+        {
+            "loss": 14.0218,
+            "grad_norm": 1.8221163749694824,
+            "learning_rate": 0.0005,
+            "epoch": 0.5387145026022555,
+            "step": 12020
+        },
+        {
+            "loss": 14.122,
+            "grad_norm": 1.8023242950439453,
+            "learning_rate": 0.0005,
+            "epoch": 0.538938593493521,
+            "step": 12025
+        },
+        {
+            "loss": 14.0216,
+            "grad_norm": 1.832963466644287,
+            "learning_rate": 0.0005,
+            "epoch": 0.5391626843847864,
+            "step": 12030
+        },
+        {
+            "loss": 14.113,
+            "grad_norm": 2.0387954711914062,
+            "learning_rate": 0.0005,
+            "epoch": 0.5393867752760519,
+            "step": 12035
+        },
+        {
+            "loss": 14.0552,
+            "grad_norm": 1.916006326675415,
+            "learning_rate": 0.0005,
+            "epoch": 0.5396108661673175,
+            "step": 12040
+        },
+        {
+            "loss": 14.0393,
+            "grad_norm": 2.2164087295532227,
+            "learning_rate": 0.0005,
+            "epoch": 0.539834957058583,
+            "step": 12045
+        },
+        {
+            "loss": 13.9528,
+            "grad_norm": 2.0238280296325684,
+            "learning_rate": 0.0005,
+            "epoch": 0.5400590479498485,
+            "step": 12050
+        },
+        {
+            "loss": 14.0354,
+            "grad_norm": 2.0497937202453613,
+            "learning_rate": 0.0005,
+            "epoch": 0.540283138841114,
+            "step": 12055
+        },
+        {
+            "loss": 14.0743,
+            "grad_norm": 1.9421876668930054,
+            "learning_rate": 0.0005,
+            "epoch": 0.5405072297323794,
+            "step": 12060
+        },
+        {
+            "loss": 14.0372,
+            "grad_norm": 1.9722645282745361,
+            "learning_rate": 0.0005,
+            "epoch": 0.5407313206236449,
+            "step": 12065
+        },
+        {
+            "loss": 14.0676,
+            "grad_norm": 1.959842562675476,
+            "learning_rate": 0.0005,
+            "epoch": 0.5409554115149104,
+            "step": 12070
+        },
+        {
+            "loss": 14.0179,
+            "grad_norm": 1.9294716119766235,
+            "learning_rate": 0.0005,
+            "epoch": 0.541179502406176,
+            "step": 12075
+        },
+        {
+            "loss": 14.0412,
+            "grad_norm": 1.8364676237106323,
+            "learning_rate": 0.0005,
+            "epoch": 0.5414035932974415,
+            "step": 12080
+        },
+        {
+            "loss": 14.0165,
+            "grad_norm": 1.904807209968567,
+            "learning_rate": 0.0005,
+            "epoch": 0.541627684188707,
+            "step": 12085
+        },
+        {
+            "loss": 13.9646,
+            "grad_norm": 1.7190061807632446,
+            "learning_rate": 0.0005,
+            "epoch": 0.5418517750799724,
+            "step": 12090
+        },
+        {
+            "loss": 13.9799,
+            "grad_norm": 1.7632275819778442,
+            "learning_rate": 0.0005,
+            "epoch": 0.5420758659712379,
+            "step": 12095
+        },
+        {
+            "loss": 14.0236,
+            "grad_norm": 1.878212332725525,
+            "learning_rate": 0.0005,
+            "epoch": 0.5422999568625034,
+            "step": 12100
+        },
+        {
+            "loss": 14.0286,
+            "grad_norm": 1.9682413339614868,
+            "learning_rate": 0.0005,
+            "epoch": 0.542524047753769,
+            "step": 12105
+        },
+        {
+            "loss": 14.0145,
+            "grad_norm": 1.8260167837142944,
+            "learning_rate": 0.0005,
+            "epoch": 0.5427481386450345,
+            "step": 12110
+        },
+        {
+            "loss": 14.0664,
+            "grad_norm": 1.9433921575546265,
+            "learning_rate": 0.0005,
+            "epoch": 0.5429722295363,
+            "step": 12115
+        },
+        {
+            "loss": 14.0621,
+            "grad_norm": 1.9075546264648438,
+            "learning_rate": 0.0005,
+            "epoch": 0.5431963204275654,
+            "step": 12120
+        },
+        {
+            "loss": 14.01,
+            "grad_norm": 1.8555830717086792,
+            "learning_rate": 0.0005,
+            "epoch": 0.5434204113188309,
+            "step": 12125
+        },
+        {
+            "loss": 14.1147,
+            "grad_norm": 1.71388578414917,
+            "learning_rate": 0.0005,
+            "epoch": 0.5436445022100964,
+            "step": 12130
+        },
+        {
+            "loss": 14.0163,
+            "grad_norm": 1.8536070585250854,
+            "learning_rate": 0.0005,
+            "epoch": 0.5438685931013619,
+            "step": 12135
+        },
+        {
+            "loss": 13.968,
+            "grad_norm": 1.7627605199813843,
+            "learning_rate": 0.0005,
+            "epoch": 0.5440926839926274,
+            "step": 12140
+        },
+        {
+            "loss": 13.9629,
+            "grad_norm": 2.0195271968841553,
+            "learning_rate": 0.0005,
+            "epoch": 0.544316774883893,
+            "step": 12145
+        },
+        {
+            "loss": 14.0757,
+            "grad_norm": 1.9609084129333496,
+            "learning_rate": 0.0005,
+            "epoch": 0.5445408657751584,
+            "step": 12150
+        },
+        {
+            "loss": 14.0866,
+            "grad_norm": 1.967761516571045,
+            "learning_rate": 0.0005,
+            "epoch": 0.5447649566664239,
+            "step": 12155
+        },
+        {
+            "loss": 14.0538,
+            "grad_norm": 1.920175552368164,
+            "learning_rate": 0.0005,
+            "epoch": 0.5449890475576894,
+            "step": 12160
+        },
+        {
+            "loss": 14.0651,
+            "grad_norm": 1.8449351787567139,
+            "learning_rate": 0.0005,
+            "epoch": 0.5452131384489549,
+            "step": 12165
+        },
+        {
+            "loss": 13.9519,
+            "grad_norm": 1.8652801513671875,
+            "learning_rate": 0.0005,
+            "epoch": 0.5454372293402204,
+            "step": 12170
+        },
+        {
+            "loss": 14.1119,
+            "grad_norm": 1.987726092338562,
+            "learning_rate": 0.0005,
+            "epoch": 0.5456613202314858,
+            "step": 12175
+        },
+        {
+            "loss": 13.9927,
+            "grad_norm": 1.8179360628128052,
+            "learning_rate": 0.0005,
+            "epoch": 0.5458854111227514,
+            "step": 12180
+        },
+        {
+            "loss": 14.0653,
+            "grad_norm": 1.8283140659332275,
+            "learning_rate": 0.0005,
+            "epoch": 0.5461095020140169,
+            "step": 12185
+        },
+        {
+            "loss": 14.0795,
+            "grad_norm": 1.810027003288269,
+            "learning_rate": 0.0005,
+            "epoch": 0.5463335929052824,
+            "step": 12190
+        },
+        {
+            "loss": 13.9509,
+            "grad_norm": 1.8754284381866455,
+            "learning_rate": 0.0005,
+            "epoch": 0.5465576837965479,
+            "step": 12195
+        },
+        {
+            "loss": 13.9429,
+            "grad_norm": 1.7635506391525269,
+            "learning_rate": 0.0005,
+            "epoch": 0.5467817746878134,
+            "step": 12200
+        },
+        {
+            "loss": 13.9315,
+            "grad_norm": 1.898576259613037,
+            "learning_rate": 0.0005,
+            "epoch": 0.5470058655790788,
+            "step": 12205
+        },
+        {
+            "loss": 14.1612,
+            "grad_norm": 2.1952714920043945,
+            "learning_rate": 0.0005,
+            "epoch": 0.5472299564703443,
+            "step": 12210
+        },
+        {
+            "loss": 14.0364,
+            "grad_norm": 2.0558507442474365,
+            "learning_rate": 0.0005,
+            "epoch": 0.5474540473616099,
+            "step": 12215
+        },
+        {
+            "loss": 14.0311,
+            "grad_norm": 2.1676862239837646,
+            "learning_rate": 0.0005,
+            "epoch": 0.5476781382528754,
+            "step": 12220
+        },
+        {
+            "loss": 14.0522,
+            "grad_norm": 2.0368027687072754,
+            "learning_rate": 0.0005,
+            "epoch": 0.5479022291441409,
+            "step": 12225
+        },
+        {
+            "loss": 14.0133,
+            "grad_norm": 2.299630641937256,
+            "learning_rate": 0.0005,
+            "epoch": 0.5481263200354064,
+            "step": 12230
+        },
+        {
+            "loss": 14.0252,
+            "grad_norm": 2.0062880516052246,
+            "learning_rate": 0.0005,
+            "epoch": 0.5483504109266718,
+            "step": 12235
+        },
+        {
+            "loss": 13.9892,
+            "grad_norm": 1.8271337747573853,
+            "learning_rate": 0.0005,
+            "epoch": 0.5485745018179373,
+            "step": 12240
+        },
+        {
+            "loss": 13.8925,
+            "grad_norm": 1.8837248086929321,
+            "learning_rate": 0.0005,
+            "epoch": 0.5487985927092028,
+            "step": 12245
+        },
+        {
+            "loss": 14.0329,
+            "grad_norm": 1.819088339805603,
+            "learning_rate": 0.0005,
+            "epoch": 0.5490226836004684,
+            "step": 12250
+        },
+        {
+            "loss": 14.0457,
+            "grad_norm": 2.0993480682373047,
+            "learning_rate": 0.0005,
+            "epoch": 0.5492467744917339,
+            "step": 12255
+        },
+        {
+            "loss": 14.05,
+            "grad_norm": 2.045747995376587,
+            "learning_rate": 0.0005,
+            "epoch": 0.5494708653829994,
+            "step": 12260
+        },
+        {
+            "loss": 13.9964,
+            "grad_norm": 1.8817187547683716,
+            "learning_rate": 0.0005,
+            "epoch": 0.5496949562742648,
+            "step": 12265
+        },
+        {
+            "loss": 13.9463,
+            "grad_norm": 1.8474345207214355,
+            "learning_rate": 0.0005,
+            "epoch": 0.5499190471655303,
+            "step": 12270
+        },
+        {
+            "loss": 14.1182,
+            "grad_norm": 1.8616013526916504,
+            "learning_rate": 0.0005,
+            "epoch": 0.5501431380567958,
+            "step": 12275
+        },
+        {
+            "loss": 13.9633,
+            "grad_norm": 1.86726713180542,
+            "learning_rate": 0.0005,
+            "epoch": 0.5503672289480613,
+            "step": 12280
+        },
+        {
+            "loss": 14.1048,
+            "grad_norm": 1.8453407287597656,
+            "learning_rate": 0.0005,
+            "epoch": 0.5505913198393269,
+            "step": 12285
+        },
+        {
+            "loss": 14.087,
+            "grad_norm": 1.8242368698120117,
+            "learning_rate": 0.0005,
+            "epoch": 0.5508154107305924,
+            "step": 12290
+        },
+        {
+            "loss": 13.8923,
+            "grad_norm": 1.7882949113845825,
+            "learning_rate": 0.0005,
+            "epoch": 0.5510395016218578,
+            "step": 12295
+        },
+        {
+            "loss": 14.0861,
+            "grad_norm": 1.9016592502593994,
+            "learning_rate": 0.0005,
+            "epoch": 0.5512635925131233,
+            "step": 12300
+        },
+        {
+            "loss": 13.9075,
+            "grad_norm": 1.852830171585083,
+            "learning_rate": 0.0005,
+            "epoch": 0.5514876834043888,
+            "step": 12305
+        },
+        {
+            "loss": 14.0973,
+            "grad_norm": 1.9363442659378052,
+            "learning_rate": 0.0005,
+            "epoch": 0.5517117742956543,
+            "step": 12310
+        },
+        {
+            "loss": 14.1186,
+            "grad_norm": 1.948805570602417,
+            "learning_rate": 0.0005,
+            "epoch": 0.5519358651869198,
+            "step": 12315
+        },
+        {
+            "loss": 13.9078,
+            "grad_norm": 1.9615259170532227,
+            "learning_rate": 0.0005,
+            "epoch": 0.5521599560781854,
+            "step": 12320
+        },
+        {
+            "loss": 14.0245,
+            "grad_norm": 2.001514196395874,
+            "learning_rate": 0.0005,
+            "epoch": 0.5523840469694508,
+            "step": 12325
+        },
+        {
+            "loss": 14.0718,
+            "grad_norm": 1.716589331626892,
+            "learning_rate": 0.0005,
+            "epoch": 0.5526081378607163,
+            "step": 12330
+        },
+        {
+            "loss": 14.0069,
+            "grad_norm": 1.803367257118225,
+            "learning_rate": 0.0005,
+            "epoch": 0.5528322287519818,
+            "step": 12335
+        },
+        {
+            "loss": 13.9906,
+            "grad_norm": 1.8775421380996704,
+            "learning_rate": 0.0005,
+            "epoch": 0.5530563196432473,
+            "step": 12340
+        },
+        {
+            "loss": 14.0023,
+            "grad_norm": 1.7530593872070312,
+            "learning_rate": 0.0005,
+            "epoch": 0.5532804105345128,
+            "step": 12345
+        },
+        {
+            "loss": 14.0228,
+            "grad_norm": 1.8927412033081055,
+            "learning_rate": 0.0005,
+            "epoch": 0.5535045014257783,
+            "step": 12350
+        },
+        {
+            "loss": 14.0209,
+            "grad_norm": 1.9034690856933594,
+            "learning_rate": 0.0005,
+            "epoch": 0.5537285923170437,
+            "step": 12355
+        },
+        {
+            "loss": 14.0166,
+            "grad_norm": 1.8328973054885864,
+            "learning_rate": 0.0005,
+            "epoch": 0.5539526832083093,
+            "step": 12360
+        },
+        {
+            "loss": 13.9525,
+            "grad_norm": 1.8188306093215942,
+            "learning_rate": 0.0005,
+            "epoch": 0.5541767740995748,
+            "step": 12365
+        },
+        {
+            "loss": 14.0075,
+            "grad_norm": 2.0496511459350586,
+            "learning_rate": 0.0005,
+            "epoch": 0.5544008649908403,
+            "step": 12370
+        },
+        {
+            "loss": 14.0544,
+            "grad_norm": 1.826499581336975,
+            "learning_rate": 0.0005,
+            "epoch": 0.5546249558821058,
+            "step": 12375
+        },
+        {
+            "loss": 14.0,
+            "grad_norm": 1.9936883449554443,
+            "learning_rate": 0.0005,
+            "epoch": 0.5548490467733713,
+            "step": 12380
+        },
+        {
+            "loss": 14.0692,
+            "grad_norm": 1.9251879453659058,
+            "learning_rate": 0.0005,
+            "epoch": 0.5550731376646367,
+            "step": 12385
+        },
+        {
+            "loss": 13.9408,
+            "grad_norm": 1.8239604234695435,
+            "learning_rate": 0.0005,
+            "epoch": 0.5552972285559022,
+            "step": 12390
+        },
+        {
+            "loss": 14.0647,
+            "grad_norm": 1.909005880355835,
+            "learning_rate": 0.0005,
+            "epoch": 0.5555213194471678,
+            "step": 12395
+        },
+        {
+            "loss": 14.0004,
+            "grad_norm": 1.7871017456054688,
+            "learning_rate": 0.0005,
+            "epoch": 0.5557454103384333,
+            "step": 12400
+        },
+        {
+            "loss": 14.1342,
+            "grad_norm": 2.068019151687622,
+            "learning_rate": 0.0005,
+            "epoch": 0.5559695012296988,
+            "step": 12405
+        },
+        {
+            "loss": 14.0307,
+            "grad_norm": 1.9286525249481201,
+            "learning_rate": 0.0005,
+            "epoch": 0.5561935921209643,
+            "step": 12410
+        },
+        {
+            "loss": 14.0578,
+            "grad_norm": 2.0839881896972656,
+            "learning_rate": 0.0005,
+            "epoch": 0.5564176830122297,
+            "step": 12415
+        },
+        {
+            "loss": 13.97,
+            "grad_norm": 1.8886492252349854,
+            "learning_rate": 0.0005,
+            "epoch": 0.5566417739034952,
+            "step": 12420
+        },
+        {
+            "loss": 14.0677,
+            "grad_norm": 1.867583155632019,
+            "learning_rate": 0.0005,
+            "epoch": 0.5568658647947607,
+            "step": 12425
+        },
+        {
+            "loss": 13.9616,
+            "grad_norm": 2.073392629623413,
+            "learning_rate": 0.0005,
+            "epoch": 0.5570899556860263,
+            "step": 12430
+        },
+        {
+            "loss": 14.0997,
+            "grad_norm": 1.8922902345657349,
+            "learning_rate": 0.0005,
+            "epoch": 0.5573140465772918,
+            "step": 12435
+        },
+        {
+            "loss": 13.9662,
+            "grad_norm": 1.8526886701583862,
+            "learning_rate": 0.0005,
+            "epoch": 0.5575381374685573,
+            "step": 12440
+        },
+        {
+            "loss": 14.1219,
+            "grad_norm": 1.8532624244689941,
+            "learning_rate": 0.0005,
+            "epoch": 0.5577622283598227,
+            "step": 12445
+        },
+        {
+            "loss": 13.9798,
+            "grad_norm": 1.8773828744888306,
+            "learning_rate": 0.0005,
+            "epoch": 0.5579863192510882,
+            "step": 12450
+        },
+        {
+            "loss": 14.035,
+            "grad_norm": 1.787448525428772,
+            "learning_rate": 0.0005,
+            "epoch": 0.5582104101423537,
+            "step": 12455
+        },
+        {
+            "loss": 14.0232,
+            "grad_norm": 1.8511303663253784,
+            "learning_rate": 0.0005,
+            "epoch": 0.5584345010336192,
+            "step": 12460
+        },
+        {
+            "loss": 13.9491,
+            "grad_norm": 1.7718552350997925,
+            "learning_rate": 0.0005,
+            "epoch": 0.5586585919248848,
+            "step": 12465
+        },
+        {
+            "loss": 14.0324,
+            "grad_norm": 1.919750690460205,
+            "learning_rate": 0.0005,
+            "epoch": 0.5588826828161503,
+            "step": 12470
+        },
+        {
+            "loss": 14.0058,
+            "grad_norm": 1.8664422035217285,
+            "learning_rate": 0.0005,
+            "epoch": 0.5591067737074157,
+            "step": 12475
+        },
+        {
+            "loss": 13.9433,
+            "grad_norm": 2.035127878189087,
+            "learning_rate": 0.0005,
+            "epoch": 0.5593308645986812,
+            "step": 12480
+        },
+        {
+            "loss": 14.0902,
+            "grad_norm": 1.8490769863128662,
+            "learning_rate": 0.0005,
+            "epoch": 0.5595549554899467,
+            "step": 12485
+        },
+        {
+            "loss": 14.0749,
+            "grad_norm": 1.7316131591796875,
+            "learning_rate": 0.0005,
+            "epoch": 0.5597790463812122,
+            "step": 12490
+        },
+        {
+            "loss": 13.9822,
+            "grad_norm": 1.8705759048461914,
+            "learning_rate": 0.0005,
+            "epoch": 0.5600031372724777,
+            "step": 12495
+        },
+        {
+            "loss": 14.1043,
+            "grad_norm": 1.9594308137893677,
+            "learning_rate": 0.0005,
+            "epoch": 0.5602272281637433,
+            "step": 12500
+        },
+        {
+            "eval_loss": 1.751859426498413,
+            "eval_runtime": 18.223,
+            "eval_samples_per_second": 899.084,
+            "eval_steps_per_second": 8.067,
+            "epoch": 0.5602272281637433,
+            "step": 12500
+        },
+        {
+            "loss": 14.0527,
+            "grad_norm": 1.8572828769683838,
+            "learning_rate": 0.0005,
+            "epoch": 0.5604513190550087,
+            "step": 12505
+        },
+        {
+            "loss": 14.0881,
+            "grad_norm": 1.770039677619934,
+            "learning_rate": 0.0005,
+            "epoch": 0.5606754099462742,
+            "step": 12510
+        },
+        {
+            "loss": 14.0535,
+            "grad_norm": 1.951517939567566,
+            "learning_rate": 0.0005,
+            "epoch": 0.5608995008375397,
+            "step": 12515
+        },
+        {
+            "loss": 14.1481,
+            "grad_norm": 1.7481681108474731,
+            "learning_rate": 0.0005,
+            "epoch": 0.5611235917288052,
+            "step": 12520
+        },
+        {
+            "loss": 13.9009,
+            "grad_norm": 1.9862737655639648,
+            "learning_rate": 0.0005,
+            "epoch": 0.5613476826200707,
+            "step": 12525
+        },
+        {
+            "loss": 13.9339,
+            "grad_norm": 2.1382763385772705,
+            "learning_rate": 0.0005,
+            "epoch": 0.5615717735113362,
+            "step": 12530
+        },
+        {
+            "loss": 13.9717,
+            "grad_norm": 1.9665172100067139,
+            "learning_rate": 0.0005,
+            "epoch": 0.5617958644026017,
+            "step": 12535
+        },
+        {
+            "loss": 14.0528,
+            "grad_norm": 1.9153531789779663,
+            "learning_rate": 0.0005,
+            "epoch": 0.5620199552938672,
+            "step": 12540
+        },
+        {
+            "loss": 13.888,
+            "grad_norm": 1.8589550256729126,
+            "learning_rate": 0.0005,
+            "epoch": 0.5622440461851327,
+            "step": 12545
+        },
+        {
+            "loss": 13.9988,
+            "grad_norm": 1.7779061794281006,
+            "learning_rate": 0.0005,
+            "epoch": 0.5624681370763982,
+            "step": 12550
+        },
+        {
+            "loss": 13.9705,
+            "grad_norm": 1.8984662294387817,
+            "learning_rate": 0.0005,
+            "epoch": 0.5626922279676637,
+            "step": 12555
+        },
+        {
+            "loss": 14.0459,
+            "grad_norm": 2.0568203926086426,
+            "learning_rate": 0.0005,
+            "epoch": 0.5629163188589292,
+            "step": 12560
+        },
+        {
+            "loss": 14.074,
+            "grad_norm": 2.0921597480773926,
+            "learning_rate": 0.0005,
+            "epoch": 0.5631404097501946,
+            "step": 12565
+        },
+        {
+            "loss": 14.0376,
+            "grad_norm": 1.914757251739502,
+            "learning_rate": 0.0005,
+            "epoch": 0.5633645006414602,
+            "step": 12570
+        },
+        {
+            "loss": 14.0212,
+            "grad_norm": 2.220177173614502,
+            "learning_rate": 0.0005,
+            "epoch": 0.5635885915327257,
+            "step": 12575
+        },
+        {
+            "loss": 13.929,
+            "grad_norm": 1.9564307928085327,
+            "learning_rate": 0.0005,
+            "epoch": 0.5638126824239912,
+            "step": 12580
+        },
+        {
+            "loss": 14.1402,
+            "grad_norm": 2.0993082523345947,
+            "learning_rate": 0.0005,
+            "epoch": 0.5640367733152567,
+            "step": 12585
+        },
+        {
+            "loss": 14.0034,
+            "grad_norm": 2.193251609802246,
+            "learning_rate": 0.0005,
+            "epoch": 0.5642608642065222,
+            "step": 12590
+        },
+        {
+            "loss": 14.013,
+            "grad_norm": 1.8814992904663086,
+            "learning_rate": 0.0005,
+            "epoch": 0.5644849550977876,
+            "step": 12595
+        },
+        {
+            "loss": 14.0452,
+            "grad_norm": 1.926684021949768,
+            "learning_rate": 0.0005,
+            "epoch": 0.5647090459890531,
+            "step": 12600
+        },
+        {
+            "loss": 13.9817,
+            "grad_norm": 1.8195796012878418,
+            "learning_rate": 0.0005,
+            "epoch": 0.5649331368803187,
+            "step": 12605
+        },
+        {
+            "loss": 14.0128,
+            "grad_norm": 1.8257880210876465,
+            "learning_rate": 0.0005,
+            "epoch": 0.5651572277715842,
+            "step": 12610
+        },
+        {
+            "loss": 13.8912,
+            "grad_norm": 1.9664603471755981,
+            "learning_rate": 0.0005,
+            "epoch": 0.5653813186628497,
+            "step": 12615
+        },
+        {
+            "loss": 14.0476,
+            "grad_norm": 1.8703941106796265,
+            "learning_rate": 0.0005,
+            "epoch": 0.5656054095541152,
+            "step": 12620
+        },
+        {
+            "loss": 13.9546,
+            "grad_norm": 1.8966342210769653,
+            "learning_rate": 0.0005,
+            "epoch": 0.5658295004453806,
+            "step": 12625
+        },
+        {
+            "loss": 14.0268,
+            "grad_norm": 1.9538925886154175,
+            "learning_rate": 0.0005,
+            "epoch": 0.5660535913366461,
+            "step": 12630
+        },
+        {
+            "loss": 13.9529,
+            "grad_norm": 1.9661394357681274,
+            "learning_rate": 0.0005,
+            "epoch": 0.5662776822279116,
+            "step": 12635
+        },
+        {
+            "loss": 14.0026,
+            "grad_norm": 1.824487566947937,
+            "learning_rate": 0.0005,
+            "epoch": 0.5665017731191772,
+            "step": 12640
+        },
+        {
+            "loss": 13.9676,
+            "grad_norm": 1.8401505947113037,
+            "learning_rate": 0.0005,
+            "epoch": 0.5667258640104427,
+            "step": 12645
+        },
+        {
+            "loss": 13.9838,
+            "grad_norm": 1.9098906517028809,
+            "learning_rate": 0.0005,
+            "epoch": 0.5669499549017082,
+            "step": 12650
+        },
+        {
+            "loss": 13.946,
+            "grad_norm": 1.9357730150222778,
+            "learning_rate": 0.0005,
+            "epoch": 0.5671740457929736,
+            "step": 12655
+        },
+        {
+            "loss": 13.9252,
+            "grad_norm": 1.9102228879928589,
+            "learning_rate": 0.0005,
+            "epoch": 0.5673981366842391,
+            "step": 12660
+        },
+        {
+            "loss": 14.0503,
+            "grad_norm": 1.755408525466919,
+            "learning_rate": 0.0005,
+            "epoch": 0.5676222275755046,
+            "step": 12665
+        },
+        {
+            "loss": 13.9132,
+            "grad_norm": 1.8137500286102295,
+            "learning_rate": 0.0005,
+            "epoch": 0.5678463184667701,
+            "step": 12670
+        },
+        {
+            "loss": 14.0714,
+            "grad_norm": 1.842359185218811,
+            "learning_rate": 0.0005,
+            "epoch": 0.5680704093580357,
+            "step": 12675
+        },
+        {
+            "loss": 13.9779,
+            "grad_norm": 1.7982245683670044,
+            "learning_rate": 0.0005,
+            "epoch": 0.5682945002493012,
+            "step": 12680
+        },
+        {
+            "loss": 14.1149,
+            "grad_norm": 1.9560648202896118,
+            "learning_rate": 0.0005,
+            "epoch": 0.5685185911405666,
+            "step": 12685
+        },
+        {
+            "loss": 13.9683,
+            "grad_norm": 1.90084707736969,
+            "learning_rate": 0.0005,
+            "epoch": 0.5687426820318321,
+            "step": 12690
+        },
+        {
+            "loss": 14.0096,
+            "grad_norm": 1.897828459739685,
+            "learning_rate": 0.0005,
+            "epoch": 0.5689667729230976,
+            "step": 12695
+        },
+        {
+            "loss": 14.0001,
+            "grad_norm": 1.7941429615020752,
+            "learning_rate": 0.0005,
+            "epoch": 0.5691908638143631,
+            "step": 12700
+        },
+        {
+            "loss": 14.0167,
+            "grad_norm": 1.9507100582122803,
+            "learning_rate": 0.0005,
+            "epoch": 0.5694149547056286,
+            "step": 12705
+        },
+        {
+            "loss": 14.0317,
+            "grad_norm": 1.8001422882080078,
+            "learning_rate": 0.0005,
+            "epoch": 0.5696390455968942,
+            "step": 12710
+        },
+        {
+            "loss": 14.0206,
+            "grad_norm": 1.9080332517623901,
+            "learning_rate": 0.0005,
+            "epoch": 0.5698631364881596,
+            "step": 12715
+        },
+        {
+            "loss": 14.0638,
+            "grad_norm": 1.915358304977417,
+            "learning_rate": 0.0005,
+            "epoch": 0.5700872273794251,
+            "step": 12720
+        },
+        {
+            "loss": 14.0261,
+            "grad_norm": 1.8281325101852417,
+            "learning_rate": 0.0005,
+            "epoch": 0.5703113182706906,
+            "step": 12725
+        },
+        {
+            "loss": 14.0883,
+            "grad_norm": 2.0495681762695312,
+            "learning_rate": 0.0005,
+            "epoch": 0.5705354091619561,
+            "step": 12730
+        },
+        {
+            "loss": 14.0626,
+            "grad_norm": 1.7946007251739502,
+            "learning_rate": 0.0005,
+            "epoch": 0.5707595000532216,
+            "step": 12735
+        },
+        {
+            "loss": 14.0204,
+            "grad_norm": 1.8077770471572876,
+            "learning_rate": 0.0005,
+            "epoch": 0.570983590944487,
+            "step": 12740
+        },
+        {
+            "loss": 13.9629,
+            "grad_norm": 1.7210853099822998,
+            "learning_rate": 0.0005,
+            "epoch": 0.5712076818357525,
+            "step": 12745
+        },
+        {
+            "loss": 13.9486,
+            "grad_norm": 1.8925403356552124,
+            "learning_rate": 0.0005,
+            "epoch": 0.5714317727270181,
+            "step": 12750
+        },
+        {
+            "loss": 13.8341,
+            "grad_norm": 1.7555067539215088,
+            "learning_rate": 0.0005,
+            "epoch": 0.5716558636182836,
+            "step": 12755
+        },
+        {
+            "loss": 13.9755,
+            "grad_norm": 1.6934796571731567,
+            "learning_rate": 0.0005,
+            "epoch": 0.5718799545095491,
+            "step": 12760
+        },
+        {
+            "loss": 14.0371,
+            "grad_norm": 1.8694788217544556,
+            "learning_rate": 0.0005,
+            "epoch": 0.5721040454008146,
+            "step": 12765
+        },
+        {
+            "loss": 14.0642,
+            "grad_norm": 1.8269613981246948,
+            "learning_rate": 0.0005,
+            "epoch": 0.57232813629208,
+            "step": 12770
+        },
+        {
+            "loss": 13.9915,
+            "grad_norm": 2.02119517326355,
+            "learning_rate": 0.0005,
+            "epoch": 0.5725522271833455,
+            "step": 12775
+        },
+        {
+            "loss": 13.8738,
+            "grad_norm": 1.83871328830719,
+            "learning_rate": 0.0005,
+            "epoch": 0.572776318074611,
+            "step": 12780
+        },
+        {
+            "loss": 14.0301,
+            "grad_norm": 1.9396083354949951,
+            "learning_rate": 0.0005,
+            "epoch": 0.5730004089658766,
+            "step": 12785
+        },
+        {
+            "loss": 14.1057,
+            "grad_norm": 1.9239839315414429,
+            "learning_rate": 0.0005,
+            "epoch": 0.5732244998571421,
+            "step": 12790
+        },
+        {
+            "loss": 14.0808,
+            "grad_norm": 1.8804762363433838,
+            "learning_rate": 0.0005,
+            "epoch": 0.5734485907484076,
+            "step": 12795
+        },
+        {
+            "loss": 14.083,
+            "grad_norm": 1.7791146039962769,
+            "learning_rate": 0.0005,
+            "epoch": 0.573672681639673,
+            "step": 12800
+        },
+        {
+            "loss": 13.977,
+            "grad_norm": 1.9019142389297485,
+            "learning_rate": 0.0005,
+            "epoch": 0.5738967725309385,
+            "step": 12805
+        },
+        {
+            "loss": 13.9416,
+            "grad_norm": 1.6902880668640137,
+            "learning_rate": 0.0005,
+            "epoch": 0.574120863422204,
+            "step": 12810
+        },
+        {
+            "loss": 13.9604,
+            "grad_norm": 1.79051673412323,
+            "learning_rate": 0.0005,
+            "epoch": 0.5743449543134695,
+            "step": 12815
+        },
+        {
+            "loss": 14.0087,
+            "grad_norm": 2.142242431640625,
+            "learning_rate": 0.0005,
+            "epoch": 0.5745690452047351,
+            "step": 12820
+        },
+        {
+            "loss": 14.189,
+            "grad_norm": 2.034118890762329,
+            "learning_rate": 0.0005,
+            "epoch": 0.5747931360960006,
+            "step": 12825
+        },
+        {
+            "loss": 14.0089,
+            "grad_norm": 1.8687961101531982,
+            "learning_rate": 0.0005,
+            "epoch": 0.575017226987266,
+            "step": 12830
+        },
+        {
+            "loss": 13.9777,
+            "grad_norm": 2.0446767807006836,
+            "learning_rate": 0.0005,
+            "epoch": 0.5752413178785315,
+            "step": 12835
+        },
+        {
+            "loss": 13.9677,
+            "grad_norm": 1.864727258682251,
+            "learning_rate": 0.0005,
+            "epoch": 0.575465408769797,
+            "step": 12840
+        },
+        {
+            "loss": 13.9624,
+            "grad_norm": 1.9575270414352417,
+            "learning_rate": 0.0005,
+            "epoch": 0.5756894996610625,
+            "step": 12845
+        },
+        {
+            "loss": 14.0536,
+            "grad_norm": 1.8159977197647095,
+            "learning_rate": 0.0005,
+            "epoch": 0.575913590552328,
+            "step": 12850
+        },
+        {
+            "loss": 13.9992,
+            "grad_norm": 1.8523368835449219,
+            "learning_rate": 0.0005,
+            "epoch": 0.5761376814435936,
+            "step": 12855
+        },
+        {
+            "loss": 13.9665,
+            "grad_norm": 1.796653151512146,
+            "learning_rate": 0.0005,
+            "epoch": 0.576361772334859,
+            "step": 12860
+        },
+        {
+            "loss": 14.0113,
+            "grad_norm": 2.0082266330718994,
+            "learning_rate": 0.0005,
+            "epoch": 0.5765858632261245,
+            "step": 12865
+        },
+        {
+            "loss": 14.0491,
+            "grad_norm": 1.9820599555969238,
+            "learning_rate": 0.0005,
+            "epoch": 0.57680995411739,
+            "step": 12870
+        },
+        {
+            "loss": 14.0307,
+            "grad_norm": 1.8778185844421387,
+            "learning_rate": 0.0005,
+            "epoch": 0.5770340450086555,
+            "step": 12875
+        },
+        {
+            "loss": 13.9243,
+            "grad_norm": 1.9270992279052734,
+            "learning_rate": 0.0005,
+            "epoch": 0.577258135899921,
+            "step": 12880
+        },
+        {
+            "loss": 14.1251,
+            "grad_norm": 1.9180136919021606,
+            "learning_rate": 0.0005,
+            "epoch": 0.5774822267911865,
+            "step": 12885
+        },
+        {
+            "loss": 13.9821,
+            "grad_norm": 1.7854799032211304,
+            "learning_rate": 0.0005,
+            "epoch": 0.577706317682452,
+            "step": 12890
+        },
+        {
+            "loss": 14.0954,
+            "grad_norm": 1.7735947370529175,
+            "learning_rate": 0.0005,
+            "epoch": 0.5779304085737175,
+            "step": 12895
+        },
+        {
+            "loss": 13.9303,
+            "grad_norm": 2.0620439052581787,
+            "learning_rate": 0.0005,
+            "epoch": 0.578154499464983,
+            "step": 12900
+        },
+        {
+            "loss": 14.1653,
+            "grad_norm": 2.0477583408355713,
+            "learning_rate": 0.0005,
+            "epoch": 0.5783785903562485,
+            "step": 12905
+        },
+        {
+            "loss": 13.9222,
+            "grad_norm": 2.063283681869507,
+            "learning_rate": 0.0005,
+            "epoch": 0.578602681247514,
+            "step": 12910
+        },
+        {
+            "loss": 13.963,
+            "grad_norm": 1.889586091041565,
+            "learning_rate": 0.0005,
+            "epoch": 0.5788267721387795,
+            "step": 12915
+        },
+        {
+            "loss": 14.0448,
+            "grad_norm": 1.8375580310821533,
+            "learning_rate": 0.0005,
+            "epoch": 0.5790508630300449,
+            "step": 12920
+        },
+        {
+            "loss": 13.9869,
+            "grad_norm": 1.8818038702011108,
+            "learning_rate": 0.0005,
+            "epoch": 0.5792749539213105,
+            "step": 12925
+        },
+        {
+            "loss": 13.9813,
+            "grad_norm": 1.8376609086990356,
+            "learning_rate": 0.0005,
+            "epoch": 0.579499044812576,
+            "step": 12930
+        },
+        {
+            "loss": 13.9736,
+            "grad_norm": 1.9093542098999023,
+            "learning_rate": 0.0005,
+            "epoch": 0.5797231357038415,
+            "step": 12935
+        },
+        {
+            "loss": 13.9723,
+            "grad_norm": 2.288410186767578,
+            "learning_rate": 0.0005,
+            "epoch": 0.579947226595107,
+            "step": 12940
+        },
+        {
+            "loss": 14.0652,
+            "grad_norm": 1.796730875968933,
+            "learning_rate": 0.0005,
+            "epoch": 0.5801713174863725,
+            "step": 12945
+        },
+        {
+            "loss": 13.9637,
+            "grad_norm": 1.9128367900848389,
+            "learning_rate": 0.0005,
+            "epoch": 0.5803954083776379,
+            "step": 12950
+        },
+        {
+            "loss": 13.9768,
+            "grad_norm": 1.967447280883789,
+            "learning_rate": 0.0005,
+            "epoch": 0.5806194992689034,
+            "step": 12955
+        },
+        {
+            "loss": 14.0629,
+            "grad_norm": 1.8460129499435425,
+            "learning_rate": 0.0005,
+            "epoch": 0.580843590160169,
+            "step": 12960
+        },
+        {
+            "loss": 14.0869,
+            "grad_norm": 2.1665172576904297,
+            "learning_rate": 0.0005,
+            "epoch": 0.5810676810514345,
+            "step": 12965
+        },
+        {
+            "loss": 14.0342,
+            "grad_norm": 1.8314615488052368,
+            "learning_rate": 0.0005,
+            "epoch": 0.5812917719427,
+            "step": 12970
+        },
+        {
+            "loss": 13.8923,
+            "grad_norm": 1.8446085453033447,
+            "learning_rate": 0.0005,
+            "epoch": 0.5815158628339655,
+            "step": 12975
+        },
+        {
+            "loss": 13.9526,
+            "grad_norm": 1.7681913375854492,
+            "learning_rate": 0.0005,
+            "epoch": 0.5817399537252309,
+            "step": 12980
+        },
+        {
+            "loss": 14.0395,
+            "grad_norm": 1.8649137020111084,
+            "learning_rate": 0.0005,
+            "epoch": 0.5819640446164964,
+            "step": 12985
+        },
+        {
+            "loss": 14.0145,
+            "grad_norm": 1.7946351766586304,
+            "learning_rate": 0.0005,
+            "epoch": 0.5821881355077619,
+            "step": 12990
+        },
+        {
+            "loss": 14.1076,
+            "grad_norm": 1.9252318143844604,
+            "learning_rate": 0.0005,
+            "epoch": 0.5824122263990275,
+            "step": 12995
+        },
+        {
+            "loss": 13.9776,
+            "grad_norm": 1.826115608215332,
+            "learning_rate": 0.0005,
+            "epoch": 0.582636317290293,
+            "step": 13000
+        },
+        {
+            "eval_loss": 1.7454519271850586,
+            "eval_runtime": 18.6945,
+            "eval_samples_per_second": 876.406,
+            "eval_steps_per_second": 7.863,
+            "epoch": 0.582636317290293,
+            "step": 13000
+        },
+        {
+            "loss": 13.9629,
+            "grad_norm": 1.9778342247009277,
+            "learning_rate": 0.0005,
+            "epoch": 0.5828604081815585,
+            "step": 13005
+        },
+        {
+            "loss": 14.0724,
+            "grad_norm": 1.7861636877059937,
+            "learning_rate": 0.0005,
+            "epoch": 0.5830844990728239,
+            "step": 13010
+        },
+        {
+            "loss": 13.9045,
+            "grad_norm": 1.8089948892593384,
+            "learning_rate": 0.0005,
+            "epoch": 0.5833085899640894,
+            "step": 13015
+        },
+        {
+            "loss": 14.0613,
+            "grad_norm": 1.9273558855056763,
+            "learning_rate": 0.0005,
+            "epoch": 0.5835326808553549,
+            "step": 13020
+        },
+        {
+            "loss": 13.9659,
+            "grad_norm": 1.7711695432662964,
+            "learning_rate": 0.0005,
+            "epoch": 0.5837567717466204,
+            "step": 13025
+        },
+        {
+            "loss": 14.0277,
+            "grad_norm": 1.847944974899292,
+            "learning_rate": 0.0005,
+            "epoch": 0.583980862637886,
+            "step": 13030
+        },
+        {
+            "loss": 13.8823,
+            "grad_norm": 1.7399357557296753,
+            "learning_rate": 0.0005,
+            "epoch": 0.5842049535291515,
+            "step": 13035
+        },
+        {
+            "loss": 13.988,
+            "grad_norm": 1.7723686695098877,
+            "learning_rate": 0.0005,
+            "epoch": 0.5844290444204169,
+            "step": 13040
+        },
+        {
+            "loss": 13.9147,
+            "grad_norm": 1.7638523578643799,
+            "learning_rate": 0.0005,
+            "epoch": 0.5846531353116824,
+            "step": 13045
+        },
+        {
+            "loss": 14.0222,
+            "grad_norm": 1.7969489097595215,
+            "learning_rate": 0.0005,
+            "epoch": 0.5848772262029479,
+            "step": 13050
+        },
+        {
+            "loss": 14.0576,
+            "grad_norm": 1.8615041971206665,
+            "learning_rate": 0.0005,
+            "epoch": 0.5851013170942134,
+            "step": 13055
+        },
+        {
+            "loss": 13.9736,
+            "grad_norm": 1.8189938068389893,
+            "learning_rate": 0.0005,
+            "epoch": 0.5853254079854789,
+            "step": 13060
+        },
+        {
+            "loss": 13.9696,
+            "grad_norm": 1.829389214515686,
+            "learning_rate": 0.0005,
+            "epoch": 0.5855494988767445,
+            "step": 13065
+        },
+        {
+            "loss": 14.008,
+            "grad_norm": 1.8700202703475952,
+            "learning_rate": 0.0005,
+            "epoch": 0.5857735897680099,
+            "step": 13070
+        },
+        {
+            "loss": 14.0049,
+            "grad_norm": 1.7231286764144897,
+            "learning_rate": 0.0005,
+            "epoch": 0.5859976806592754,
+            "step": 13075
+        },
+        {
+            "loss": 13.9735,
+            "grad_norm": 2.0502021312713623,
+            "learning_rate": 0.0005,
+            "epoch": 0.5862217715505409,
+            "step": 13080
+        },
+        {
+            "loss": 14.0124,
+            "grad_norm": 2.1990959644317627,
+            "learning_rate": 0.0005,
+            "epoch": 0.5864458624418064,
+            "step": 13085
+        },
+        {
+            "loss": 14.1032,
+            "grad_norm": 1.902991771697998,
+            "learning_rate": 0.0005,
+            "epoch": 0.5866699533330719,
+            "step": 13090
+        },
+        {
+            "loss": 13.9135,
+            "grad_norm": 1.738110065460205,
+            "learning_rate": 0.0005,
+            "epoch": 0.5868940442243374,
+            "step": 13095
+        },
+        {
+            "loss": 14.0366,
+            "grad_norm": 1.736744999885559,
+            "learning_rate": 0.0005,
+            "epoch": 0.5871181351156028,
+            "step": 13100
+        },
+        {
+            "loss": 14.0341,
+            "grad_norm": 1.728366732597351,
+            "learning_rate": 0.0005,
+            "epoch": 0.5873422260068684,
+            "step": 13105
+        },
+        {
+            "loss": 14.0086,
+            "grad_norm": 1.9115058183670044,
+            "learning_rate": 0.0005,
+            "epoch": 0.5875663168981339,
+            "step": 13110
+        },
+        {
+            "loss": 13.9843,
+            "grad_norm": 1.8320019245147705,
+            "learning_rate": 0.0005,
+            "epoch": 0.5877904077893994,
+            "step": 13115
+        },
+        {
+            "loss": 13.9944,
+            "grad_norm": 1.8176792860031128,
+            "learning_rate": 0.0005,
+            "epoch": 0.5880144986806649,
+            "step": 13120
+        },
+        {
+            "loss": 14.0356,
+            "grad_norm": 1.802194356918335,
+            "learning_rate": 0.0005,
+            "epoch": 0.5882385895719304,
+            "step": 13125
+        },
+        {
+            "loss": 13.9914,
+            "grad_norm": 1.7907404899597168,
+            "learning_rate": 0.0005,
+            "epoch": 0.5884626804631958,
+            "step": 13130
+        },
+        {
+            "loss": 13.9473,
+            "grad_norm": 1.9506714344024658,
+            "learning_rate": 0.0005,
+            "epoch": 0.5886867713544613,
+            "step": 13135
+        },
+        {
+            "loss": 13.9552,
+            "grad_norm": 1.9168850183486938,
+            "learning_rate": 0.0005,
+            "epoch": 0.5889108622457269,
+            "step": 13140
+        },
+        {
+            "loss": 14.1582,
+            "grad_norm": 1.8277353048324585,
+            "learning_rate": 0.0005,
+            "epoch": 0.5891349531369924,
+            "step": 13145
+        },
+        {
+            "loss": 14.0282,
+            "grad_norm": 1.944557785987854,
+            "learning_rate": 0.0005,
+            "epoch": 0.5893590440282579,
+            "step": 13150
+        },
+        {
+            "loss": 14.017,
+            "grad_norm": 1.851028323173523,
+            "learning_rate": 0.0005,
+            "epoch": 0.5895831349195234,
+            "step": 13155
+        },
+        {
+            "loss": 14.046,
+            "grad_norm": 1.9107221364974976,
+            "learning_rate": 0.0005,
+            "epoch": 0.5898072258107888,
+            "step": 13160
+        },
+        {
+            "loss": 13.96,
+            "grad_norm": 1.908125638961792,
+            "learning_rate": 0.0005,
+            "epoch": 0.5900313167020543,
+            "step": 13165
+        },
+        {
+            "loss": 14.0432,
+            "grad_norm": 1.729802131652832,
+            "learning_rate": 0.0005,
+            "epoch": 0.5902554075933198,
+            "step": 13170
+        },
+        {
+            "loss": 13.8967,
+            "grad_norm": 1.9202781915664673,
+            "learning_rate": 0.0005,
+            "epoch": 0.5904794984845854,
+            "step": 13175
+        },
+        {
+            "loss": 13.9397,
+            "grad_norm": 1.7435154914855957,
+            "learning_rate": 0.0005,
+            "epoch": 0.5907035893758509,
+            "step": 13180
+        },
+        {
+            "loss": 13.9726,
+            "grad_norm": 1.8927922248840332,
+            "learning_rate": 0.0005,
+            "epoch": 0.5909276802671164,
+            "step": 13185
+        },
+        {
+            "loss": 13.9359,
+            "grad_norm": 1.952462911605835,
+            "learning_rate": 0.0005,
+            "epoch": 0.5911517711583818,
+            "step": 13190
+        },
+        {
+            "loss": 13.9552,
+            "grad_norm": 1.815459132194519,
+            "learning_rate": 0.0005,
+            "epoch": 0.5913758620496473,
+            "step": 13195
+        },
+        {
+            "loss": 13.9568,
+            "grad_norm": 1.8877557516098022,
+            "learning_rate": 0.0005,
+            "epoch": 0.5915999529409128,
+            "step": 13200
+        },
+        {
+            "loss": 13.9949,
+            "grad_norm": 1.7916780710220337,
+            "learning_rate": 0.0005,
+            "epoch": 0.5918240438321783,
+            "step": 13205
+        },
+        {
+            "loss": 14.0393,
+            "grad_norm": 1.8304307460784912,
+            "learning_rate": 0.0005,
+            "epoch": 0.5920481347234439,
+            "step": 13210
+        },
+        {
+            "loss": 14.0511,
+            "grad_norm": 1.842038869857788,
+            "learning_rate": 0.0005,
+            "epoch": 0.5922722256147094,
+            "step": 13215
+        },
+        {
+            "loss": 13.9725,
+            "grad_norm": 1.7629725933074951,
+            "learning_rate": 0.0005,
+            "epoch": 0.5924963165059748,
+            "step": 13220
+        },
+        {
+            "loss": 14.0556,
+            "grad_norm": 1.8662147521972656,
+            "learning_rate": 0.0005,
+            "epoch": 0.5927204073972403,
+            "step": 13225
+        },
+        {
+            "loss": 14.1007,
+            "grad_norm": 1.7180095911026,
+            "learning_rate": 0.0005,
+            "epoch": 0.5929444982885058,
+            "step": 13230
+        },
+        {
+            "loss": 14.0402,
+            "grad_norm": 1.8103516101837158,
+            "learning_rate": 0.0005,
+            "epoch": 0.5931685891797713,
+            "step": 13235
+        },
+        {
+            "loss": 13.9343,
+            "grad_norm": 1.8062591552734375,
+            "learning_rate": 0.0005,
+            "epoch": 0.5933926800710368,
+            "step": 13240
+        },
+        {
+            "loss": 14.0915,
+            "grad_norm": 1.8626552820205688,
+            "learning_rate": 0.0005,
+            "epoch": 0.5936167709623024,
+            "step": 13245
+        },
+        {
+            "loss": 14.0781,
+            "grad_norm": 1.9887112379074097,
+            "learning_rate": 0.0005,
+            "epoch": 0.5938408618535678,
+            "step": 13250
+        },
+        {
+            "loss": 13.9957,
+            "grad_norm": 1.8482990264892578,
+            "learning_rate": 0.0005,
+            "epoch": 0.5940649527448333,
+            "step": 13255
+        },
+        {
+            "loss": 14.0905,
+            "grad_norm": 1.8576593399047852,
+            "learning_rate": 0.0005,
+            "epoch": 0.5942890436360988,
+            "step": 13260
+        },
+        {
+            "loss": 14.0453,
+            "grad_norm": 1.986657738685608,
+            "learning_rate": 0.0005,
+            "epoch": 0.5945131345273643,
+            "step": 13265
+        },
+        {
+            "loss": 13.964,
+            "grad_norm": 1.9155220985412598,
+            "learning_rate": 0.0005,
+            "epoch": 0.5947372254186298,
+            "step": 13270
+        },
+        {
+            "loss": 14.0482,
+            "grad_norm": 1.9993773698806763,
+            "learning_rate": 0.0005,
+            "epoch": 0.5949613163098952,
+            "step": 13275
+        },
+        {
+            "loss": 13.9583,
+            "grad_norm": 1.8450108766555786,
+            "learning_rate": 0.0005,
+            "epoch": 0.5951854072011608,
+            "step": 13280
+        },
+        {
+            "loss": 13.9949,
+            "grad_norm": 1.7589229345321655,
+            "learning_rate": 0.0005,
+            "epoch": 0.5954094980924263,
+            "step": 13285
+        },
+        {
+            "loss": 13.9284,
+            "grad_norm": 1.9227980375289917,
+            "learning_rate": 0.0005,
+            "epoch": 0.5956335889836918,
+            "step": 13290
+        },
+        {
+            "loss": 14.0239,
+            "grad_norm": 1.8273468017578125,
+            "learning_rate": 0.0005,
+            "epoch": 0.5958576798749573,
+            "step": 13295
+        },
+        {
+            "loss": 13.975,
+            "grad_norm": 1.8974274396896362,
+            "learning_rate": 0.0005,
+            "epoch": 0.5960817707662228,
+            "step": 13300
+        },
+        {
+            "loss": 13.9155,
+            "grad_norm": 1.7669709920883179,
+            "learning_rate": 0.0005,
+            "epoch": 0.5963058616574882,
+            "step": 13305
+        },
+        {
+            "loss": 13.9947,
+            "grad_norm": 1.719914436340332,
+            "learning_rate": 0.0005,
+            "epoch": 0.5965299525487537,
+            "step": 13310
+        },
+        {
+            "loss": 13.9589,
+            "grad_norm": 1.8317246437072754,
+            "learning_rate": 0.0005,
+            "epoch": 0.5967540434400193,
+            "step": 13315
+        },
+        {
+            "loss": 14.0374,
+            "grad_norm": 1.9633820056915283,
+            "learning_rate": 0.0005,
+            "epoch": 0.5969781343312848,
+            "step": 13320
+        },
+        {
+            "loss": 14.0707,
+            "grad_norm": 1.788787841796875,
+            "learning_rate": 0.0005,
+            "epoch": 0.5972022252225503,
+            "step": 13325
+        },
+        {
+            "loss": 13.9841,
+            "grad_norm": 1.7529159784317017,
+            "learning_rate": 0.0005,
+            "epoch": 0.5974263161138158,
+            "step": 13330
+        },
+        {
+            "loss": 13.9827,
+            "grad_norm": 1.8291631937026978,
+            "learning_rate": 0.0005,
+            "epoch": 0.5976504070050812,
+            "step": 13335
+        },
+        {
+            "loss": 13.9554,
+            "grad_norm": 1.8712011575698853,
+            "learning_rate": 0.0005,
+            "epoch": 0.5978744978963467,
+            "step": 13340
+        },
+        {
+            "loss": 14.015,
+            "grad_norm": 2.0833263397216797,
+            "learning_rate": 0.0005,
+            "epoch": 0.5980985887876122,
+            "step": 13345
+        },
+        {
+            "loss": 14.1187,
+            "grad_norm": 1.7845981121063232,
+            "learning_rate": 0.0005,
+            "epoch": 0.5983226796788778,
+            "step": 13350
+        },
+        {
+            "loss": 13.9722,
+            "grad_norm": 1.9722967147827148,
+            "learning_rate": 0.0005,
+            "epoch": 0.5985467705701433,
+            "step": 13355
+        },
+        {
+            "loss": 14.0666,
+            "grad_norm": 1.8057105541229248,
+            "learning_rate": 0.0005,
+            "epoch": 0.5987708614614088,
+            "step": 13360
+        },
+        {
+            "loss": 13.9814,
+            "grad_norm": 1.8602194786071777,
+            "learning_rate": 0.0005,
+            "epoch": 0.5989949523526742,
+            "step": 13365
+        },
+        {
+            "loss": 13.988,
+            "grad_norm": 1.906585931777954,
+            "learning_rate": 0.0005,
+            "epoch": 0.5992190432439397,
+            "step": 13370
+        },
+        {
+            "loss": 14.0361,
+            "grad_norm": 1.843865156173706,
+            "learning_rate": 0.0005,
+            "epoch": 0.5994431341352052,
+            "step": 13375
+        },
+        {
+            "loss": 14.067,
+            "grad_norm": 1.943974256515503,
+            "learning_rate": 0.0005,
+            "epoch": 0.5996672250264707,
+            "step": 13380
+        },
+        {
+            "loss": 14.0264,
+            "grad_norm": 1.87297785282135,
+            "learning_rate": 0.0005,
+            "epoch": 0.5998913159177363,
+            "step": 13385
+        },
+        {
+            "loss": 13.963,
+            "grad_norm": 1.9725892543792725,
+            "learning_rate": 0.0005,
+            "epoch": 0.6001154068090018,
+            "step": 13390
+        },
+        {
+            "loss": 14.0057,
+            "grad_norm": 1.7576072216033936,
+            "learning_rate": 0.0005,
+            "epoch": 0.6003394977002672,
+            "step": 13395
+        },
+        {
+            "loss": 14.0163,
+            "grad_norm": 1.7891968488693237,
+            "learning_rate": 0.0005,
+            "epoch": 0.6005635885915327,
+            "step": 13400
+        },
+        {
+            "loss": 14.0749,
+            "grad_norm": 2.0627856254577637,
+            "learning_rate": 0.0005,
+            "epoch": 0.6007876794827982,
+            "step": 13405
+        },
+        {
+            "loss": 13.9621,
+            "grad_norm": 2.045072317123413,
+            "learning_rate": 0.0005,
+            "epoch": 0.6010117703740637,
+            "step": 13410
+        },
+        {
+            "loss": 14.0363,
+            "grad_norm": 1.9276081323623657,
+            "learning_rate": 0.0005,
+            "epoch": 0.6012358612653292,
+            "step": 13415
+        },
+        {
+            "loss": 14.0716,
+            "grad_norm": 1.9165892601013184,
+            "learning_rate": 0.0005,
+            "epoch": 0.6014599521565948,
+            "step": 13420
+        },
+        {
+            "loss": 14.0879,
+            "grad_norm": 1.9438608884811401,
+            "learning_rate": 0.0005,
+            "epoch": 0.6016840430478602,
+            "step": 13425
+        },
+        {
+            "loss": 13.9669,
+            "grad_norm": 1.997025966644287,
+            "learning_rate": 0.0005,
+            "epoch": 0.6019081339391257,
+            "step": 13430
+        },
+        {
+            "loss": 13.9591,
+            "grad_norm": 2.0243873596191406,
+            "learning_rate": 0.0005,
+            "epoch": 0.6021322248303912,
+            "step": 13435
+        },
+        {
+            "loss": 13.9436,
+            "grad_norm": 2.027860641479492,
+            "learning_rate": 0.0005,
+            "epoch": 0.6023563157216567,
+            "step": 13440
+        },
+        {
+            "loss": 13.7903,
+            "grad_norm": 1.785510540008545,
+            "learning_rate": 0.0005,
+            "epoch": 0.6025804066129222,
+            "step": 13445
+        },
+        {
+            "loss": 14.0328,
+            "grad_norm": 1.863256812095642,
+            "learning_rate": 0.0005,
+            "epoch": 0.6028044975041877,
+            "step": 13450
+        },
+        {
+            "loss": 13.9454,
+            "grad_norm": 1.8799083232879639,
+            "learning_rate": 0.0005,
+            "epoch": 0.6030285883954531,
+            "step": 13455
+        },
+        {
+            "loss": 13.9203,
+            "grad_norm": 1.7613584995269775,
+            "learning_rate": 0.0005,
+            "epoch": 0.6032526792867187,
+            "step": 13460
+        },
+        {
+            "loss": 13.993,
+            "grad_norm": 1.7297017574310303,
+            "learning_rate": 0.0005,
+            "epoch": 0.6034767701779842,
+            "step": 13465
+        },
+        {
+            "loss": 13.976,
+            "grad_norm": 1.9386945962905884,
+            "learning_rate": 0.0005,
+            "epoch": 0.6037008610692497,
+            "step": 13470
+        },
+        {
+            "loss": 13.8954,
+            "grad_norm": 1.883724570274353,
+            "learning_rate": 0.0005,
+            "epoch": 0.6039249519605152,
+            "step": 13475
+        },
+        {
+            "loss": 13.9777,
+            "grad_norm": 1.9041953086853027,
+            "learning_rate": 0.0005,
+            "epoch": 0.6041490428517807,
+            "step": 13480
+        },
+        {
+            "loss": 14.0164,
+            "grad_norm": 1.890601396560669,
+            "learning_rate": 0.0005,
+            "epoch": 0.6043731337430461,
+            "step": 13485
+        },
+        {
+            "loss": 13.9782,
+            "grad_norm": 1.8261315822601318,
+            "learning_rate": 0.0005,
+            "epoch": 0.6045972246343116,
+            "step": 13490
+        },
+        {
+            "loss": 14.0556,
+            "grad_norm": 1.8594465255737305,
+            "learning_rate": 0.0005,
+            "epoch": 0.6048213155255772,
+            "step": 13495
+        },
+        {
+            "loss": 13.9321,
+            "grad_norm": 1.9538732767105103,
+            "learning_rate": 0.0005,
+            "epoch": 0.6050454064168427,
+            "step": 13500
+        },
+        {
+            "eval_loss": 1.7396843433380127,
+            "eval_runtime": 18.3572,
+            "eval_samples_per_second": 892.513,
+            "eval_steps_per_second": 8.008,
+            "epoch": 0.6050454064168427,
+            "step": 13500
+        },
+        {
+            "loss": 14.0172,
+            "grad_norm": 1.9603897333145142,
+            "learning_rate": 0.0005,
+            "epoch": 0.6052694973081082,
+            "step": 13505
+        },
+        {
+            "loss": 13.9596,
+            "grad_norm": 1.8882741928100586,
+            "learning_rate": 0.0005,
+            "epoch": 0.6054935881993737,
+            "step": 13510
+        },
+        {
+            "loss": 13.9479,
+            "grad_norm": 1.9250017404556274,
+            "learning_rate": 0.0005,
+            "epoch": 0.6057176790906391,
+            "step": 13515
+        },
+        {
+            "loss": 13.9714,
+            "grad_norm": 1.8301010131835938,
+            "learning_rate": 0.0005,
+            "epoch": 0.6059417699819046,
+            "step": 13520
+        },
+        {
+            "loss": 13.9986,
+            "grad_norm": 1.8147294521331787,
+            "learning_rate": 0.0005,
+            "epoch": 0.6061658608731701,
+            "step": 13525
+        },
+        {
+            "loss": 13.9523,
+            "grad_norm": 1.8113244771957397,
+            "learning_rate": 0.0005,
+            "epoch": 0.6063899517644357,
+            "step": 13530
+        },
+        {
+            "loss": 13.9375,
+            "grad_norm": 1.8791850805282593,
+            "learning_rate": 0.0005,
+            "epoch": 0.6066140426557012,
+            "step": 13535
+        },
+        {
+            "loss": 13.9489,
+            "grad_norm": 1.732534646987915,
+            "learning_rate": 0.0005,
+            "epoch": 0.6068381335469667,
+            "step": 13540
+        },
+        {
+            "loss": 13.8999,
+            "grad_norm": 1.9004950523376465,
+            "learning_rate": 0.0005,
+            "epoch": 0.6070622244382321,
+            "step": 13545
+        },
+        {
+            "loss": 14.0194,
+            "grad_norm": 1.957031011581421,
+            "learning_rate": 0.0005,
+            "epoch": 0.6072863153294976,
+            "step": 13550
+        },
+        {
+            "loss": 14.0431,
+            "grad_norm": 1.8446530103683472,
+            "learning_rate": 0.0005,
+            "epoch": 0.6075104062207631,
+            "step": 13555
+        },
+        {
+            "loss": 13.9807,
+            "grad_norm": 2.006579875946045,
+            "learning_rate": 0.0005,
+            "epoch": 0.6077344971120286,
+            "step": 13560
+        },
+        {
+            "loss": 14.0643,
+            "grad_norm": 1.957889199256897,
+            "learning_rate": 0.0005,
+            "epoch": 0.6079585880032942,
+            "step": 13565
+        },
+        {
+            "loss": 14.0157,
+            "grad_norm": 1.978819489479065,
+            "learning_rate": 0.0005,
+            "epoch": 0.6081826788945597,
+            "step": 13570
+        },
+        {
+            "loss": 14.0688,
+            "grad_norm": 1.799669623374939,
+            "learning_rate": 0.0005,
+            "epoch": 0.6084067697858251,
+            "step": 13575
+        },
+        {
+            "loss": 14.0564,
+            "grad_norm": 1.74842369556427,
+            "learning_rate": 0.0005,
+            "epoch": 0.6086308606770906,
+            "step": 13580
+        },
+        {
+            "loss": 13.9357,
+            "grad_norm": 1.9134894609451294,
+            "learning_rate": 0.0005,
+            "epoch": 0.6088549515683561,
+            "step": 13585
+        },
+        {
+            "loss": 13.9969,
+            "grad_norm": 1.7899885177612305,
+            "learning_rate": 0.0005,
+            "epoch": 0.6090790424596216,
+            "step": 13590
+        },
+        {
+            "loss": 14.1657,
+            "grad_norm": 1.8893680572509766,
+            "learning_rate": 0.0005,
+            "epoch": 0.6093031333508871,
+            "step": 13595
+        },
+        {
+            "loss": 14.0536,
+            "grad_norm": 1.9131115674972534,
+            "learning_rate": 0.0005,
+            "epoch": 0.6095272242421527,
+            "step": 13600
+        },
+        {
+            "loss": 13.9266,
+            "grad_norm": 1.7662969827651978,
+            "learning_rate": 0.0005,
+            "epoch": 0.6097513151334181,
+            "step": 13605
+        },
+        {
+            "loss": 13.9748,
+            "grad_norm": 1.6508381366729736,
+            "learning_rate": 0.0005,
+            "epoch": 0.6099754060246836,
+            "step": 13610
+        },
+        {
+            "loss": 14.0392,
+            "grad_norm": 1.7770224809646606,
+            "learning_rate": 0.0005,
+            "epoch": 0.6101994969159491,
+            "step": 13615
+        },
+        {
+            "loss": 13.9233,
+            "grad_norm": 1.875481367111206,
+            "learning_rate": 0.0005,
+            "epoch": 0.6104235878072146,
+            "step": 13620
+        },
+        {
+            "loss": 13.9507,
+            "grad_norm": 1.7938653230667114,
+            "learning_rate": 0.0005,
+            "epoch": 0.6106476786984801,
+            "step": 13625
+        },
+        {
+            "loss": 13.9316,
+            "grad_norm": 1.8137539625167847,
+            "learning_rate": 0.0005,
+            "epoch": 0.6108717695897457,
+            "step": 13630
+        },
+        {
+            "loss": 13.9899,
+            "grad_norm": 1.780452847480774,
+            "learning_rate": 0.0005,
+            "epoch": 0.611095860481011,
+            "step": 13635
+        },
+        {
+            "loss": 14.0376,
+            "grad_norm": 1.930036187171936,
+            "learning_rate": 0.0005,
+            "epoch": 0.6113199513722766,
+            "step": 13640
+        },
+        {
+            "loss": 13.9412,
+            "grad_norm": 1.8009343147277832,
+            "learning_rate": 0.0005,
+            "epoch": 0.6115440422635421,
+            "step": 13645
+        },
+        {
+            "loss": 14.0176,
+            "grad_norm": 1.777569055557251,
+            "learning_rate": 0.0005,
+            "epoch": 0.6117681331548076,
+            "step": 13650
+        },
+        {
+            "loss": 13.9528,
+            "grad_norm": 1.8455289602279663,
+            "learning_rate": 0.0005,
+            "epoch": 0.6119922240460731,
+            "step": 13655
+        },
+        {
+            "loss": 13.9233,
+            "grad_norm": 1.8569191694259644,
+            "learning_rate": 0.0005,
+            "epoch": 0.6122163149373386,
+            "step": 13660
+        },
+        {
+            "loss": 14.0166,
+            "grad_norm": 1.8818286657333374,
+            "learning_rate": 0.0005,
+            "epoch": 0.612440405828604,
+            "step": 13665
+        },
+        {
+            "loss": 14.0259,
+            "grad_norm": 1.8745702505111694,
+            "learning_rate": 0.0005,
+            "epoch": 0.6126644967198696,
+            "step": 13670
+        },
+        {
+            "loss": 13.8863,
+            "grad_norm": 1.8713172674179077,
+            "learning_rate": 0.0005,
+            "epoch": 0.6128885876111351,
+            "step": 13675
+        },
+        {
+            "loss": 14.083,
+            "grad_norm": 1.9167075157165527,
+            "learning_rate": 0.0005,
+            "epoch": 0.6131126785024006,
+            "step": 13680
+        },
+        {
+            "loss": 13.9466,
+            "grad_norm": 1.9366717338562012,
+            "learning_rate": 0.0005,
+            "epoch": 0.6133367693936661,
+            "step": 13685
+        },
+        {
+            "loss": 14.0146,
+            "grad_norm": 2.0661909580230713,
+            "learning_rate": 0.0005,
+            "epoch": 0.6135608602849316,
+            "step": 13690
+        },
+        {
+            "loss": 14.0099,
+            "grad_norm": 2.0465962886810303,
+            "learning_rate": 0.0005,
+            "epoch": 0.613784951176197,
+            "step": 13695
+        },
+        {
+            "loss": 13.9691,
+            "grad_norm": 2.050764560699463,
+            "learning_rate": 0.0005,
+            "epoch": 0.6140090420674625,
+            "step": 13700
+        },
+        {
+            "loss": 13.962,
+            "grad_norm": 2.0589582920074463,
+            "learning_rate": 0.0005,
+            "epoch": 0.6142331329587281,
+            "step": 13705
+        },
+        {
+            "loss": 13.9126,
+            "grad_norm": 1.9636064767837524,
+            "learning_rate": 0.0005,
+            "epoch": 0.6144572238499936,
+            "step": 13710
+        },
+        {
+            "loss": 13.894,
+            "grad_norm": 1.9355812072753906,
+            "learning_rate": 0.0005,
+            "epoch": 0.6146813147412591,
+            "step": 13715
+        },
+        {
+            "loss": 13.935,
+            "grad_norm": 1.7583673000335693,
+            "learning_rate": 0.0005,
+            "epoch": 0.6149054056325246,
+            "step": 13720
+        },
+        {
+            "loss": 14.0261,
+            "grad_norm": 1.8463855981826782,
+            "learning_rate": 0.0005,
+            "epoch": 0.61512949652379,
+            "step": 13725
+        },
+        {
+            "loss": 14.0147,
+            "grad_norm": 1.794784665107727,
+            "learning_rate": 0.0005,
+            "epoch": 0.6153535874150555,
+            "step": 13730
+        },
+        {
+            "loss": 14.0213,
+            "grad_norm": 1.9001660346984863,
+            "learning_rate": 0.0005,
+            "epoch": 0.615577678306321,
+            "step": 13735
+        },
+        {
+            "loss": 13.9914,
+            "grad_norm": 1.800702691078186,
+            "learning_rate": 0.0005,
+            "epoch": 0.6158017691975866,
+            "step": 13740
+        },
+        {
+            "loss": 14.0857,
+            "grad_norm": 1.8461076021194458,
+            "learning_rate": 0.0005,
+            "epoch": 0.6160258600888521,
+            "step": 13745
+        },
+        {
+            "loss": 13.9006,
+            "grad_norm": 1.7148480415344238,
+            "learning_rate": 0.0005,
+            "epoch": 0.6162499509801176,
+            "step": 13750
+        },
+        {
+            "loss": 13.9633,
+            "grad_norm": 1.772823452949524,
+            "learning_rate": 0.0005,
+            "epoch": 0.616474041871383,
+            "step": 13755
+        },
+        {
+            "loss": 13.9953,
+            "grad_norm": 1.8532960414886475,
+            "learning_rate": 0.0005,
+            "epoch": 0.6166981327626485,
+            "step": 13760
+        },
+        {
+            "loss": 13.9837,
+            "grad_norm": 1.9373810291290283,
+            "learning_rate": 0.0005,
+            "epoch": 0.616922223653914,
+            "step": 13765
+        },
+        {
+            "loss": 14.0104,
+            "grad_norm": 1.789506435394287,
+            "learning_rate": 0.0005,
+            "epoch": 0.6171463145451795,
+            "step": 13770
+        },
+        {
+            "loss": 13.9058,
+            "grad_norm": 1.9801504611968994,
+            "learning_rate": 0.0005,
+            "epoch": 0.6173704054364451,
+            "step": 13775
+        },
+        {
+            "loss": 14.1375,
+            "grad_norm": 2.04109263420105,
+            "learning_rate": 0.0005,
+            "epoch": 0.6175944963277106,
+            "step": 13780
+        },
+        {
+            "loss": 13.953,
+            "grad_norm": 1.7348535060882568,
+            "learning_rate": 0.0005,
+            "epoch": 0.617818587218976,
+            "step": 13785
+        },
+        {
+            "loss": 14.066,
+            "grad_norm": 1.89817214012146,
+            "learning_rate": 0.0005,
+            "epoch": 0.6180426781102415,
+            "step": 13790
+        },
+        {
+            "loss": 13.9058,
+            "grad_norm": 2.1307947635650635,
+            "learning_rate": 0.0005,
+            "epoch": 0.618266769001507,
+            "step": 13795
+        },
+        {
+            "loss": 13.9577,
+            "grad_norm": 2.0067522525787354,
+            "learning_rate": 0.0005,
+            "epoch": 0.6184908598927725,
+            "step": 13800
+        },
+        {
+            "loss": 13.9795,
+            "grad_norm": 1.8235889673233032,
+            "learning_rate": 0.0005,
+            "epoch": 0.618714950784038,
+            "step": 13805
+        },
+        {
+            "loss": 13.9652,
+            "grad_norm": 1.71876060962677,
+            "learning_rate": 0.0005,
+            "epoch": 0.6189390416753036,
+            "step": 13810
+        },
+        {
+            "loss": 13.9877,
+            "grad_norm": 1.8106869459152222,
+            "learning_rate": 0.0005,
+            "epoch": 0.619163132566569,
+            "step": 13815
+        },
+        {
+            "loss": 13.9324,
+            "grad_norm": 1.8171489238739014,
+            "learning_rate": 0.0005,
+            "epoch": 0.6193872234578345,
+            "step": 13820
+        },
+        {
+            "loss": 14.0892,
+            "grad_norm": 1.7866055965423584,
+            "learning_rate": 0.0005,
+            "epoch": 0.6196113143491,
+            "step": 13825
+        },
+        {
+            "loss": 14.0149,
+            "grad_norm": 1.8658759593963623,
+            "learning_rate": 0.0005,
+            "epoch": 0.6198354052403655,
+            "step": 13830
+        },
+        {
+            "loss": 14.0664,
+            "grad_norm": 2.031806468963623,
+            "learning_rate": 0.0005,
+            "epoch": 0.620059496131631,
+            "step": 13835
+        },
+        {
+            "loss": 14.1421,
+            "grad_norm": 1.8138434886932373,
+            "learning_rate": 0.0005,
+            "epoch": 0.6202835870228964,
+            "step": 13840
+        },
+        {
+            "loss": 14.0283,
+            "grad_norm": 1.8643549680709839,
+            "learning_rate": 0.0005,
+            "epoch": 0.620507677914162,
+            "step": 13845
+        },
+        {
+            "loss": 13.8748,
+            "grad_norm": 2.06784987449646,
+            "learning_rate": 0.0005,
+            "epoch": 0.6207317688054275,
+            "step": 13850
+        },
+        {
+            "loss": 13.9531,
+            "grad_norm": 1.9382271766662598,
+            "learning_rate": 0.0005,
+            "epoch": 0.620955859696693,
+            "step": 13855
+        },
+        {
+            "loss": 14.0412,
+            "grad_norm": 1.8744487762451172,
+            "learning_rate": 0.0005,
+            "epoch": 0.6211799505879585,
+            "step": 13860
+        },
+        {
+            "loss": 13.8106,
+            "grad_norm": 1.8505882024765015,
+            "learning_rate": 0.0005,
+            "epoch": 0.621404041479224,
+            "step": 13865
+        },
+        {
+            "loss": 13.9749,
+            "grad_norm": 1.9018405675888062,
+            "learning_rate": 0.0005,
+            "epoch": 0.6216281323704894,
+            "step": 13870
+        },
+        {
+            "loss": 13.958,
+            "grad_norm": 1.9398044347763062,
+            "learning_rate": 0.0005,
+            "epoch": 0.6218522232617549,
+            "step": 13875
+        },
+        {
+            "loss": 13.9535,
+            "grad_norm": 1.8147361278533936,
+            "learning_rate": 0.0005,
+            "epoch": 0.6220763141530204,
+            "step": 13880
+        },
+        {
+            "loss": 13.911,
+            "grad_norm": 1.730082631111145,
+            "learning_rate": 0.0005,
+            "epoch": 0.622300405044286,
+            "step": 13885
+        },
+        {
+            "loss": 13.961,
+            "grad_norm": 2.0570127964019775,
+            "learning_rate": 0.0005,
+            "epoch": 0.6225244959355515,
+            "step": 13890
+        },
+        {
+            "loss": 14.05,
+            "grad_norm": 1.9781049489974976,
+            "learning_rate": 0.0005,
+            "epoch": 0.622748586826817,
+            "step": 13895
+        },
+        {
+            "loss": 14.0566,
+            "grad_norm": 1.8938955068588257,
+            "learning_rate": 0.0005,
+            "epoch": 0.6229726777180824,
+            "step": 13900
+        },
+        {
+            "loss": 14.0239,
+            "grad_norm": 1.8804343938827515,
+            "learning_rate": 0.0005,
+            "epoch": 0.6231967686093479,
+            "step": 13905
+        },
+        {
+            "loss": 14.009,
+            "grad_norm": 1.9425048828125,
+            "learning_rate": 0.0005,
+            "epoch": 0.6234208595006134,
+            "step": 13910
+        },
+        {
+            "loss": 14.0123,
+            "grad_norm": 1.753278136253357,
+            "learning_rate": 0.0005,
+            "epoch": 0.623644950391879,
+            "step": 13915
+        },
+        {
+            "loss": 13.9572,
+            "grad_norm": 1.8765408992767334,
+            "learning_rate": 0.0005,
+            "epoch": 0.6238690412831445,
+            "step": 13920
+        },
+        {
+            "loss": 14.0642,
+            "grad_norm": 1.867653250694275,
+            "learning_rate": 0.0005,
+            "epoch": 0.62409313217441,
+            "step": 13925
+        },
+        {
+            "loss": 13.9205,
+            "grad_norm": 1.8997493982315063,
+            "learning_rate": 0.0005,
+            "epoch": 0.6243172230656754,
+            "step": 13930
+        },
+        {
+            "loss": 13.9988,
+            "grad_norm": 1.8226754665374756,
+            "learning_rate": 0.0005,
+            "epoch": 0.6245413139569409,
+            "step": 13935
+        },
+        {
+            "loss": 13.9536,
+            "grad_norm": 1.774327039718628,
+            "learning_rate": 0.0005,
+            "epoch": 0.6247654048482064,
+            "step": 13940
+        },
+        {
+            "loss": 13.958,
+            "grad_norm": 2.0624618530273438,
+            "learning_rate": 0.0005,
+            "epoch": 0.6249894957394719,
+            "step": 13945
+        },
+        {
+            "loss": 14.0178,
+            "grad_norm": 2.0096333026885986,
+            "learning_rate": 0.0005,
+            "epoch": 0.6252135866307375,
+            "step": 13950
+        },
+        {
+            "loss": 13.9385,
+            "grad_norm": 1.836005449295044,
+            "learning_rate": 0.0005,
+            "epoch": 0.625437677522003,
+            "step": 13955
+        },
+        {
+            "loss": 13.9841,
+            "grad_norm": 1.7480518817901611,
+            "learning_rate": 0.0005,
+            "epoch": 0.6256617684132684,
+            "step": 13960
+        },
+        {
+            "loss": 14.0596,
+            "grad_norm": 1.9773201942443848,
+            "learning_rate": 0.0005,
+            "epoch": 0.6258858593045339,
+            "step": 13965
+        },
+        {
+            "loss": 13.9208,
+            "grad_norm": 1.8041434288024902,
+            "learning_rate": 0.0005,
+            "epoch": 0.6261099501957994,
+            "step": 13970
+        },
+        {
+            "loss": 14.0366,
+            "grad_norm": 1.7311369180679321,
+            "learning_rate": 0.0005,
+            "epoch": 0.6263340410870649,
+            "step": 13975
+        },
+        {
+            "loss": 13.996,
+            "grad_norm": 1.7201324701309204,
+            "learning_rate": 0.0005,
+            "epoch": 0.6265581319783304,
+            "step": 13980
+        },
+        {
+            "loss": 14.0056,
+            "grad_norm": 1.814854621887207,
+            "learning_rate": 0.0005,
+            "epoch": 0.626782222869596,
+            "step": 13985
+        },
+        {
+            "loss": 13.9742,
+            "grad_norm": 1.9238992929458618,
+            "learning_rate": 0.0005,
+            "epoch": 0.6270063137608614,
+            "step": 13990
+        },
+        {
+            "loss": 13.9946,
+            "grad_norm": 1.7817552089691162,
+            "learning_rate": 0.0005,
+            "epoch": 0.6272304046521269,
+            "step": 13995
+        },
+        {
+            "loss": 13.9743,
+            "grad_norm": 1.9954774379730225,
+            "learning_rate": 0.0005,
+            "epoch": 0.6274544955433924,
+            "step": 14000
+        },
+        {
+            "eval_loss": 1.7412445545196533,
+            "eval_runtime": 18.6224,
+            "eval_samples_per_second": 879.8,
+            "eval_steps_per_second": 7.894,
+            "epoch": 0.6274544955433924,
+            "step": 14000
+        },
+        {
+            "loss": 14.0041,
+            "grad_norm": 1.7871489524841309,
+            "learning_rate": 0.0005,
+            "epoch": 0.6276785864346579,
+            "step": 14005
+        },
+        {
+            "loss": 14.0718,
+            "grad_norm": 1.705079436302185,
+            "learning_rate": 0.0005,
+            "epoch": 0.6279026773259234,
+            "step": 14010
+        },
+        {
+            "loss": 13.9973,
+            "grad_norm": 1.8414729833602905,
+            "learning_rate": 0.0005,
+            "epoch": 0.6281267682171889,
+            "step": 14015
+        },
+        {
+            "loss": 14.0205,
+            "grad_norm": 1.7663222551345825,
+            "learning_rate": 0.0005,
+            "epoch": 0.6283508591084543,
+            "step": 14020
+        },
+        {
+            "loss": 14.0191,
+            "grad_norm": 1.9664376974105835,
+            "learning_rate": 0.0005,
+            "epoch": 0.6285749499997199,
+            "step": 14025
+        },
+        {
+            "loss": 13.9239,
+            "grad_norm": 1.706502079963684,
+            "learning_rate": 0.0005,
+            "epoch": 0.6287990408909854,
+            "step": 14030
+        },
+        {
+            "loss": 14.0895,
+            "grad_norm": 1.858054757118225,
+            "learning_rate": 0.0005,
+            "epoch": 0.6290231317822509,
+            "step": 14035
+        },
+        {
+            "loss": 14.0412,
+            "grad_norm": 1.9600058794021606,
+            "learning_rate": 0.0005,
+            "epoch": 0.6292472226735164,
+            "step": 14040
+        },
+        {
+            "loss": 13.9395,
+            "grad_norm": 2.096877098083496,
+            "learning_rate": 0.0005,
+            "epoch": 0.6294713135647819,
+            "step": 14045
+        },
+        {
+            "loss": 13.9903,
+            "grad_norm": 1.7399251461029053,
+            "learning_rate": 0.0005,
+            "epoch": 0.6296954044560473,
+            "step": 14050
+        },
+        {
+            "loss": 14.0175,
+            "grad_norm": 1.7890634536743164,
+            "learning_rate": 0.0005,
+            "epoch": 0.6299194953473128,
+            "step": 14055
+        },
+        {
+            "loss": 13.9969,
+            "grad_norm": 1.7637101411819458,
+            "learning_rate": 0.0005,
+            "epoch": 0.6301435862385784,
+            "step": 14060
+        },
+        {
+            "loss": 13.9807,
+            "grad_norm": 1.8002638816833496,
+            "learning_rate": 0.0005,
+            "epoch": 0.6303676771298439,
+            "step": 14065
+        },
+        {
+            "loss": 14.1129,
+            "grad_norm": 1.9056965112686157,
+            "learning_rate": 0.0005,
+            "epoch": 0.6305917680211094,
+            "step": 14070
+        },
+        {
+            "loss": 13.936,
+            "grad_norm": 1.9890260696411133,
+            "learning_rate": 0.0005,
+            "epoch": 0.6308158589123749,
+            "step": 14075
+        },
+        {
+            "loss": 14.0304,
+            "grad_norm": 1.753678321838379,
+            "learning_rate": 0.0005,
+            "epoch": 0.6310399498036403,
+            "step": 14080
+        },
+        {
+            "loss": 13.9602,
+            "grad_norm": 1.7128273248672485,
+            "learning_rate": 0.0005,
+            "epoch": 0.6312640406949058,
+            "step": 14085
+        },
+        {
+            "loss": 13.9127,
+            "grad_norm": 1.706416130065918,
+            "learning_rate": 0.0005,
+            "epoch": 0.6314881315861713,
+            "step": 14090
+        },
+        {
+            "loss": 13.8752,
+            "grad_norm": 1.8593871593475342,
+            "learning_rate": 0.0005,
+            "epoch": 0.6317122224774369,
+            "step": 14095
+        },
+        {
+            "loss": 14.0063,
+            "grad_norm": 2.070998430252075,
+            "learning_rate": 0.0005,
+            "epoch": 0.6319363133687024,
+            "step": 14100
+        },
+        {
+            "loss": 14.0104,
+            "grad_norm": 1.9629713296890259,
+            "learning_rate": 0.0005,
+            "epoch": 0.6321604042599679,
+            "step": 14105
+        },
+        {
+            "loss": 13.929,
+            "grad_norm": 1.7538399696350098,
+            "learning_rate": 0.0005,
+            "epoch": 0.6323844951512333,
+            "step": 14110
+        },
+        {
+            "loss": 14.0087,
+            "grad_norm": 1.8609179258346558,
+            "learning_rate": 0.0005,
+            "epoch": 0.6326085860424988,
+            "step": 14115
+        },
+        {
+            "loss": 13.9808,
+            "grad_norm": 1.7343677282333374,
+            "learning_rate": 0.0005,
+            "epoch": 0.6328326769337643,
+            "step": 14120
+        },
+        {
+            "loss": 13.999,
+            "grad_norm": 1.8937780857086182,
+            "learning_rate": 0.0005,
+            "epoch": 0.6330567678250298,
+            "step": 14125
+        },
+        {
+            "loss": 13.9897,
+            "grad_norm": 1.700460433959961,
+            "learning_rate": 0.0005,
+            "epoch": 0.6332808587162954,
+            "step": 14130
+        },
+        {
+            "loss": 13.915,
+            "grad_norm": 1.805824875831604,
+            "learning_rate": 0.0005,
+            "epoch": 0.6335049496075609,
+            "step": 14135
+        },
+        {
+            "loss": 14.0701,
+            "grad_norm": 1.755516767501831,
+            "learning_rate": 0.0005,
+            "epoch": 0.6337290404988263,
+            "step": 14140
+        },
+        {
+            "loss": 13.9003,
+            "grad_norm": 1.7965532541275024,
+            "learning_rate": 0.0005,
+            "epoch": 0.6339531313900918,
+            "step": 14145
+        },
+        {
+            "loss": 13.886,
+            "grad_norm": 1.8935853242874146,
+            "learning_rate": 0.0005,
+            "epoch": 0.6341772222813573,
+            "step": 14150
+        },
+        {
+            "loss": 13.956,
+            "grad_norm": 1.919097900390625,
+            "learning_rate": 0.0005,
+            "epoch": 0.6344013131726228,
+            "step": 14155
+        },
+        {
+            "loss": 14.0817,
+            "grad_norm": 2.0242080688476562,
+            "learning_rate": 0.0005,
+            "epoch": 0.6346254040638883,
+            "step": 14160
+        },
+        {
+            "loss": 14.0245,
+            "grad_norm": 1.946304440498352,
+            "learning_rate": 0.0005,
+            "epoch": 0.6348494949551539,
+            "step": 14165
+        },
+        {
+            "loss": 13.8935,
+            "grad_norm": 1.9550292491912842,
+            "learning_rate": 0.0005,
+            "epoch": 0.6350735858464193,
+            "step": 14170
+        },
+        {
+            "loss": 14.0761,
+            "grad_norm": 2.2620325088500977,
+            "learning_rate": 0.0005,
+            "epoch": 0.6352976767376848,
+            "step": 14175
+        },
+        {
+            "loss": 14.0249,
+            "grad_norm": 2.0270233154296875,
+            "learning_rate": 0.0005,
+            "epoch": 0.6355217676289503,
+            "step": 14180
+        },
+        {
+            "loss": 14.0,
+            "grad_norm": 2.0550220012664795,
+            "learning_rate": 0.0005,
+            "epoch": 0.6357458585202158,
+            "step": 14185
+        },
+        {
+            "loss": 14.0812,
+            "grad_norm": 2.0032031536102295,
+            "learning_rate": 0.0005,
+            "epoch": 0.6359699494114813,
+            "step": 14190
+        },
+        {
+            "loss": 13.9631,
+            "grad_norm": 1.7279571294784546,
+            "learning_rate": 0.0005,
+            "epoch": 0.6361940403027468,
+            "step": 14195
+        },
+        {
+            "loss": 13.9304,
+            "grad_norm": 1.76564359664917,
+            "learning_rate": 0.0005,
+            "epoch": 0.6364181311940122,
+            "step": 14200
+        },
+        {
+            "loss": 13.9507,
+            "grad_norm": 1.8951811790466309,
+            "learning_rate": 0.0005,
+            "epoch": 0.6366422220852778,
+            "step": 14205
+        },
+        {
+            "loss": 13.9794,
+            "grad_norm": 1.7524734735488892,
+            "learning_rate": 0.0005,
+            "epoch": 0.6368663129765433,
+            "step": 14210
+        },
+        {
+            "loss": 14.092,
+            "grad_norm": 1.8108766078948975,
+            "learning_rate": 0.0005,
+            "epoch": 0.6370904038678088,
+            "step": 14215
+        },
+        {
+            "loss": 14.0535,
+            "grad_norm": 1.8466496467590332,
+            "learning_rate": 0.0005,
+            "epoch": 0.6373144947590743,
+            "step": 14220
+        },
+        {
+            "loss": 14.0813,
+            "grad_norm": 1.866255760192871,
+            "learning_rate": 0.0005,
+            "epoch": 0.6375385856503398,
+            "step": 14225
+        },
+        {
+            "loss": 13.8828,
+            "grad_norm": 1.8671491146087646,
+            "learning_rate": 0.0005,
+            "epoch": 0.6377626765416052,
+            "step": 14230
+        },
+        {
+            "loss": 13.8839,
+            "grad_norm": 1.6825112104415894,
+            "learning_rate": 0.0005,
+            "epoch": 0.6379867674328707,
+            "step": 14235
+        },
+        {
+            "loss": 14.0688,
+            "grad_norm": 1.8906571865081787,
+            "learning_rate": 0.0005,
+            "epoch": 0.6382108583241363,
+            "step": 14240
+        },
+        {
+            "loss": 14.0267,
+            "grad_norm": 1.99757719039917,
+            "learning_rate": 0.0005,
+            "epoch": 0.6384349492154018,
+            "step": 14245
+        },
+        {
+            "loss": 13.9307,
+            "grad_norm": 1.889869213104248,
+            "learning_rate": 0.0005,
+            "epoch": 0.6386590401066673,
+            "step": 14250
+        },
+        {
+            "loss": 13.9939,
+            "grad_norm": 1.7477375268936157,
+            "learning_rate": 0.0005,
+            "epoch": 0.6388831309979328,
+            "step": 14255
+        },
+        {
+            "loss": 13.9191,
+            "grad_norm": 1.7579962015151978,
+            "learning_rate": 0.0005,
+            "epoch": 0.6391072218891982,
+            "step": 14260
+        },
+        {
+            "loss": 14.0355,
+            "grad_norm": 1.838754653930664,
+            "learning_rate": 0.0005,
+            "epoch": 0.6393313127804637,
+            "step": 14265
+        },
+        {
+            "loss": 13.9155,
+            "grad_norm": 1.835952877998352,
+            "learning_rate": 0.0005,
+            "epoch": 0.6395554036717293,
+            "step": 14270
+        },
+        {
+            "loss": 13.92,
+            "grad_norm": 1.85426926612854,
+            "learning_rate": 0.0005,
+            "epoch": 0.6397794945629948,
+            "step": 14275
+        },
+        {
+            "loss": 14.0431,
+            "grad_norm": 1.7805200815200806,
+            "learning_rate": 0.0005,
+            "epoch": 0.6400035854542603,
+            "step": 14280
+        },
+        {
+            "loss": 13.9615,
+            "grad_norm": 1.7967098951339722,
+            "learning_rate": 0.0005,
+            "epoch": 0.6402276763455258,
+            "step": 14285
+        },
+        {
+            "loss": 13.9886,
+            "grad_norm": 1.6179386377334595,
+            "learning_rate": 0.0005,
+            "epoch": 0.6404517672367912,
+            "step": 14290
+        },
+        {
+            "loss": 14.0352,
+            "grad_norm": 1.9071606397628784,
+            "learning_rate": 0.0005,
+            "epoch": 0.6406758581280567,
+            "step": 14295
+        },
+        {
+            "loss": 13.9369,
+            "grad_norm": 1.928659439086914,
+            "learning_rate": 0.0005,
+            "epoch": 0.6408999490193222,
+            "step": 14300
+        },
+        {
+            "loss": 14.0594,
+            "grad_norm": 1.7425745725631714,
+            "learning_rate": 0.0005,
+            "epoch": 0.6411240399105878,
+            "step": 14305
+        },
+        {
+            "loss": 13.9505,
+            "grad_norm": 1.9480642080307007,
+            "learning_rate": 0.0005,
+            "epoch": 0.6413481308018533,
+            "step": 14310
+        },
+        {
+            "loss": 13.9178,
+            "grad_norm": 1.7811546325683594,
+            "learning_rate": 0.0005,
+            "epoch": 0.6415722216931188,
+            "step": 14315
+        },
+        {
+            "loss": 13.9264,
+            "grad_norm": 1.7841154336929321,
+            "learning_rate": 0.0005,
+            "epoch": 0.6417963125843842,
+            "step": 14320
+        },
+        {
+            "loss": 13.963,
+            "grad_norm": 1.872770071029663,
+            "learning_rate": 0.0005,
+            "epoch": 0.6420204034756497,
+            "step": 14325
+        },
+        {
+            "loss": 14.0144,
+            "grad_norm": 2.028611660003662,
+            "learning_rate": 0.0005,
+            "epoch": 0.6422444943669152,
+            "step": 14330
+        },
+        {
+            "loss": 13.9069,
+            "grad_norm": 1.810651421546936,
+            "learning_rate": 0.0005,
+            "epoch": 0.6424685852581807,
+            "step": 14335
+        },
+        {
+            "loss": 13.9686,
+            "grad_norm": 1.918397307395935,
+            "learning_rate": 0.0005,
+            "epoch": 0.6426926761494463,
+            "step": 14340
+        },
+        {
+            "loss": 14.0204,
+            "grad_norm": 1.9416550397872925,
+            "learning_rate": 0.0005,
+            "epoch": 0.6429167670407118,
+            "step": 14345
+        },
+        {
+            "loss": 13.9854,
+            "grad_norm": 1.8357208967208862,
+            "learning_rate": 0.0005,
+            "epoch": 0.6431408579319772,
+            "step": 14350
+        },
+        {
+            "loss": 13.9148,
+            "grad_norm": 1.8995227813720703,
+            "learning_rate": 0.0005,
+            "epoch": 0.6433649488232427,
+            "step": 14355
+        },
+        {
+            "loss": 14.0029,
+            "grad_norm": 1.7630736827850342,
+            "learning_rate": 0.0005,
+            "epoch": 0.6435890397145082,
+            "step": 14360
+        },
+        {
+            "loss": 14.0908,
+            "grad_norm": 1.9035780429840088,
+            "learning_rate": 0.0005,
+            "epoch": 0.6438131306057737,
+            "step": 14365
+        },
+        {
+            "loss": 14.0767,
+            "grad_norm": 2.125985860824585,
+            "learning_rate": 0.0005,
+            "epoch": 0.6440372214970392,
+            "step": 14370
+        },
+        {
+            "loss": 14.0037,
+            "grad_norm": 2.025540351867676,
+            "learning_rate": 0.0005,
+            "epoch": 0.6442613123883046,
+            "step": 14375
+        },
+        {
+            "loss": 14.1204,
+            "grad_norm": 1.9417158365249634,
+            "learning_rate": 0.0005,
+            "epoch": 0.6444854032795702,
+            "step": 14380
+        },
+        {
+            "loss": 13.9539,
+            "grad_norm": 1.753354549407959,
+            "learning_rate": 0.0005,
+            "epoch": 0.6447094941708357,
+            "step": 14385
+        },
+        {
+            "loss": 13.9972,
+            "grad_norm": 1.866416573524475,
+            "learning_rate": 0.0005,
+            "epoch": 0.6449335850621012,
+            "step": 14390
+        },
+        {
+            "loss": 13.9683,
+            "grad_norm": 1.8368868827819824,
+            "learning_rate": 0.0005,
+            "epoch": 0.6451576759533667,
+            "step": 14395
+        },
+        {
+            "loss": 13.8554,
+            "grad_norm": 1.7871224880218506,
+            "learning_rate": 0.0005,
+            "epoch": 0.6453817668446322,
+            "step": 14400
+        },
+        {
+            "loss": 13.8737,
+            "grad_norm": 1.802868127822876,
+            "learning_rate": 0.0005,
+            "epoch": 0.6456058577358976,
+            "step": 14405
+        },
+        {
+            "loss": 13.911,
+            "grad_norm": 1.8559720516204834,
+            "learning_rate": 0.0005,
+            "epoch": 0.6458299486271631,
+            "step": 14410
+        },
+        {
+            "loss": 13.9026,
+            "grad_norm": 2.0018739700317383,
+            "learning_rate": 0.0005,
+            "epoch": 0.6460540395184287,
+            "step": 14415
+        },
+        {
+            "loss": 14.0628,
+            "grad_norm": 1.8371291160583496,
+            "learning_rate": 0.0005,
+            "epoch": 0.6462781304096942,
+            "step": 14420
+        },
+        {
+            "loss": 14.0768,
+            "grad_norm": 2.005053758621216,
+            "learning_rate": 0.0005,
+            "epoch": 0.6465022213009597,
+            "step": 14425
+        },
+        {
+            "loss": 13.9509,
+            "grad_norm": 1.8181270360946655,
+            "learning_rate": 0.0005,
+            "epoch": 0.6467263121922252,
+            "step": 14430
+        },
+        {
+            "loss": 13.9489,
+            "grad_norm": 1.8325514793395996,
+            "learning_rate": 0.0005,
+            "epoch": 0.6469504030834906,
+            "step": 14435
+        },
+        {
+            "loss": 13.9991,
+            "grad_norm": 1.8346320390701294,
+            "learning_rate": 0.0005,
+            "epoch": 0.6471744939747561,
+            "step": 14440
+        },
+        {
+            "loss": 14.0045,
+            "grad_norm": 1.7492088079452515,
+            "learning_rate": 0.0005,
+            "epoch": 0.6473985848660216,
+            "step": 14445
+        },
+        {
+            "loss": 14.0071,
+            "grad_norm": 1.8489519357681274,
+            "learning_rate": 0.0005,
+            "epoch": 0.6476226757572872,
+            "step": 14450
+        },
+        {
+            "loss": 13.874,
+            "grad_norm": 1.8009483814239502,
+            "learning_rate": 0.0005,
+            "epoch": 0.6478467666485527,
+            "step": 14455
+        },
+        {
+            "loss": 13.8814,
+            "grad_norm": 1.7434006929397583,
+            "learning_rate": 0.0005,
+            "epoch": 0.6480708575398182,
+            "step": 14460
+        },
+        {
+            "loss": 13.9751,
+            "grad_norm": 1.910570502281189,
+            "learning_rate": 0.0005,
+            "epoch": 0.6482949484310836,
+            "step": 14465
+        },
+        {
+            "loss": 14.0008,
+            "grad_norm": 1.9025352001190186,
+            "learning_rate": 0.0005,
+            "epoch": 0.6485190393223491,
+            "step": 14470
+        },
+        {
+            "loss": 13.9886,
+            "grad_norm": 1.918157696723938,
+            "learning_rate": 0.0005,
+            "epoch": 0.6487431302136146,
+            "step": 14475
+        },
+        {
+            "loss": 13.9403,
+            "grad_norm": 1.9863786697387695,
+            "learning_rate": 0.0005,
+            "epoch": 0.6489672211048801,
+            "step": 14480
+        },
+        {
+            "loss": 13.9518,
+            "grad_norm": 1.8702291250228882,
+            "learning_rate": 0.0005,
+            "epoch": 0.6491913119961457,
+            "step": 14485
+        },
+        {
+            "loss": 14.0877,
+            "grad_norm": 2.1532609462738037,
+            "learning_rate": 0.0005,
+            "epoch": 0.6494154028874112,
+            "step": 14490
+        },
+        {
+            "loss": 13.9791,
+            "grad_norm": 1.9106634855270386,
+            "learning_rate": 0.0005,
+            "epoch": 0.6496394937786766,
+            "step": 14495
+        },
+        {
+            "loss": 13.9435,
+            "grad_norm": 1.7779203653335571,
+            "learning_rate": 0.0005,
+            "epoch": 0.6498635846699421,
+            "step": 14500
+        },
+        {
+            "eval_loss": 1.741438627243042,
+            "eval_runtime": 18.8225,
+            "eval_samples_per_second": 870.449,
+            "eval_steps_per_second": 7.81,
+            "epoch": 0.6498635846699421,
+            "step": 14500
+        },
+        {
+            "loss": 13.8422,
+            "grad_norm": 1.825476884841919,
+            "learning_rate": 0.0005,
+            "epoch": 0.6500876755612076,
+            "step": 14505
+        },
+        {
+            "loss": 13.8998,
+            "grad_norm": 2.1719396114349365,
+            "learning_rate": 0.0005,
+            "epoch": 0.6503117664524731,
+            "step": 14510
+        },
+        {
+            "loss": 13.9174,
+            "grad_norm": 1.8220202922821045,
+            "learning_rate": 0.0005,
+            "epoch": 0.6505358573437386,
+            "step": 14515
+        },
+        {
+            "loss": 13.8703,
+            "grad_norm": 1.6797312498092651,
+            "learning_rate": 0.0005,
+            "epoch": 0.6507599482350042,
+            "step": 14520
+        },
+        {
+            "loss": 13.9191,
+            "grad_norm": 1.8802516460418701,
+            "learning_rate": 0.0005,
+            "epoch": 0.6509840391262696,
+            "step": 14525
+        },
+        {
+            "loss": 13.9308,
+            "grad_norm": 1.7543238401412964,
+            "learning_rate": 0.0005,
+            "epoch": 0.6512081300175351,
+            "step": 14530
+        },
+        {
+            "loss": 13.9549,
+            "grad_norm": 1.8070513010025024,
+            "learning_rate": 0.0005,
+            "epoch": 0.6514322209088006,
+            "step": 14535
+        },
+        {
+            "loss": 13.8917,
+            "grad_norm": 1.7234477996826172,
+            "learning_rate": 0.0005,
+            "epoch": 0.6516563118000661,
+            "step": 14540
+        },
+        {
+            "loss": 14.0706,
+            "grad_norm": 1.919089674949646,
+            "learning_rate": 0.0005,
+            "epoch": 0.6518804026913316,
+            "step": 14545
+        },
+        {
+            "loss": 13.9334,
+            "grad_norm": 1.6644461154937744,
+            "learning_rate": 0.0005,
+            "epoch": 0.6521044935825971,
+            "step": 14550
+        },
+        {
+            "loss": 13.9985,
+            "grad_norm": 1.8210339546203613,
+            "learning_rate": 0.0005,
+            "epoch": 0.6523285844738625,
+            "step": 14555
+        },
+        {
+            "loss": 14.1057,
+            "grad_norm": 1.849721074104309,
+            "learning_rate": 0.0005,
+            "epoch": 0.6525526753651281,
+            "step": 14560
+        },
+        {
+            "loss": 13.9989,
+            "grad_norm": 1.818102478981018,
+            "learning_rate": 0.0005,
+            "epoch": 0.6527767662563936,
+            "step": 14565
+        },
+        {
+            "loss": 13.8813,
+            "grad_norm": 1.669271469116211,
+            "learning_rate": 0.0005,
+            "epoch": 0.6530008571476591,
+            "step": 14570
+        },
+        {
+            "loss": 14.047,
+            "grad_norm": 1.731544852256775,
+            "learning_rate": 0.0005,
+            "epoch": 0.6532249480389246,
+            "step": 14575
+        },
+        {
+            "loss": 13.9806,
+            "grad_norm": 1.7254154682159424,
+            "learning_rate": 0.0005,
+            "epoch": 0.6534490389301901,
+            "step": 14580
+        },
+        {
+            "loss": 13.9701,
+            "grad_norm": 1.8537356853485107,
+            "learning_rate": 0.0005,
+            "epoch": 0.6536731298214555,
+            "step": 14585
+        },
+        {
+            "loss": 13.9786,
+            "grad_norm": 1.8968428373336792,
+            "learning_rate": 0.0005,
+            "epoch": 0.653897220712721,
+            "step": 14590
+        },
+        {
+            "loss": 13.9947,
+            "grad_norm": 1.9824936389923096,
+            "learning_rate": 0.0005,
+            "epoch": 0.6541213116039866,
+            "step": 14595
+        },
+        {
+            "loss": 13.9541,
+            "grad_norm": 2.01820969581604,
+            "learning_rate": 0.0005,
+            "epoch": 0.6543454024952521,
+            "step": 14600
+        },
+        {
+            "loss": 13.8774,
+            "grad_norm": 1.703083872795105,
+            "learning_rate": 0.0005,
+            "epoch": 0.6545694933865176,
+            "step": 14605
+        },
+        {
+            "loss": 13.8502,
+            "grad_norm": 1.8082938194274902,
+            "learning_rate": 0.0005,
+            "epoch": 0.6547935842777831,
+            "step": 14610
+        },
+        {
+            "loss": 13.9328,
+            "grad_norm": 1.763091802597046,
+            "learning_rate": 0.0005,
+            "epoch": 0.6550176751690485,
+            "step": 14615
+        },
+        {
+            "loss": 14.0332,
+            "grad_norm": 2.1249189376831055,
+            "learning_rate": 0.0005,
+            "epoch": 0.655241766060314,
+            "step": 14620
+        },
+        {
+            "loss": 13.955,
+            "grad_norm": 1.8449853658676147,
+            "learning_rate": 0.0005,
+            "epoch": 0.6554658569515796,
+            "step": 14625
+        },
+        {
+            "loss": 14.0941,
+            "grad_norm": 1.8855788707733154,
+            "learning_rate": 0.0005,
+            "epoch": 0.6556899478428451,
+            "step": 14630
+        },
+        {
+            "loss": 13.9477,
+            "grad_norm": 1.754029631614685,
+            "learning_rate": 0.0005,
+            "epoch": 0.6559140387341106,
+            "step": 14635
+        },
+        {
+            "loss": 14.0136,
+            "grad_norm": 1.8334318399429321,
+            "learning_rate": 0.0005,
+            "epoch": 0.6561381296253761,
+            "step": 14640
+        },
+        {
+            "loss": 14.1017,
+            "grad_norm": 1.7702960968017578,
+            "learning_rate": 0.0005,
+            "epoch": 0.6563622205166415,
+            "step": 14645
+        },
+        {
+            "loss": 13.9838,
+            "grad_norm": 2.070499897003174,
+            "learning_rate": 0.0005,
+            "epoch": 0.656586311407907,
+            "step": 14650
+        },
+        {
+            "loss": 14.0847,
+            "grad_norm": 2.111660957336426,
+            "learning_rate": 0.0005,
+            "epoch": 0.6568104022991725,
+            "step": 14655
+        },
+        {
+            "loss": 13.9565,
+            "grad_norm": 1.7753303050994873,
+            "learning_rate": 0.0005,
+            "epoch": 0.657034493190438,
+            "step": 14660
+        },
+        {
+            "loss": 14.1332,
+            "grad_norm": 1.8608185052871704,
+            "learning_rate": 0.0005,
+            "epoch": 0.6572585840817036,
+            "step": 14665
+        },
+        {
+            "loss": 13.9903,
+            "grad_norm": 1.9424304962158203,
+            "learning_rate": 0.0005,
+            "epoch": 0.6574826749729691,
+            "step": 14670
+        },
+        {
+            "loss": 14.0011,
+            "grad_norm": 1.7471083402633667,
+            "learning_rate": 0.0005,
+            "epoch": 0.6577067658642345,
+            "step": 14675
+        },
+        {
+            "loss": 13.9287,
+            "grad_norm": 1.8507441282272339,
+            "learning_rate": 0.0005,
+            "epoch": 0.6579308567555,
+            "step": 14680
+        },
+        {
+            "loss": 13.9519,
+            "grad_norm": 1.78011953830719,
+            "learning_rate": 0.0005,
+            "epoch": 0.6581549476467655,
+            "step": 14685
+        },
+        {
+            "loss": 13.9735,
+            "grad_norm": 1.7394940853118896,
+            "learning_rate": 0.0005,
+            "epoch": 0.658379038538031,
+            "step": 14690
+        },
+        {
+            "loss": 13.9678,
+            "grad_norm": 1.8042924404144287,
+            "learning_rate": 0.0005,
+            "epoch": 0.6586031294292966,
+            "step": 14695
+        },
+        {
+            "loss": 14.0196,
+            "grad_norm": 1.7354576587677002,
+            "learning_rate": 0.0005,
+            "epoch": 0.6588272203205621,
+            "step": 14700
+        },
+        {
+            "loss": 13.9881,
+            "grad_norm": 1.9211896657943726,
+            "learning_rate": 0.0005,
+            "epoch": 0.6590513112118275,
+            "step": 14705
+        },
+        {
+            "loss": 14.0377,
+            "grad_norm": 2.0172202587127686,
+            "learning_rate": 0.0005,
+            "epoch": 0.659275402103093,
+            "step": 14710
+        },
+        {
+            "loss": 13.8951,
+            "grad_norm": 1.9813729524612427,
+            "learning_rate": 0.0005,
+            "epoch": 0.6594994929943585,
+            "step": 14715
+        },
+        {
+            "loss": 13.9769,
+            "grad_norm": 1.8304388523101807,
+            "learning_rate": 0.0005,
+            "epoch": 0.659723583885624,
+            "step": 14720
+        },
+        {
+            "loss": 14.1043,
+            "grad_norm": 1.95379638671875,
+            "learning_rate": 0.0005,
+            "epoch": 0.6599476747768895,
+            "step": 14725
+        },
+        {
+            "loss": 13.8718,
+            "grad_norm": 1.9209054708480835,
+            "learning_rate": 0.0005,
+            "epoch": 0.660171765668155,
+            "step": 14730
+        },
+        {
+            "loss": 13.9663,
+            "grad_norm": 2.017179250717163,
+            "learning_rate": 0.0005,
+            "epoch": 0.6603958565594205,
+            "step": 14735
+        },
+        {
+            "loss": 13.9567,
+            "grad_norm": 1.9263821840286255,
+            "learning_rate": 0.0005,
+            "epoch": 0.660619947450686,
+            "step": 14740
+        },
+        {
+            "loss": 14.017,
+            "grad_norm": 1.9511022567749023,
+            "learning_rate": 0.0005,
+            "epoch": 0.6608440383419515,
+            "step": 14745
+        },
+        {
+            "loss": 13.9642,
+            "grad_norm": 2.0967555046081543,
+            "learning_rate": 0.0005,
+            "epoch": 0.661068129233217,
+            "step": 14750
+        },
+        {
+            "loss": 13.9685,
+            "grad_norm": 1.841191053390503,
+            "learning_rate": 0.0005,
+            "epoch": 0.6612922201244825,
+            "step": 14755
+        },
+        {
+            "loss": 13.9074,
+            "grad_norm": 1.949305534362793,
+            "learning_rate": 0.0005,
+            "epoch": 0.661516311015748,
+            "step": 14760
+        },
+        {
+            "loss": 13.8287,
+            "grad_norm": 2.0074777603149414,
+            "learning_rate": 0.0005,
+            "epoch": 0.6617404019070134,
+            "step": 14765
+        },
+        {
+            "loss": 14.0528,
+            "grad_norm": 2.0784366130828857,
+            "learning_rate": 0.0005,
+            "epoch": 0.661964492798279,
+            "step": 14770
+        },
+        {
+            "loss": 13.9003,
+            "grad_norm": 1.922762155532837,
+            "learning_rate": 0.0005,
+            "epoch": 0.6621885836895445,
+            "step": 14775
+        },
+        {
+            "loss": 13.9591,
+            "grad_norm": 2.1148183345794678,
+            "learning_rate": 0.0005,
+            "epoch": 0.66241267458081,
+            "step": 14780
+        },
+        {
+            "loss": 13.9369,
+            "grad_norm": 1.88979172706604,
+            "learning_rate": 0.0005,
+            "epoch": 0.6626367654720755,
+            "step": 14785
+        },
+        {
+            "loss": 13.9334,
+            "grad_norm": 2.0521533489227295,
+            "learning_rate": 0.0005,
+            "epoch": 0.662860856363341,
+            "step": 14790
+        },
+        {
+            "loss": 13.8078,
+            "grad_norm": 1.8560912609100342,
+            "learning_rate": 0.0005,
+            "epoch": 0.6630849472546064,
+            "step": 14795
+        },
+        {
+            "loss": 13.992,
+            "grad_norm": 1.895782232284546,
+            "learning_rate": 0.0005,
+            "epoch": 0.6633090381458719,
+            "step": 14800
+        },
+        {
+            "loss": 13.915,
+            "grad_norm": 1.9879924058914185,
+            "learning_rate": 0.0005,
+            "epoch": 0.6635331290371375,
+            "step": 14805
+        },
+        {
+            "loss": 14.0203,
+            "grad_norm": 2.0573923587799072,
+            "learning_rate": 0.0005,
+            "epoch": 0.663757219928403,
+            "step": 14810
+        },
+        {
+            "loss": 13.9983,
+            "grad_norm": 1.8449429273605347,
+            "learning_rate": 0.0005,
+            "epoch": 0.6639813108196685,
+            "step": 14815
+        },
+        {
+            "loss": 13.9999,
+            "grad_norm": 1.8708304166793823,
+            "learning_rate": 0.0005,
+            "epoch": 0.664205401710934,
+            "step": 14820
+        },
+        {
+            "loss": 13.9573,
+            "grad_norm": 1.8508751392364502,
+            "learning_rate": 0.0005,
+            "epoch": 0.6644294926021994,
+            "step": 14825
+        },
+        {
+            "loss": 14.1518,
+            "grad_norm": 1.816766381263733,
+            "learning_rate": 0.0005,
+            "epoch": 0.6646535834934649,
+            "step": 14830
+        },
+        {
+            "loss": 13.9542,
+            "grad_norm": 1.728379487991333,
+            "learning_rate": 0.0005,
+            "epoch": 0.6648776743847304,
+            "step": 14835
+        },
+        {
+            "loss": 13.9146,
+            "grad_norm": 1.8660993576049805,
+            "learning_rate": 0.0005,
+            "epoch": 0.665101765275996,
+            "step": 14840
+        },
+        {
+            "loss": 13.866,
+            "grad_norm": 1.7663236856460571,
+            "learning_rate": 0.0005,
+            "epoch": 0.6653258561672615,
+            "step": 14845
+        },
+        {
+            "loss": 14.0455,
+            "grad_norm": 1.6998153924942017,
+            "learning_rate": 0.0005,
+            "epoch": 0.665549947058527,
+            "step": 14850
+        },
+        {
+            "loss": 14.0245,
+            "grad_norm": 1.9014002084732056,
+            "learning_rate": 0.0005,
+            "epoch": 0.6657740379497924,
+            "step": 14855
+        },
+        {
+            "loss": 14.0883,
+            "grad_norm": 2.1450681686401367,
+            "learning_rate": 0.0005,
+            "epoch": 0.6659981288410579,
+            "step": 14860
+        },
+        {
+            "loss": 14.0052,
+            "grad_norm": 2.028628349304199,
+            "learning_rate": 0.0005,
+            "epoch": 0.6662222197323234,
+            "step": 14865
+        },
+        {
+            "loss": 13.9657,
+            "grad_norm": 1.8996095657348633,
+            "learning_rate": 0.0005,
+            "epoch": 0.6664463106235889,
+            "step": 14870
+        },
+        {
+            "loss": 13.99,
+            "grad_norm": 1.8172988891601562,
+            "learning_rate": 0.0005,
+            "epoch": 0.6666704015148545,
+            "step": 14875
+        },
+        {
+            "loss": 13.9786,
+            "grad_norm": 1.9849042892456055,
+            "learning_rate": 0.0005,
+            "epoch": 0.66689449240612,
+            "step": 14880
+        },
+        {
+            "loss": 13.9666,
+            "grad_norm": 1.7792553901672363,
+            "learning_rate": 0.0005,
+            "epoch": 0.6671185832973854,
+            "step": 14885
+        },
+        {
+            "loss": 14.0166,
+            "grad_norm": 1.7328038215637207,
+            "learning_rate": 0.0005,
+            "epoch": 0.6673426741886509,
+            "step": 14890
+        },
+        {
+            "loss": 13.9913,
+            "grad_norm": 1.7132660150527954,
+            "learning_rate": 0.0005,
+            "epoch": 0.6675667650799164,
+            "step": 14895
+        },
+        {
+            "loss": 13.9751,
+            "grad_norm": 1.872312307357788,
+            "learning_rate": 0.0005,
+            "epoch": 0.6677908559711819,
+            "step": 14900
+        },
+        {
+            "loss": 13.8644,
+            "grad_norm": 1.7622418403625488,
+            "learning_rate": 0.0005,
+            "epoch": 0.6680149468624474,
+            "step": 14905
+        },
+        {
+            "loss": 14.0471,
+            "grad_norm": 1.7850992679595947,
+            "learning_rate": 0.0005,
+            "epoch": 0.668239037753713,
+            "step": 14910
+        },
+        {
+            "loss": 13.9102,
+            "grad_norm": 1.8782780170440674,
+            "learning_rate": 0.0005,
+            "epoch": 0.6684631286449784,
+            "step": 14915
+        },
+        {
+            "loss": 13.9136,
+            "grad_norm": 2.188697099685669,
+            "learning_rate": 0.0005,
+            "epoch": 0.6686872195362439,
+            "step": 14920
+        },
+        {
+            "loss": 13.9138,
+            "grad_norm": 1.8725215196609497,
+            "learning_rate": 0.0005,
+            "epoch": 0.6689113104275094,
+            "step": 14925
+        },
+        {
+            "loss": 13.9785,
+            "grad_norm": 1.9379011392593384,
+            "learning_rate": 0.0005,
+            "epoch": 0.6691354013187749,
+            "step": 14930
+        },
+        {
+            "loss": 13.9834,
+            "grad_norm": 1.845558762550354,
+            "learning_rate": 0.0005,
+            "epoch": 0.6693594922100404,
+            "step": 14935
+        },
+        {
+            "loss": 13.8606,
+            "grad_norm": 1.8981108665466309,
+            "learning_rate": 0.0005,
+            "epoch": 0.6695835831013058,
+            "step": 14940
+        },
+        {
+            "loss": 13.9827,
+            "grad_norm": 2.040807008743286,
+            "learning_rate": 0.0005,
+            "epoch": 0.6698076739925714,
+            "step": 14945
+        },
+        {
+            "loss": 14.0103,
+            "grad_norm": 1.7667853832244873,
+            "learning_rate": 0.0005,
+            "epoch": 0.6700317648838369,
+            "step": 14950
+        },
+        {
+            "loss": 13.9359,
+            "grad_norm": 1.7468235492706299,
+            "learning_rate": 0.0005,
+            "epoch": 0.6702558557751024,
+            "step": 14955
+        },
+        {
+            "loss": 14.0798,
+            "grad_norm": 1.8681683540344238,
+            "learning_rate": 0.0005,
+            "epoch": 0.6704799466663679,
+            "step": 14960
+        },
+        {
+            "loss": 13.9911,
+            "grad_norm": 1.8676295280456543,
+            "learning_rate": 0.0005,
+            "epoch": 0.6707040375576334,
+            "step": 14965
+        },
+        {
+            "loss": 13.9214,
+            "grad_norm": 2.245338201522827,
+            "learning_rate": 0.0005,
+            "epoch": 0.6709281284488988,
+            "step": 14970
+        },
+        {
+            "loss": 13.954,
+            "grad_norm": 1.9026623964309692,
+            "learning_rate": 0.0005,
+            "epoch": 0.6711522193401643,
+            "step": 14975
+        },
+        {
+            "loss": 13.9763,
+            "grad_norm": 1.7272088527679443,
+            "learning_rate": 0.0005,
+            "epoch": 0.6713763102314299,
+            "step": 14980
+        },
+        {
+            "loss": 14.0194,
+            "grad_norm": 1.7873562574386597,
+            "learning_rate": 0.0005,
+            "epoch": 0.6716004011226954,
+            "step": 14985
+        },
+        {
+            "loss": 13.8973,
+            "grad_norm": 1.8993101119995117,
+            "learning_rate": 0.0005,
+            "epoch": 0.6718244920139609,
+            "step": 14990
+        },
+        {
+            "loss": 13.932,
+            "grad_norm": 1.8094158172607422,
+            "learning_rate": 0.0005,
+            "epoch": 0.6720485829052264,
+            "step": 14995
+        },
+        {
+            "loss": 13.9708,
+            "grad_norm": 2.0334019660949707,
+            "learning_rate": 0.0005,
+            "epoch": 0.6722726737964918,
+            "step": 15000
+        },
+        {
+            "eval_loss": 1.7388238906860352,
+            "eval_runtime": 18.2399,
+            "eval_samples_per_second": 898.252,
+            "eval_steps_per_second": 8.059,
+            "epoch": 0.6722726737964918,
+            "step": 15000
+        },
+        {
+            "loss": 13.9549,
+            "grad_norm": 1.8386287689208984,
+            "learning_rate": 0.0005,
+            "epoch": 0.6724967646877573,
+            "step": 15005
+        },
+        {
+            "loss": 13.8927,
+            "grad_norm": 1.9849205017089844,
+            "learning_rate": 0.0005,
+            "epoch": 0.6727208555790228,
+            "step": 15010
+        },
+        {
+            "loss": 13.9941,
+            "grad_norm": 1.9122897386550903,
+            "learning_rate": 0.0005,
+            "epoch": 0.6729449464702884,
+            "step": 15015
+        },
+        {
+            "loss": 13.9255,
+            "grad_norm": 1.7986644506454468,
+            "learning_rate": 0.0005,
+            "epoch": 0.6731690373615539,
+            "step": 15020
+        },
+        {
+            "loss": 13.9692,
+            "grad_norm": 1.9500049352645874,
+            "learning_rate": 0.0005,
+            "epoch": 0.6733931282528194,
+            "step": 15025
+        },
+        {
+            "loss": 13.8283,
+            "grad_norm": 1.7760783433914185,
+            "learning_rate": 0.0005,
+            "epoch": 0.6736172191440848,
+            "step": 15030
+        },
+        {
+            "loss": 13.8723,
+            "grad_norm": 1.832217812538147,
+            "learning_rate": 0.0005,
+            "epoch": 0.6738413100353503,
+            "step": 15035
+        },
+        {
+            "loss": 14.027,
+            "grad_norm": 1.9263299703598022,
+            "learning_rate": 0.0005,
+            "epoch": 0.6740654009266158,
+            "step": 15040
+        },
+        {
+            "loss": 13.9448,
+            "grad_norm": 1.7831065654754639,
+            "learning_rate": 0.0005,
+            "epoch": 0.6742894918178813,
+            "step": 15045
+        },
+        {
+            "loss": 13.9431,
+            "grad_norm": 1.9155333042144775,
+            "learning_rate": 0.0005,
+            "epoch": 0.6745135827091469,
+            "step": 15050
+        },
+        {
+            "loss": 13.943,
+            "grad_norm": 1.8451348543167114,
+            "learning_rate": 0.0005,
+            "epoch": 0.6747376736004124,
+            "step": 15055
+        },
+        {
+            "loss": 13.9246,
+            "grad_norm": 1.8482345342636108,
+            "learning_rate": 0.0005,
+            "epoch": 0.6749617644916778,
+            "step": 15060
+        },
+        {
+            "loss": 13.9124,
+            "grad_norm": 1.8485702276229858,
+            "learning_rate": 0.0005,
+            "epoch": 0.6751858553829433,
+            "step": 15065
+        },
+        {
+            "loss": 13.9741,
+            "grad_norm": 1.697227120399475,
+            "learning_rate": 0.0005,
+            "epoch": 0.6754099462742088,
+            "step": 15070
+        },
+        {
+            "loss": 13.973,
+            "grad_norm": 1.7353583574295044,
+            "learning_rate": 0.0005,
+            "epoch": 0.6756340371654743,
+            "step": 15075
+        },
+        {
+            "loss": 13.9429,
+            "grad_norm": 2.0220677852630615,
+            "learning_rate": 0.0005,
+            "epoch": 0.6758581280567398,
+            "step": 15080
+        },
+        {
+            "loss": 13.9565,
+            "grad_norm": 1.9071723222732544,
+            "learning_rate": 0.0005,
+            "epoch": 0.6760822189480054,
+            "step": 15085
+        },
+        {
+            "loss": 13.9441,
+            "grad_norm": 1.8981002569198608,
+            "learning_rate": 0.0005,
+            "epoch": 0.6763063098392708,
+            "step": 15090
+        },
+        {
+            "loss": 13.8322,
+            "grad_norm": 1.9765799045562744,
+            "learning_rate": 0.0005,
+            "epoch": 0.6765304007305363,
+            "step": 15095
+        },
+        {
+            "loss": 13.9005,
+            "grad_norm": 1.9476999044418335,
+            "learning_rate": 0.0005,
+            "epoch": 0.6767544916218018,
+            "step": 15100
+        },
+        {
+            "loss": 14.0017,
+            "grad_norm": 2.263842821121216,
+            "learning_rate": 0.0005,
+            "epoch": 0.6769785825130673,
+            "step": 15105
+        },
+        {
+            "loss": 13.8683,
+            "grad_norm": 1.9773280620574951,
+            "learning_rate": 0.0005,
+            "epoch": 0.6772026734043328,
+            "step": 15110
+        },
+        {
+            "loss": 13.9226,
+            "grad_norm": 1.9092910289764404,
+            "learning_rate": 0.0005,
+            "epoch": 0.6774267642955983,
+            "step": 15115
+        },
+        {
+            "loss": 13.8926,
+            "grad_norm": 2.052339553833008,
+            "learning_rate": 0.0005,
+            "epoch": 0.6776508551868637,
+            "step": 15120
+        },
+        {
+            "loss": 13.9082,
+            "grad_norm": 1.8536814451217651,
+            "learning_rate": 0.0005,
+            "epoch": 0.6778749460781293,
+            "step": 15125
+        },
+        {
+            "loss": 13.9144,
+            "grad_norm": 1.7902272939682007,
+            "learning_rate": 0.0005,
+            "epoch": 0.6780990369693948,
+            "step": 15130
+        },
+        {
+            "loss": 13.9273,
+            "grad_norm": 1.7612346410751343,
+            "learning_rate": 0.0005,
+            "epoch": 0.6783231278606603,
+            "step": 15135
+        },
+        {
+            "loss": 13.8805,
+            "grad_norm": 1.8486686944961548,
+            "learning_rate": 0.0005,
+            "epoch": 0.6785472187519258,
+            "step": 15140
+        },
+        {
+            "loss": 13.9629,
+            "grad_norm": 1.8244805335998535,
+            "learning_rate": 0.0005,
+            "epoch": 0.6787713096431913,
+            "step": 15145
+        },
+        {
+            "loss": 13.9529,
+            "grad_norm": 1.757832407951355,
+            "learning_rate": 0.0005,
+            "epoch": 0.6789954005344567,
+            "step": 15150
+        },
+        {
+            "loss": 14.0222,
+            "grad_norm": 1.7775987386703491,
+            "learning_rate": 0.0005,
+            "epoch": 0.6792194914257222,
+            "step": 15155
+        },
+        {
+            "loss": 13.9338,
+            "grad_norm": 1.7991831302642822,
+            "learning_rate": 0.0005,
+            "epoch": 0.6794435823169878,
+            "step": 15160
+        },
+        {
+            "loss": 13.9246,
+            "grad_norm": 1.7518317699432373,
+            "learning_rate": 0.0005,
+            "epoch": 0.6796676732082533,
+            "step": 15165
+        },
+        {
+            "loss": 14.0307,
+            "grad_norm": 1.7775969505310059,
+            "learning_rate": 0.0005,
+            "epoch": 0.6798917640995188,
+            "step": 15170
+        },
+        {
+            "loss": 14.0334,
+            "grad_norm": 1.9538028240203857,
+            "learning_rate": 0.0005,
+            "epoch": 0.6801158549907843,
+            "step": 15175
+        },
+        {
+            "loss": 14.0499,
+            "grad_norm": 1.7270337343215942,
+            "learning_rate": 0.0005,
+            "epoch": 0.6803399458820497,
+            "step": 15180
+        },
+        {
+            "loss": 13.8981,
+            "grad_norm": 1.7424579858779907,
+            "learning_rate": 0.0005,
+            "epoch": 0.6805640367733152,
+            "step": 15185
+        },
+        {
+            "loss": 14.0244,
+            "grad_norm": 1.9071272611618042,
+            "learning_rate": 0.0005,
+            "epoch": 0.6807881276645807,
+            "step": 15190
+        },
+        {
+            "loss": 14.0035,
+            "grad_norm": 1.7010276317596436,
+            "learning_rate": 0.0005,
+            "epoch": 0.6810122185558463,
+            "step": 15195
+        },
+        {
+            "loss": 13.9618,
+            "grad_norm": 1.6710230112075806,
+            "learning_rate": 0.0005,
+            "epoch": 0.6812363094471118,
+            "step": 15200
+        },
+        {
+            "loss": 14.0453,
+            "grad_norm": 1.7517222166061401,
+            "learning_rate": 0.0005,
+            "epoch": 0.6814604003383773,
+            "step": 15205
+        },
+        {
+            "loss": 13.961,
+            "grad_norm": 1.8641791343688965,
+            "learning_rate": 0.0005,
+            "epoch": 0.6816844912296427,
+            "step": 15210
+        },
+        {
+            "loss": 14.0737,
+            "grad_norm": 1.7744903564453125,
+            "learning_rate": 0.0005,
+            "epoch": 0.6819085821209082,
+            "step": 15215
+        },
+        {
+            "loss": 13.9684,
+            "grad_norm": 1.6295077800750732,
+            "learning_rate": 0.0005,
+            "epoch": 0.6821326730121737,
+            "step": 15220
+        },
+        {
+            "loss": 13.8636,
+            "grad_norm": 1.8711214065551758,
+            "learning_rate": 0.0005,
+            "epoch": 0.6823567639034392,
+            "step": 15225
+        },
+        {
+            "loss": 13.9437,
+            "grad_norm": 1.8667914867401123,
+            "learning_rate": 0.0005,
+            "epoch": 0.6825808547947048,
+            "step": 15230
+        },
+        {
+            "loss": 14.0297,
+            "grad_norm": 1.7881425619125366,
+            "learning_rate": 0.0005,
+            "epoch": 0.6828049456859703,
+            "step": 15235
+        },
+        {
+            "loss": 13.9598,
+            "grad_norm": 2.000901460647583,
+            "learning_rate": 0.0005,
+            "epoch": 0.6830290365772357,
+            "step": 15240
+        },
+        {
+            "loss": 13.9195,
+            "grad_norm": 1.8856955766677856,
+            "learning_rate": 0.0005,
+            "epoch": 0.6832531274685012,
+            "step": 15245
+        },
+        {
+            "loss": 13.9155,
+            "grad_norm": 1.8447706699371338,
+            "learning_rate": 0.0005,
+            "epoch": 0.6834772183597667,
+            "step": 15250
+        },
+        {
+            "loss": 13.8523,
+            "grad_norm": 1.8459632396697998,
+            "learning_rate": 0.0005,
+            "epoch": 0.6837013092510322,
+            "step": 15255
+        },
+        {
+            "loss": 13.8948,
+            "grad_norm": 1.915228009223938,
+            "learning_rate": 0.0005,
+            "epoch": 0.6839254001422977,
+            "step": 15260
+        },
+        {
+            "loss": 13.9318,
+            "grad_norm": 1.9591858386993408,
+            "learning_rate": 0.0005,
+            "epoch": 0.6841494910335633,
+            "step": 15265
+        },
+        {
+            "loss": 13.9371,
+            "grad_norm": 1.8608108758926392,
+            "learning_rate": 0.0005,
+            "epoch": 0.6843735819248287,
+            "step": 15270
+        },
+        {
+            "loss": 14.0048,
+            "grad_norm": 1.806567907333374,
+            "learning_rate": 0.0005,
+            "epoch": 0.6845976728160942,
+            "step": 15275
+        },
+        {
+            "loss": 13.989,
+            "grad_norm": 1.7369107007980347,
+            "learning_rate": 0.0005,
+            "epoch": 0.6848217637073597,
+            "step": 15280
+        },
+        {
+            "loss": 13.9489,
+            "grad_norm": 1.8193225860595703,
+            "learning_rate": 0.0005,
+            "epoch": 0.6850458545986252,
+            "step": 15285
+        },
+        {
+            "loss": 13.8948,
+            "grad_norm": 1.8266242742538452,
+            "learning_rate": 0.0005,
+            "epoch": 0.6852699454898907,
+            "step": 15290
+        },
+        {
+            "loss": 14.0153,
+            "grad_norm": 1.8363949060440063,
+            "learning_rate": 0.0005,
+            "epoch": 0.6854940363811562,
+            "step": 15295
+        },
+        {
+            "loss": 14.0074,
+            "grad_norm": 1.840710997581482,
+            "learning_rate": 0.0005,
+            "epoch": 0.6857181272724217,
+            "step": 15300
+        },
+        {
+            "loss": 14.0341,
+            "grad_norm": 1.8140329122543335,
+            "learning_rate": 0.0005,
+            "epoch": 0.6859422181636872,
+            "step": 15305
+        },
+        {
+            "loss": 14.004,
+            "grad_norm": 1.7864294052124023,
+            "learning_rate": 0.0005,
+            "epoch": 0.6861663090549527,
+            "step": 15310
+        },
+        {
+            "loss": 13.959,
+            "grad_norm": 1.866920828819275,
+            "learning_rate": 0.0005,
+            "epoch": 0.6863903999462182,
+            "step": 15315
+        },
+        {
+            "loss": 13.96,
+            "grad_norm": 1.7539829015731812,
+            "learning_rate": 0.0005,
+            "epoch": 0.6866144908374837,
+            "step": 15320
+        },
+        {
+            "loss": 13.9228,
+            "grad_norm": 1.7904281616210938,
+            "learning_rate": 0.0005,
+            "epoch": 0.6868385817287492,
+            "step": 15325
+        },
+        {
+            "loss": 14.0567,
+            "grad_norm": 1.8457413911819458,
+            "learning_rate": 0.0005,
+            "epoch": 0.6870626726200146,
+            "step": 15330
+        },
+        {
+            "loss": 13.9857,
+            "grad_norm": 1.89803946018219,
+            "learning_rate": 0.0005,
+            "epoch": 0.6872867635112802,
+            "step": 15335
+        },
+        {
+            "loss": 13.8866,
+            "grad_norm": 1.962873101234436,
+            "learning_rate": 0.0005,
+            "epoch": 0.6875108544025457,
+            "step": 15340
+        },
+        {
+            "loss": 13.9703,
+            "grad_norm": 1.8302043676376343,
+            "learning_rate": 0.0005,
+            "epoch": 0.6877349452938112,
+            "step": 15345
+        },
+        {
+            "loss": 13.9963,
+            "grad_norm": 1.7163232564926147,
+            "learning_rate": 0.0005,
+            "epoch": 0.6879590361850767,
+            "step": 15350
+        },
+        {
+            "loss": 14.0319,
+            "grad_norm": 1.7698335647583008,
+            "learning_rate": 0.0005,
+            "epoch": 0.6881831270763422,
+            "step": 15355
+        },
+        {
+            "loss": 13.8204,
+            "grad_norm": 1.7495076656341553,
+            "learning_rate": 0.0005,
+            "epoch": 0.6884072179676076,
+            "step": 15360
+        },
+        {
+            "loss": 14.0892,
+            "grad_norm": 2.0041182041168213,
+            "learning_rate": 0.0005,
+            "epoch": 0.6886313088588731,
+            "step": 15365
+        },
+        {
+            "loss": 13.9974,
+            "grad_norm": 1.8604499101638794,
+            "learning_rate": 0.0005,
+            "epoch": 0.6888553997501387,
+            "step": 15370
+        },
+        {
+            "loss": 14.0546,
+            "grad_norm": 1.8129802942276,
+            "learning_rate": 0.0005,
+            "epoch": 0.6890794906414042,
+            "step": 15375
+        },
+        {
+            "loss": 13.9537,
+            "grad_norm": 1.8278909921646118,
+            "learning_rate": 0.0005,
+            "epoch": 0.6893035815326697,
+            "step": 15380
+        },
+        {
+            "loss": 13.9457,
+            "grad_norm": 1.8413242101669312,
+            "learning_rate": 0.0005,
+            "epoch": 0.6895276724239352,
+            "step": 15385
+        },
+        {
+            "loss": 13.9907,
+            "grad_norm": 1.9248651266098022,
+            "learning_rate": 0.0005,
+            "epoch": 0.6897517633152006,
+            "step": 15390
+        },
+        {
+            "loss": 13.8831,
+            "grad_norm": 1.6882672309875488,
+            "learning_rate": 0.0005,
+            "epoch": 0.6899758542064661,
+            "step": 15395
+        },
+        {
+            "loss": 13.9649,
+            "grad_norm": 2.0407419204711914,
+            "learning_rate": 0.0005,
+            "epoch": 0.6901999450977316,
+            "step": 15400
+        },
+        {
+            "loss": 14.0037,
+            "grad_norm": 1.9546421766281128,
+            "learning_rate": 0.0005,
+            "epoch": 0.6904240359889972,
+            "step": 15405
+        },
+        {
+            "loss": 14.0474,
+            "grad_norm": 1.8498766422271729,
+            "learning_rate": 0.0005,
+            "epoch": 0.6906481268802627,
+            "step": 15410
+        },
+        {
+            "loss": 14.0266,
+            "grad_norm": 1.8743306398391724,
+            "learning_rate": 0.0005,
+            "epoch": 0.6908722177715282,
+            "step": 15415
+        },
+        {
+            "loss": 13.9242,
+            "grad_norm": 1.8721219301223755,
+            "learning_rate": 0.0005,
+            "epoch": 0.6910963086627936,
+            "step": 15420
+        },
+        {
+            "loss": 13.9808,
+            "grad_norm": 1.8055014610290527,
+            "learning_rate": 0.0005,
+            "epoch": 0.6913203995540591,
+            "step": 15425
+        },
+        {
+            "loss": 13.934,
+            "grad_norm": 1.7605301141738892,
+            "learning_rate": 0.0005,
+            "epoch": 0.6915444904453246,
+            "step": 15430
+        },
+        {
+            "loss": 13.8644,
+            "grad_norm": 1.6423132419586182,
+            "learning_rate": 0.0005,
+            "epoch": 0.6917685813365901,
+            "step": 15435
+        },
+        {
+            "loss": 13.9572,
+            "grad_norm": 1.7309364080429077,
+            "learning_rate": 0.0005,
+            "epoch": 0.6919926722278557,
+            "step": 15440
+        },
+        {
+            "loss": 14.0399,
+            "grad_norm": 1.644025206565857,
+            "learning_rate": 0.0005,
+            "epoch": 0.6922167631191212,
+            "step": 15445
+        },
+        {
+            "loss": 13.9544,
+            "grad_norm": 1.9230233430862427,
+            "learning_rate": 0.0005,
+            "epoch": 0.6924408540103866,
+            "step": 15450
+        },
+        {
+            "loss": 13.9615,
+            "grad_norm": 1.8161972761154175,
+            "learning_rate": 0.0005,
+            "epoch": 0.6926649449016521,
+            "step": 15455
+        },
+        {
+            "loss": 13.9694,
+            "grad_norm": 1.834704041481018,
+            "learning_rate": 0.0005,
+            "epoch": 0.6928890357929176,
+            "step": 15460
+        },
+        {
+            "loss": 13.8995,
+            "grad_norm": 1.961165428161621,
+            "learning_rate": 0.0005,
+            "epoch": 0.6931131266841831,
+            "step": 15465
+        },
+        {
+            "loss": 13.841,
+            "grad_norm": 1.9340486526489258,
+            "learning_rate": 0.0005,
+            "epoch": 0.6933372175754486,
+            "step": 15470
+        },
+        {
+            "loss": 14.0106,
+            "grad_norm": 1.840572476387024,
+            "learning_rate": 0.0005,
+            "epoch": 0.693561308466714,
+            "step": 15475
+        },
+        {
+            "loss": 14.0718,
+            "grad_norm": 1.8550306558609009,
+            "learning_rate": 0.0005,
+            "epoch": 0.6937853993579796,
+            "step": 15480
+        },
+        {
+            "loss": 13.9573,
+            "grad_norm": 1.7712377309799194,
+            "learning_rate": 0.0005,
+            "epoch": 0.6940094902492451,
+            "step": 15485
+        },
+        {
+            "loss": 13.8744,
+            "grad_norm": 1.6746519804000854,
+            "learning_rate": 0.0005,
+            "epoch": 0.6942335811405106,
+            "step": 15490
+        },
+        {
+            "loss": 13.9816,
+            "grad_norm": 1.7693535089492798,
+            "learning_rate": 0.0005,
+            "epoch": 0.6944576720317761,
+            "step": 15495
+        },
+        {
+            "loss": 13.9546,
+            "grad_norm": 1.792589783668518,
+            "learning_rate": 0.0005,
+            "epoch": 0.6946817629230416,
+            "step": 15500
+        },
+        {
+            "eval_loss": 1.7409549951553345,
+            "eval_runtime": 19.0449,
+            "eval_samples_per_second": 860.281,
+            "eval_steps_per_second": 7.719,
+            "epoch": 0.6946817629230416,
+            "step": 15500
+        },
+        {
+            "loss": 13.8749,
+            "grad_norm": 1.7078382968902588,
+            "learning_rate": 0.0005,
+            "epoch": 0.694905853814307,
+            "step": 15505
+        },
+        {
+            "loss": 13.949,
+            "grad_norm": 1.8692890405654907,
+            "learning_rate": 0.0005,
+            "epoch": 0.6951299447055725,
+            "step": 15510
+        },
+        {
+            "loss": 13.9288,
+            "grad_norm": 1.9514987468719482,
+            "learning_rate": 0.0005,
+            "epoch": 0.6953540355968381,
+            "step": 15515
+        },
+        {
+            "loss": 13.9926,
+            "grad_norm": 1.945313811302185,
+            "learning_rate": 0.0005,
+            "epoch": 0.6955781264881036,
+            "step": 15520
+        },
+        {
+            "loss": 13.9282,
+            "grad_norm": 1.845374584197998,
+            "learning_rate": 0.0005,
+            "epoch": 0.6958022173793691,
+            "step": 15525
+        },
+        {
+            "loss": 14.0666,
+            "grad_norm": 1.8912990093231201,
+            "learning_rate": 0.0005,
+            "epoch": 0.6960263082706346,
+            "step": 15530
+        },
+        {
+            "loss": 13.9056,
+            "grad_norm": 1.8223357200622559,
+            "learning_rate": 0.0005,
+            "epoch": 0.6962503991619,
+            "step": 15535
+        },
+        {
+            "loss": 13.9665,
+            "grad_norm": 1.7922083139419556,
+            "learning_rate": 0.0005,
+            "epoch": 0.6964744900531655,
+            "step": 15540
+        },
+        {
+            "loss": 13.9223,
+            "grad_norm": 1.9126029014587402,
+            "learning_rate": 0.0005,
+            "epoch": 0.696698580944431,
+            "step": 15545
+        },
+        {
+            "loss": 14.0502,
+            "grad_norm": 1.927013874053955,
+            "learning_rate": 0.0005,
+            "epoch": 0.6969226718356966,
+            "step": 15550
+        },
+        {
+            "loss": 13.9175,
+            "grad_norm": 1.7995493412017822,
+            "learning_rate": 0.0005,
+            "epoch": 0.6971467627269621,
+            "step": 15555
+        },
+        {
+            "loss": 14.0,
+            "grad_norm": 1.9804531335830688,
+            "learning_rate": 0.0005,
+            "epoch": 0.6973708536182276,
+            "step": 15560
+        },
+        {
+            "loss": 13.945,
+            "grad_norm": 2.095614194869995,
+            "learning_rate": 0.0005,
+            "epoch": 0.697594944509493,
+            "step": 15565
+        },
+        {
+            "loss": 13.946,
+            "grad_norm": 1.8693947792053223,
+            "learning_rate": 0.0005,
+            "epoch": 0.6978190354007585,
+            "step": 15570
+        },
+        {
+            "loss": 13.9907,
+            "grad_norm": 1.7160202264785767,
+            "learning_rate": 0.0005,
+            "epoch": 0.698043126292024,
+            "step": 15575
+        },
+        {
+            "loss": 13.9814,
+            "grad_norm": 1.9428715705871582,
+            "learning_rate": 0.0005,
+            "epoch": 0.6982672171832895,
+            "step": 15580
+        },
+        {
+            "loss": 13.9458,
+            "grad_norm": 1.8247159719467163,
+            "learning_rate": 0.0005,
+            "epoch": 0.6984913080745551,
+            "step": 15585
+        },
+        {
+            "loss": 13.894,
+            "grad_norm": 1.742426872253418,
+            "learning_rate": 0.0005,
+            "epoch": 0.6987153989658206,
+            "step": 15590
+        },
+        {
+            "loss": 13.9269,
+            "grad_norm": 1.9104456901550293,
+            "learning_rate": 0.0005,
+            "epoch": 0.698939489857086,
+            "step": 15595
+        },
+        {
+            "loss": 13.9191,
+            "grad_norm": 1.842275619506836,
+            "learning_rate": 0.0005,
+            "epoch": 0.6991635807483515,
+            "step": 15600
+        },
+        {
+            "loss": 14.024,
+            "grad_norm": 1.793081283569336,
+            "learning_rate": 0.0005,
+            "epoch": 0.699387671639617,
+            "step": 15605
+        },
+        {
+            "loss": 13.8893,
+            "grad_norm": 1.7978259325027466,
+            "learning_rate": 0.0005,
+            "epoch": 0.6996117625308825,
+            "step": 15610
+        },
+        {
+            "loss": 13.9192,
+            "grad_norm": 2.2107441425323486,
+            "learning_rate": 0.0005,
+            "epoch": 0.699835853422148,
+            "step": 15615
+        },
+        {
+            "loss": 14.043,
+            "grad_norm": 2.238145112991333,
+            "learning_rate": 0.0005,
+            "epoch": 0.7000599443134136,
+            "step": 15620
+        },
+        {
+            "loss": 13.9411,
+            "grad_norm": 2.0489559173583984,
+            "learning_rate": 0.0005,
+            "epoch": 0.700284035204679,
+            "step": 15625
+        },
+        {
+            "loss": 13.8079,
+            "grad_norm": 1.8659436702728271,
+            "learning_rate": 0.0005,
+            "epoch": 0.7005081260959445,
+            "step": 15630
+        },
+        {
+            "loss": 13.9392,
+            "grad_norm": 1.7955938577651978,
+            "learning_rate": 0.0005,
+            "epoch": 0.70073221698721,
+            "step": 15635
+        },
+        {
+            "loss": 13.9242,
+            "grad_norm": 1.8826559782028198,
+            "learning_rate": 0.0005,
+            "epoch": 0.7009563078784755,
+            "step": 15640
+        },
+        {
+            "loss": 13.9201,
+            "grad_norm": 1.858927607536316,
+            "learning_rate": 0.0005,
+            "epoch": 0.701180398769741,
+            "step": 15645
+        },
+        {
+            "loss": 13.9107,
+            "grad_norm": 1.839073896408081,
+            "learning_rate": 0.0005,
+            "epoch": 0.7014044896610065,
+            "step": 15650
+        },
+        {
+            "loss": 13.9317,
+            "grad_norm": 1.8229670524597168,
+            "learning_rate": 0.0005,
+            "epoch": 0.701628580552272,
+            "step": 15655
+        },
+        {
+            "loss": 13.91,
+            "grad_norm": 1.8194196224212646,
+            "learning_rate": 0.0005,
+            "epoch": 0.7018526714435375,
+            "step": 15660
+        },
+        {
+            "loss": 14.1022,
+            "grad_norm": 1.8115154504776,
+            "learning_rate": 0.0005,
+            "epoch": 0.702076762334803,
+            "step": 15665
+        },
+        {
+            "loss": 13.9124,
+            "grad_norm": 1.8422259092330933,
+            "learning_rate": 0.0005,
+            "epoch": 0.7023008532260685,
+            "step": 15670
+        },
+        {
+            "loss": 13.8756,
+            "grad_norm": 1.6191073656082153,
+            "learning_rate": 0.0005,
+            "epoch": 0.702524944117334,
+            "step": 15675
+        },
+        {
+            "loss": 13.9934,
+            "grad_norm": 1.692289113998413,
+            "learning_rate": 0.0005,
+            "epoch": 0.7027490350085995,
+            "step": 15680
+        },
+        {
+            "loss": 13.9409,
+            "grad_norm": 2.0738937854766846,
+            "learning_rate": 0.0005,
+            "epoch": 0.7029731258998649,
+            "step": 15685
+        },
+        {
+            "loss": 13.8589,
+            "grad_norm": 2.1270904541015625,
+            "learning_rate": 0.0005,
+            "epoch": 0.7031972167911305,
+            "step": 15690
+        },
+        {
+            "loss": 13.9472,
+            "grad_norm": 1.8903199434280396,
+            "learning_rate": 0.0005,
+            "epoch": 0.703421307682396,
+            "step": 15695
+        },
+        {
+            "loss": 13.9934,
+            "grad_norm": 1.8520416021347046,
+            "learning_rate": 0.0005,
+            "epoch": 0.7036453985736615,
+            "step": 15700
+        },
+        {
+            "loss": 13.8661,
+            "grad_norm": 1.8308141231536865,
+            "learning_rate": 0.0005,
+            "epoch": 0.703869489464927,
+            "step": 15705
+        },
+        {
+            "loss": 13.9935,
+            "grad_norm": 1.8230431079864502,
+            "learning_rate": 0.0005,
+            "epoch": 0.7040935803561925,
+            "step": 15710
+        },
+        {
+            "loss": 13.8179,
+            "grad_norm": 1.8655309677124023,
+            "learning_rate": 0.0005,
+            "epoch": 0.7043176712474579,
+            "step": 15715
+        },
+        {
+            "loss": 14.0309,
+            "grad_norm": 1.8119747638702393,
+            "learning_rate": 0.0005,
+            "epoch": 0.7045417621387234,
+            "step": 15720
+        },
+        {
+            "loss": 13.982,
+            "grad_norm": 1.8590638637542725,
+            "learning_rate": 0.0005,
+            "epoch": 0.704765853029989,
+            "step": 15725
+        },
+        {
+            "loss": 13.8865,
+            "grad_norm": 1.9631258249282837,
+            "learning_rate": 0.0005,
+            "epoch": 0.7049899439212545,
+            "step": 15730
+        },
+        {
+            "loss": 13.8423,
+            "grad_norm": 1.8070425987243652,
+            "learning_rate": 0.0005,
+            "epoch": 0.70521403481252,
+            "step": 15735
+        },
+        {
+            "loss": 13.9572,
+            "grad_norm": 1.8490524291992188,
+            "learning_rate": 0.0005,
+            "epoch": 0.7054381257037855,
+            "step": 15740
+        },
+        {
+            "loss": 13.9006,
+            "grad_norm": 1.8480288982391357,
+            "learning_rate": 0.0005,
+            "epoch": 0.7056622165950509,
+            "step": 15745
+        },
+        {
+            "loss": 13.9108,
+            "grad_norm": 1.937630295753479,
+            "learning_rate": 0.0005,
+            "epoch": 0.7058863074863164,
+            "step": 15750
+        },
+        {
+            "loss": 14.0033,
+            "grad_norm": 1.6791224479675293,
+            "learning_rate": 0.0005,
+            "epoch": 0.7061103983775819,
+            "step": 15755
+        },
+        {
+            "loss": 13.8934,
+            "grad_norm": 1.8505795001983643,
+            "learning_rate": 0.0005,
+            "epoch": 0.7063344892688475,
+            "step": 15760
+        },
+        {
+            "loss": 13.8907,
+            "grad_norm": 1.9696643352508545,
+            "learning_rate": 0.0005,
+            "epoch": 0.706558580160113,
+            "step": 15765
+        },
+        {
+            "loss": 13.9201,
+            "grad_norm": 1.9013051986694336,
+            "learning_rate": 0.0005,
+            "epoch": 0.7067826710513785,
+            "step": 15770
+        },
+        {
+            "loss": 13.9411,
+            "grad_norm": 2.011505365371704,
+            "learning_rate": 0.0005,
+            "epoch": 0.7070067619426439,
+            "step": 15775
+        },
+        {
+            "loss": 14.0902,
+            "grad_norm": 1.9483027458190918,
+            "learning_rate": 0.0005,
+            "epoch": 0.7072308528339094,
+            "step": 15780
+        },
+        {
+            "loss": 13.9785,
+            "grad_norm": 1.9703984260559082,
+            "learning_rate": 0.0005,
+            "epoch": 0.7074549437251749,
+            "step": 15785
+        },
+        {
+            "loss": 13.9829,
+            "grad_norm": 1.9221055507659912,
+            "learning_rate": 0.0005,
+            "epoch": 0.7076790346164404,
+            "step": 15790
+        },
+        {
+            "loss": 13.9331,
+            "grad_norm": 1.96420156955719,
+            "learning_rate": 0.0005,
+            "epoch": 0.707903125507706,
+            "step": 15795
+        },
+        {
+            "loss": 13.9016,
+            "grad_norm": 1.7140847444534302,
+            "learning_rate": 0.0005,
+            "epoch": 0.7081272163989715,
+            "step": 15800
+        },
+        {
+            "loss": 13.9334,
+            "grad_norm": 1.9515072107315063,
+            "learning_rate": 0.0005,
+            "epoch": 0.7083513072902369,
+            "step": 15805
+        },
+        {
+            "loss": 13.9258,
+            "grad_norm": 1.709166407585144,
+            "learning_rate": 0.0005,
+            "epoch": 0.7085753981815024,
+            "step": 15810
+        },
+        {
+            "loss": 13.9812,
+            "grad_norm": 1.7489274740219116,
+            "learning_rate": 0.0005,
+            "epoch": 0.7087994890727679,
+            "step": 15815
+        },
+        {
+            "loss": 13.9408,
+            "grad_norm": 2.133345127105713,
+            "learning_rate": 0.0005,
+            "epoch": 0.7090235799640334,
+            "step": 15820
+        },
+        {
+            "loss": 13.9426,
+            "grad_norm": 1.8625730276107788,
+            "learning_rate": 0.0005,
+            "epoch": 0.7092476708552989,
+            "step": 15825
+        },
+        {
+            "loss": 13.9732,
+            "grad_norm": 1.6733115911483765,
+            "learning_rate": 0.0005,
+            "epoch": 0.7094717617465645,
+            "step": 15830
+        },
+        {
+            "loss": 14.0664,
+            "grad_norm": 1.8962249755859375,
+            "learning_rate": 0.0005,
+            "epoch": 0.7096958526378299,
+            "step": 15835
+        },
+        {
+            "loss": 14.0338,
+            "grad_norm": 1.839948058128357,
+            "learning_rate": 0.0005,
+            "epoch": 0.7099199435290954,
+            "step": 15840
+        },
+        {
+            "loss": 13.9143,
+            "grad_norm": 1.7848726511001587,
+            "learning_rate": 0.0005,
+            "epoch": 0.7101440344203609,
+            "step": 15845
+        },
+        {
+            "loss": 13.8563,
+            "grad_norm": 1.9577068090438843,
+            "learning_rate": 0.0005,
+            "epoch": 0.7103681253116264,
+            "step": 15850
+        },
+        {
+            "loss": 13.9217,
+            "grad_norm": 1.9215444326400757,
+            "learning_rate": 0.0005,
+            "epoch": 0.7105922162028919,
+            "step": 15855
+        },
+        {
+            "loss": 13.9757,
+            "grad_norm": 1.9216139316558838,
+            "learning_rate": 0.0005,
+            "epoch": 0.7108163070941574,
+            "step": 15860
+        },
+        {
+            "loss": 13.9598,
+            "grad_norm": 1.9376696348190308,
+            "learning_rate": 0.0005,
+            "epoch": 0.7110403979854228,
+            "step": 15865
+        },
+        {
+            "loss": 13.9578,
+            "grad_norm": 1.8964476585388184,
+            "learning_rate": 0.0005,
+            "epoch": 0.7112644888766884,
+            "step": 15870
+        },
+        {
+            "loss": 13.9611,
+            "grad_norm": 1.7792725563049316,
+            "learning_rate": 0.0005,
+            "epoch": 0.7114885797679539,
+            "step": 15875
+        },
+        {
+            "loss": 14.0214,
+            "grad_norm": 1.8994535207748413,
+            "learning_rate": 0.0005,
+            "epoch": 0.7117126706592194,
+            "step": 15880
+        },
+        {
+            "loss": 13.9837,
+            "grad_norm": 1.8821009397506714,
+            "learning_rate": 0.0005,
+            "epoch": 0.7119367615504849,
+            "step": 15885
+        },
+        {
+            "loss": 14.0434,
+            "grad_norm": 1.858879804611206,
+            "learning_rate": 0.0005,
+            "epoch": 0.7121608524417504,
+            "step": 15890
+        },
+        {
+            "loss": 13.9523,
+            "grad_norm": 1.7259469032287598,
+            "learning_rate": 0.0005,
+            "epoch": 0.7123849433330158,
+            "step": 15895
+        },
+        {
+            "loss": 13.8842,
+            "grad_norm": 1.9253323078155518,
+            "learning_rate": 0.0005,
+            "epoch": 0.7126090342242813,
+            "step": 15900
+        },
+        {
+            "loss": 13.9879,
+            "grad_norm": 1.9537336826324463,
+            "learning_rate": 0.0005,
+            "epoch": 0.7128331251155469,
+            "step": 15905
+        },
+        {
+            "loss": 13.9387,
+            "grad_norm": 1.8602211475372314,
+            "learning_rate": 0.0005,
+            "epoch": 0.7130572160068124,
+            "step": 15910
+        },
+        {
+            "loss": 13.8637,
+            "grad_norm": 1.9752931594848633,
+            "learning_rate": 0.0005,
+            "epoch": 0.7132813068980779,
+            "step": 15915
+        },
+        {
+            "loss": 13.7979,
+            "grad_norm": 1.6928125619888306,
+            "learning_rate": 0.0005,
+            "epoch": 0.7135053977893434,
+            "step": 15920
+        },
+        {
+            "loss": 13.9197,
+            "grad_norm": 2.161198854446411,
+            "learning_rate": 0.0005,
+            "epoch": 0.7137294886806088,
+            "step": 15925
+        },
+        {
+            "loss": 13.993,
+            "grad_norm": 1.9235631227493286,
+            "learning_rate": 0.0005,
+            "epoch": 0.7139535795718743,
+            "step": 15930
+        },
+        {
+            "loss": 13.8041,
+            "grad_norm": 1.714410424232483,
+            "learning_rate": 0.0005,
+            "epoch": 0.7141776704631398,
+            "step": 15935
+        },
+        {
+            "loss": 13.9278,
+            "grad_norm": 1.8958978652954102,
+            "learning_rate": 0.0005,
+            "epoch": 0.7144017613544054,
+            "step": 15940
+        },
+        {
+            "loss": 13.934,
+            "grad_norm": 1.819340705871582,
+            "learning_rate": 0.0005,
+            "epoch": 0.7146258522456709,
+            "step": 15945
+        },
+        {
+            "loss": 13.8556,
+            "grad_norm": 1.6928635835647583,
+            "learning_rate": 0.0005,
+            "epoch": 0.7148499431369364,
+            "step": 15950
+        },
+        {
+            "loss": 13.913,
+            "grad_norm": 1.6891690492630005,
+            "learning_rate": 0.0005,
+            "epoch": 0.7150740340282018,
+            "step": 15955
+        },
+        {
+            "loss": 13.967,
+            "grad_norm": 1.8896440267562866,
+            "learning_rate": 0.0005,
+            "epoch": 0.7152981249194673,
+            "step": 15960
+        },
+        {
+            "loss": 13.9234,
+            "grad_norm": 1.8856749534606934,
+            "learning_rate": 0.0005,
+            "epoch": 0.7155222158107328,
+            "step": 15965
+        },
+        {
+            "loss": 14.0063,
+            "grad_norm": 1.8018680810928345,
+            "learning_rate": 0.0005,
+            "epoch": 0.7157463067019983,
+            "step": 15970
+        },
+        {
+            "loss": 13.8262,
+            "grad_norm": 1.8745312690734863,
+            "learning_rate": 0.0005,
+            "epoch": 0.7159703975932639,
+            "step": 15975
+        },
+        {
+            "loss": 13.9973,
+            "grad_norm": 1.877651572227478,
+            "learning_rate": 0.0005,
+            "epoch": 0.7161944884845294,
+            "step": 15980
+        },
+        {
+            "loss": 13.9225,
+            "grad_norm": 1.7573840618133545,
+            "learning_rate": 0.0005,
+            "epoch": 0.7164185793757948,
+            "step": 15985
+        },
+        {
+            "loss": 13.9996,
+            "grad_norm": 1.7413743734359741,
+            "learning_rate": 0.0005,
+            "epoch": 0.7166426702670603,
+            "step": 15990
+        },
+        {
+            "loss": 13.9276,
+            "grad_norm": 1.8496475219726562,
+            "learning_rate": 0.0005,
+            "epoch": 0.7168667611583258,
+            "step": 15995
+        },
+        {
+            "loss": 14.0066,
+            "grad_norm": 1.8488664627075195,
+            "learning_rate": 0.0005,
+            "epoch": 0.7170908520495913,
+            "step": 16000
+        },
+        {
+            "eval_loss": 1.7406035661697388,
+            "eval_runtime": 18.3143,
+            "eval_samples_per_second": 894.602,
+            "eval_steps_per_second": 8.027,
+            "epoch": 0.7170908520495913,
+            "step": 16000
+        },
+        {
+            "loss": 14.0059,
+            "grad_norm": 1.8590061664581299,
+            "learning_rate": 0.0005,
+            "epoch": 0.7173149429408568,
+            "step": 16005
+        },
+        {
+            "loss": 13.9813,
+            "grad_norm": 2.056410074234009,
+            "learning_rate": 0.0005,
+            "epoch": 0.7175390338321224,
+            "step": 16010
+        },
+        {
+            "loss": 13.9623,
+            "grad_norm": 1.8297349214553833,
+            "learning_rate": 0.0005,
+            "epoch": 0.7177631247233878,
+            "step": 16015
+        },
+        {
+            "loss": 14.0261,
+            "grad_norm": 1.8574327230453491,
+            "learning_rate": 0.0005,
+            "epoch": 0.7179872156146533,
+            "step": 16020
+        },
+        {
+            "loss": 13.9457,
+            "grad_norm": 1.8076415061950684,
+            "learning_rate": 0.0005,
+            "epoch": 0.7182113065059188,
+            "step": 16025
+        },
+        {
+            "loss": 13.8557,
+            "grad_norm": 1.8541926145553589,
+            "learning_rate": 0.0005,
+            "epoch": 0.7184353973971843,
+            "step": 16030
+        },
+        {
+            "loss": 13.9438,
+            "grad_norm": 1.8825889825820923,
+            "learning_rate": 0.0005,
+            "epoch": 0.7186594882884498,
+            "step": 16035
+        },
+        {
+            "loss": 13.891,
+            "grad_norm": 1.7605642080307007,
+            "learning_rate": 0.0005,
+            "epoch": 0.7188835791797152,
+            "step": 16040
+        },
+        {
+            "loss": 13.9985,
+            "grad_norm": 1.8549124002456665,
+            "learning_rate": 0.0005,
+            "epoch": 0.7191076700709808,
+            "step": 16045
+        },
+        {
+            "loss": 14.0044,
+            "grad_norm": 1.8616878986358643,
+            "learning_rate": 0.0005,
+            "epoch": 0.7193317609622463,
+            "step": 16050
+        },
+        {
+            "loss": 13.986,
+            "grad_norm": 2.120058059692383,
+            "learning_rate": 0.0005,
+            "epoch": 0.7195558518535118,
+            "step": 16055
+        },
+        {
+            "loss": 13.9578,
+            "grad_norm": 1.7615188360214233,
+            "learning_rate": 0.0005,
+            "epoch": 0.7197799427447773,
+            "step": 16060
+        },
+        {
+            "loss": 13.9921,
+            "grad_norm": 1.9902580976486206,
+            "learning_rate": 0.0005,
+            "epoch": 0.7200040336360428,
+            "step": 16065
+        },
+        {
+            "loss": 13.8947,
+            "grad_norm": 1.999809980392456,
+            "learning_rate": 0.0005,
+            "epoch": 0.7202281245273082,
+            "step": 16070
+        },
+        {
+            "loss": 13.9399,
+            "grad_norm": 1.9101618528366089,
+            "learning_rate": 0.0005,
+            "epoch": 0.7204522154185737,
+            "step": 16075
+        },
+        {
+            "loss": 13.9123,
+            "grad_norm": 1.882951259613037,
+            "learning_rate": 0.0005,
+            "epoch": 0.7206763063098393,
+            "step": 16080
+        },
+        {
+            "loss": 13.8909,
+            "grad_norm": 1.6697813272476196,
+            "learning_rate": 0.0005,
+            "epoch": 0.7209003972011048,
+            "step": 16085
+        },
+        {
+            "loss": 13.9208,
+            "grad_norm": 1.7573597431182861,
+            "learning_rate": 0.0005,
+            "epoch": 0.7211244880923703,
+            "step": 16090
+        },
+        {
+            "loss": 13.9518,
+            "grad_norm": 1.8712494373321533,
+            "learning_rate": 0.0005,
+            "epoch": 0.7213485789836358,
+            "step": 16095
+        },
+        {
+            "loss": 14.035,
+            "grad_norm": 1.8593941926956177,
+            "learning_rate": 0.0005,
+            "epoch": 0.7215726698749012,
+            "step": 16100
+        },
+        {
+            "loss": 13.9139,
+            "grad_norm": 1.979432463645935,
+            "learning_rate": 0.0005,
+            "epoch": 0.7217967607661667,
+            "step": 16105
+        },
+        {
+            "loss": 13.9773,
+            "grad_norm": 1.9317654371261597,
+            "learning_rate": 0.0005,
+            "epoch": 0.7220208516574322,
+            "step": 16110
+        },
+        {
+            "loss": 13.9748,
+            "grad_norm": 1.9029157161712646,
+            "learning_rate": 0.0005,
+            "epoch": 0.7222449425486978,
+            "step": 16115
+        },
+        {
+            "loss": 13.975,
+            "grad_norm": 1.7725335359573364,
+            "learning_rate": 0.0005,
+            "epoch": 0.7224690334399633,
+            "step": 16120
+        },
+        {
+            "loss": 13.9674,
+            "grad_norm": 1.7740458250045776,
+            "learning_rate": 0.0005,
+            "epoch": 0.7226931243312288,
+            "step": 16125
+        },
+        {
+            "loss": 13.892,
+            "grad_norm": 1.7745329141616821,
+            "learning_rate": 0.0005,
+            "epoch": 0.7229172152224942,
+            "step": 16130
+        },
+        {
+            "loss": 13.9921,
+            "grad_norm": 1.7700062990188599,
+            "learning_rate": 0.0005,
+            "epoch": 0.7231413061137597,
+            "step": 16135
+        },
+        {
+            "loss": 14.0545,
+            "grad_norm": 1.8467191457748413,
+            "learning_rate": 0.0005,
+            "epoch": 0.7233653970050252,
+            "step": 16140
+        },
+        {
+            "loss": 13.9745,
+            "grad_norm": 1.7570888996124268,
+            "learning_rate": 0.0005,
+            "epoch": 0.7235894878962907,
+            "step": 16145
+        },
+        {
+            "loss": 13.9612,
+            "grad_norm": 1.7153857946395874,
+            "learning_rate": 0.0005,
+            "epoch": 0.7238135787875563,
+            "step": 16150
+        },
+        {
+            "loss": 13.9667,
+            "grad_norm": 1.918713092803955,
+            "learning_rate": 0.0005,
+            "epoch": 0.7240376696788218,
+            "step": 16155
+        },
+        {
+            "loss": 14.0746,
+            "grad_norm": 1.8288605213165283,
+            "learning_rate": 0.0005,
+            "epoch": 0.7242617605700872,
+            "step": 16160
+        },
+        {
+            "loss": 13.9114,
+            "grad_norm": 1.7235784530639648,
+            "learning_rate": 0.0005,
+            "epoch": 0.7244858514613527,
+            "step": 16165
+        },
+        {
+            "loss": 13.9968,
+            "grad_norm": 1.6746882200241089,
+            "learning_rate": 0.0005,
+            "epoch": 0.7247099423526182,
+            "step": 16170
+        },
+        {
+            "loss": 13.9439,
+            "grad_norm": 1.731046438217163,
+            "learning_rate": 0.0005,
+            "epoch": 0.7249340332438837,
+            "step": 16175
+        },
+        {
+            "loss": 13.9765,
+            "grad_norm": 1.830639123916626,
+            "learning_rate": 0.0005,
+            "epoch": 0.7251581241351492,
+            "step": 16180
+        },
+        {
+            "loss": 13.9256,
+            "grad_norm": 2.0223536491394043,
+            "learning_rate": 0.0005,
+            "epoch": 0.7253822150264148,
+            "step": 16185
+        },
+        {
+            "loss": 13.891,
+            "grad_norm": 1.7806121110916138,
+            "learning_rate": 0.0005,
+            "epoch": 0.7256063059176802,
+            "step": 16190
+        },
+        {
+            "loss": 14.0454,
+            "grad_norm": 1.6691431999206543,
+            "learning_rate": 0.0005,
+            "epoch": 0.7258303968089457,
+            "step": 16195
+        },
+        {
+            "loss": 13.8613,
+            "grad_norm": 1.7030301094055176,
+            "learning_rate": 0.0005,
+            "epoch": 0.7260544877002112,
+            "step": 16200
+        },
+        {
+            "loss": 13.869,
+            "grad_norm": 1.8289581537246704,
+            "learning_rate": 0.0005,
+            "epoch": 0.7262785785914767,
+            "step": 16205
+        },
+        {
+            "loss": 13.9307,
+            "grad_norm": 1.9004987478256226,
+            "learning_rate": 0.0005,
+            "epoch": 0.7265026694827422,
+            "step": 16210
+        },
+        {
+            "loss": 13.9059,
+            "grad_norm": 1.8068387508392334,
+            "learning_rate": 0.0005,
+            "epoch": 0.7267267603740077,
+            "step": 16215
+        },
+        {
+            "loss": 13.8845,
+            "grad_norm": 1.8290519714355469,
+            "learning_rate": 0.0005,
+            "epoch": 0.7269508512652731,
+            "step": 16220
+        },
+        {
+            "loss": 13.9848,
+            "grad_norm": 1.8770573139190674,
+            "learning_rate": 0.0005,
+            "epoch": 0.7271749421565387,
+            "step": 16225
+        },
+        {
+            "loss": 13.9106,
+            "grad_norm": 1.7803281545639038,
+            "learning_rate": 0.0005,
+            "epoch": 0.7273990330478042,
+            "step": 16230
+        },
+        {
+            "loss": 13.9031,
+            "grad_norm": 1.7252570390701294,
+            "learning_rate": 0.0005,
+            "epoch": 0.7276231239390697,
+            "step": 16235
+        },
+        {
+            "loss": 13.882,
+            "grad_norm": 1.8499126434326172,
+            "learning_rate": 0.0005,
+            "epoch": 0.7278472148303352,
+            "step": 16240
+        },
+        {
+            "loss": 13.9478,
+            "grad_norm": 1.8788185119628906,
+            "learning_rate": 0.0005,
+            "epoch": 0.7280713057216007,
+            "step": 16245
+        },
+        {
+            "loss": 13.9407,
+            "grad_norm": 1.9770047664642334,
+            "learning_rate": 0.0005,
+            "epoch": 0.7282953966128661,
+            "step": 16250
+        },
+        {
+            "loss": 13.9473,
+            "grad_norm": 1.9661725759506226,
+            "learning_rate": 0.0005,
+            "epoch": 0.7285194875041316,
+            "step": 16255
+        },
+        {
+            "loss": 13.9523,
+            "grad_norm": 1.7707750797271729,
+            "learning_rate": 0.0005,
+            "epoch": 0.7287435783953972,
+            "step": 16260
+        },
+        {
+            "loss": 13.9735,
+            "grad_norm": 1.7866833209991455,
+            "learning_rate": 0.0005,
+            "epoch": 0.7289676692866627,
+            "step": 16265
+        },
+        {
+            "loss": 13.9603,
+            "grad_norm": 1.8298813104629517,
+            "learning_rate": 0.0005,
+            "epoch": 0.7291917601779282,
+            "step": 16270
+        },
+        {
+            "loss": 13.9271,
+            "grad_norm": 1.8025341033935547,
+            "learning_rate": 0.0005,
+            "epoch": 0.7294158510691937,
+            "step": 16275
+        },
+        {
+            "loss": 13.9085,
+            "grad_norm": 1.7539923191070557,
+            "learning_rate": 0.0005,
+            "epoch": 0.7296399419604591,
+            "step": 16280
+        },
+        {
+            "loss": 13.9229,
+            "grad_norm": 1.9807707071304321,
+            "learning_rate": 0.0005,
+            "epoch": 0.7298640328517246,
+            "step": 16285
+        },
+        {
+            "loss": 13.9094,
+            "grad_norm": 1.8446540832519531,
+            "learning_rate": 0.0005,
+            "epoch": 0.7300881237429901,
+            "step": 16290
+        },
+        {
+            "loss": 13.9222,
+            "grad_norm": 1.7621265649795532,
+            "learning_rate": 0.0005,
+            "epoch": 0.7303122146342557,
+            "step": 16295
+        },
+        {
+            "loss": 13.9597,
+            "grad_norm": 1.9226858615875244,
+            "learning_rate": 0.0005,
+            "epoch": 0.7305363055255212,
+            "step": 16300
+        },
+        {
+            "loss": 13.8855,
+            "grad_norm": 1.8753238916397095,
+            "learning_rate": 0.0005,
+            "epoch": 0.7307603964167867,
+            "step": 16305
+        },
+        {
+            "loss": 13.9531,
+            "grad_norm": 1.7832648754119873,
+            "learning_rate": 0.0005,
+            "epoch": 0.7309844873080521,
+            "step": 16310
+        },
+        {
+            "loss": 13.9558,
+            "grad_norm": 1.9437285661697388,
+            "learning_rate": 0.0005,
+            "epoch": 0.7312085781993176,
+            "step": 16315
+        },
+        {
+            "loss": 13.9162,
+            "grad_norm": 1.9407806396484375,
+            "learning_rate": 0.0005,
+            "epoch": 0.7314326690905831,
+            "step": 16320
+        },
+        {
+            "loss": 13.9994,
+            "grad_norm": 1.8932803869247437,
+            "learning_rate": 0.0005,
+            "epoch": 0.7316567599818486,
+            "step": 16325
+        },
+        {
+            "loss": 14.0211,
+            "grad_norm": 1.7515097856521606,
+            "learning_rate": 0.0005,
+            "epoch": 0.7318808508731142,
+            "step": 16330
+        },
+        {
+            "loss": 13.9126,
+            "grad_norm": 1.8044368028640747,
+            "learning_rate": 0.0005,
+            "epoch": 0.7321049417643797,
+            "step": 16335
+        },
+        {
+            "loss": 14.0328,
+            "grad_norm": 1.7515829801559448,
+            "learning_rate": 0.0005,
+            "epoch": 0.7323290326556451,
+            "step": 16340
+        },
+        {
+            "loss": 13.9308,
+            "grad_norm": 1.720150351524353,
+            "learning_rate": 0.0005,
+            "epoch": 0.7325531235469106,
+            "step": 16345
+        },
+        {
+            "loss": 14.0479,
+            "grad_norm": 1.8786776065826416,
+            "learning_rate": 0.0005,
+            "epoch": 0.7327772144381761,
+            "step": 16350
+        },
+        {
+            "loss": 14.0188,
+            "grad_norm": 1.9431655406951904,
+            "learning_rate": 0.0005,
+            "epoch": 0.7330013053294416,
+            "step": 16355
+        },
+        {
+            "loss": 13.9009,
+            "grad_norm": 1.9714356660842896,
+            "learning_rate": 0.0005,
+            "epoch": 0.7332253962207071,
+            "step": 16360
+        },
+        {
+            "loss": 13.8668,
+            "grad_norm": 1.8676221370697021,
+            "learning_rate": 0.0005,
+            "epoch": 0.7334494871119727,
+            "step": 16365
+        },
+        {
+            "loss": 14.0674,
+            "grad_norm": 1.862158179283142,
+            "learning_rate": 0.0005,
+            "epoch": 0.7336735780032381,
+            "step": 16370
+        },
+        {
+            "loss": 13.8538,
+            "grad_norm": 1.813812255859375,
+            "learning_rate": 0.0005,
+            "epoch": 0.7338976688945036,
+            "step": 16375
+        },
+        {
+            "loss": 14.0629,
+            "grad_norm": 2.1444220542907715,
+            "learning_rate": 0.0005,
+            "epoch": 0.7341217597857691,
+            "step": 16380
+        },
+        {
+            "loss": 13.9987,
+            "grad_norm": 1.824510097503662,
+            "learning_rate": 0.0005,
+            "epoch": 0.7343458506770346,
+            "step": 16385
+        },
+        {
+            "loss": 13.9475,
+            "grad_norm": 1.9032855033874512,
+            "learning_rate": 0.0005,
+            "epoch": 0.7345699415683001,
+            "step": 16390
+        },
+        {
+            "loss": 13.8946,
+            "grad_norm": 1.7234930992126465,
+            "learning_rate": 0.0005,
+            "epoch": 0.7347940324595656,
+            "step": 16395
+        },
+        {
+            "loss": 13.9531,
+            "grad_norm": 1.8535950183868408,
+            "learning_rate": 0.0005,
+            "epoch": 0.735018123350831,
+            "step": 16400
+        },
+        {
+            "loss": 13.935,
+            "grad_norm": 1.7756191492080688,
+            "learning_rate": 0.0005,
+            "epoch": 0.7352422142420966,
+            "step": 16405
+        },
+        {
+            "loss": 14.0069,
+            "grad_norm": 1.6882721185684204,
+            "learning_rate": 0.0005,
+            "epoch": 0.7354663051333621,
+            "step": 16410
+        },
+        {
+            "loss": 13.9706,
+            "grad_norm": 1.888295292854309,
+            "learning_rate": 0.0005,
+            "epoch": 0.7356903960246276,
+            "step": 16415
+        },
+        {
+            "loss": 14.0556,
+            "grad_norm": 1.7977948188781738,
+            "learning_rate": 0.0005,
+            "epoch": 0.7359144869158931,
+            "step": 16420
+        },
+        {
+            "loss": 13.952,
+            "grad_norm": 1.8391417264938354,
+            "learning_rate": 0.0005,
+            "epoch": 0.7361385778071586,
+            "step": 16425
+        },
+        {
+            "loss": 13.911,
+            "grad_norm": 1.754473328590393,
+            "learning_rate": 0.0005,
+            "epoch": 0.736362668698424,
+            "step": 16430
+        },
+        {
+            "loss": 13.9549,
+            "grad_norm": 1.7115532159805298,
+            "learning_rate": 0.0005,
+            "epoch": 0.7365867595896896,
+            "step": 16435
+        },
+        {
+            "loss": 13.9322,
+            "grad_norm": 1.6821171045303345,
+            "learning_rate": 0.0005,
+            "epoch": 0.7368108504809551,
+            "step": 16440
+        },
+        {
+            "loss": 13.9352,
+            "grad_norm": 1.6488306522369385,
+            "learning_rate": 0.0005,
+            "epoch": 0.7370349413722206,
+            "step": 16445
+        },
+        {
+            "loss": 13.8723,
+            "grad_norm": 1.7560571432113647,
+            "learning_rate": 0.0005,
+            "epoch": 0.7372590322634861,
+            "step": 16450
+        },
+        {
+            "loss": 13.9244,
+            "grad_norm": 1.8154276609420776,
+            "learning_rate": 0.0005,
+            "epoch": 0.7374831231547516,
+            "step": 16455
+        },
+        {
+            "loss": 14.1038,
+            "grad_norm": 2.00632905960083,
+            "learning_rate": 0.0005,
+            "epoch": 0.737707214046017,
+            "step": 16460
+        },
+        {
+            "loss": 13.986,
+            "grad_norm": 1.9674841165542603,
+            "learning_rate": 0.0005,
+            "epoch": 0.7379313049372825,
+            "step": 16465
+        },
+        {
+            "loss": 13.9672,
+            "grad_norm": 1.891155481338501,
+            "learning_rate": 0.0005,
+            "epoch": 0.738155395828548,
+            "step": 16470
+        },
+        {
+            "loss": 14.0534,
+            "grad_norm": 1.940896987915039,
+            "learning_rate": 0.0005,
+            "epoch": 0.7383794867198136,
+            "step": 16475
+        },
+        {
+            "loss": 13.9107,
+            "grad_norm": 1.842924952507019,
+            "learning_rate": 0.0005,
+            "epoch": 0.7386035776110791,
+            "step": 16480
+        },
+        {
+            "loss": 13.9201,
+            "grad_norm": 1.7718983888626099,
+            "learning_rate": 0.0005,
+            "epoch": 0.7388276685023446,
+            "step": 16485
+        },
+        {
+            "loss": 13.8871,
+            "grad_norm": 1.7668657302856445,
+            "learning_rate": 0.0005,
+            "epoch": 0.73905175939361,
+            "step": 16490
+        },
+        {
+            "loss": 14.0596,
+            "grad_norm": 1.8025968074798584,
+            "learning_rate": 0.0005,
+            "epoch": 0.7392758502848755,
+            "step": 16495
+        },
+        {
+            "loss": 13.7597,
+            "grad_norm": 1.8243893384933472,
+            "learning_rate": 0.0005,
+            "epoch": 0.739499941176141,
+            "step": 16500
+        },
+        {
+            "eval_loss": 1.7385435104370117,
+            "eval_runtime": 18.6914,
+            "eval_samples_per_second": 876.555,
+            "eval_steps_per_second": 7.865,
+            "epoch": 0.739499941176141,
+            "step": 16500
+        },
+        {
+            "loss": 13.9621,
+            "grad_norm": 1.63633394241333,
+            "learning_rate": 0.0005,
+            "epoch": 0.7397240320674066,
+            "step": 16505
+        },
+        {
+            "loss": 13.7676,
+            "grad_norm": 1.9245598316192627,
+            "learning_rate": 0.0005,
+            "epoch": 0.7399481229586721,
+            "step": 16510
+        },
+        {
+            "loss": 13.9617,
+            "grad_norm": 1.7750827074050903,
+            "learning_rate": 0.0005,
+            "epoch": 0.7401722138499376,
+            "step": 16515
+        },
+        {
+            "loss": 13.9345,
+            "grad_norm": 1.7867636680603027,
+            "learning_rate": 0.0005,
+            "epoch": 0.740396304741203,
+            "step": 16520
+        },
+        {
+            "loss": 13.9694,
+            "grad_norm": 1.8457869291305542,
+            "learning_rate": 0.0005,
+            "epoch": 0.7406203956324685,
+            "step": 16525
+        },
+        {
+            "loss": 13.8827,
+            "grad_norm": 1.8203812837600708,
+            "learning_rate": 0.0005,
+            "epoch": 0.740844486523734,
+            "step": 16530
+        },
+        {
+            "loss": 13.9653,
+            "grad_norm": 1.8373428583145142,
+            "learning_rate": 0.0005,
+            "epoch": 0.7410685774149995,
+            "step": 16535
+        },
+        {
+            "loss": 13.8845,
+            "grad_norm": 1.7521898746490479,
+            "learning_rate": 0.0005,
+            "epoch": 0.7412926683062651,
+            "step": 16540
+        },
+        {
+            "loss": 13.97,
+            "grad_norm": 1.7927263975143433,
+            "learning_rate": 0.0005,
+            "epoch": 0.7415167591975306,
+            "step": 16545
+        },
+        {
+            "loss": 13.986,
+            "grad_norm": 2.0001494884490967,
+            "learning_rate": 0.0005,
+            "epoch": 0.741740850088796,
+            "step": 16550
+        },
+        {
+            "loss": 13.9242,
+            "grad_norm": 1.772683024406433,
+            "learning_rate": 0.0005,
+            "epoch": 0.7419649409800615,
+            "step": 16555
+        },
+        {
+            "loss": 13.9403,
+            "grad_norm": 1.6826038360595703,
+            "learning_rate": 0.0005,
+            "epoch": 0.742189031871327,
+            "step": 16560
+        },
+        {
+            "loss": 13.92,
+            "grad_norm": 1.8097511529922485,
+            "learning_rate": 0.0005,
+            "epoch": 0.7424131227625925,
+            "step": 16565
+        },
+        {
+            "loss": 13.9673,
+            "grad_norm": 1.8825819492340088,
+            "learning_rate": 0.0005,
+            "epoch": 0.742637213653858,
+            "step": 16570
+        },
+        {
+            "loss": 13.8282,
+            "grad_norm": 1.8934838771820068,
+            "learning_rate": 0.0005,
+            "epoch": 0.7428613045451236,
+            "step": 16575
+        },
+        {
+            "loss": 13.9764,
+            "grad_norm": 1.666270136833191,
+            "learning_rate": 0.0005,
+            "epoch": 0.743085395436389,
+            "step": 16580
+        },
+        {
+            "loss": 13.9107,
+            "grad_norm": 1.7280218601226807,
+            "learning_rate": 0.0005,
+            "epoch": 0.7433094863276545,
+            "step": 16585
+        },
+        {
+            "loss": 13.8924,
+            "grad_norm": 1.6558258533477783,
+            "learning_rate": 0.0005,
+            "epoch": 0.74353357721892,
+            "step": 16590
+        },
+        {
+            "loss": 13.8897,
+            "grad_norm": 1.7129403352737427,
+            "learning_rate": 0.0005,
+            "epoch": 0.7437576681101855,
+            "step": 16595
+        },
+        {
+            "loss": 13.8267,
+            "grad_norm": 1.8275370597839355,
+            "learning_rate": 0.0005,
+            "epoch": 0.743981759001451,
+            "step": 16600
+        },
+        {
+            "loss": 13.8312,
+            "grad_norm": 1.9228730201721191,
+            "learning_rate": 0.0005,
+            "epoch": 0.7442058498927164,
+            "step": 16605
+        },
+        {
+            "loss": 14.0149,
+            "grad_norm": 1.7469290494918823,
+            "learning_rate": 0.0005,
+            "epoch": 0.744429940783982,
+            "step": 16610
+        },
+        {
+            "loss": 13.8832,
+            "grad_norm": 1.7500646114349365,
+            "learning_rate": 0.0005,
+            "epoch": 0.7446540316752475,
+            "step": 16615
+        },
+        {
+            "loss": 13.9315,
+            "grad_norm": 1.7939316034317017,
+            "learning_rate": 0.0005,
+            "epoch": 0.744878122566513,
+            "step": 16620
+        },
+        {
+            "loss": 13.82,
+            "grad_norm": 1.9013673067092896,
+            "learning_rate": 0.0005,
+            "epoch": 0.7451022134577785,
+            "step": 16625
+        },
+        {
+            "loss": 13.8628,
+            "grad_norm": 1.7872364521026611,
+            "learning_rate": 0.0005,
+            "epoch": 0.745326304349044,
+            "step": 16630
+        },
+        {
+            "loss": 13.9992,
+            "grad_norm": 1.8590223789215088,
+            "learning_rate": 0.0005,
+            "epoch": 0.7455503952403094,
+            "step": 16635
+        },
+        {
+            "loss": 14.049,
+            "grad_norm": 1.8022938966751099,
+            "learning_rate": 0.0005,
+            "epoch": 0.7457744861315749,
+            "step": 16640
+        },
+        {
+            "loss": 13.9455,
+            "grad_norm": 1.7616809606552124,
+            "learning_rate": 0.0005,
+            "epoch": 0.7459985770228404,
+            "step": 16645
+        },
+        {
+            "loss": 13.8782,
+            "grad_norm": 1.9642635583877563,
+            "learning_rate": 0.0005,
+            "epoch": 0.746222667914106,
+            "step": 16650
+        },
+        {
+            "loss": 13.9376,
+            "grad_norm": 1.945994257926941,
+            "learning_rate": 0.0005,
+            "epoch": 0.7464467588053715,
+            "step": 16655
+        },
+        {
+            "loss": 13.9022,
+            "grad_norm": 1.802782416343689,
+            "learning_rate": 0.0005,
+            "epoch": 0.746670849696637,
+            "step": 16660
+        },
+        {
+            "loss": 14.0735,
+            "grad_norm": 1.7638579607009888,
+            "learning_rate": 0.0005,
+            "epoch": 0.7468949405879024,
+            "step": 16665
+        },
+        {
+            "loss": 13.8999,
+            "grad_norm": 1.7529107332229614,
+            "learning_rate": 0.0005,
+            "epoch": 0.7471190314791679,
+            "step": 16670
+        },
+        {
+            "loss": 13.9178,
+            "grad_norm": 1.919114112854004,
+            "learning_rate": 0.0005,
+            "epoch": 0.7473431223704334,
+            "step": 16675
+        },
+        {
+            "loss": 13.8848,
+            "grad_norm": 1.8599441051483154,
+            "learning_rate": 0.0005,
+            "epoch": 0.747567213261699,
+            "step": 16680
+        },
+        {
+            "loss": 13.9828,
+            "grad_norm": 1.8337881565093994,
+            "learning_rate": 0.0005,
+            "epoch": 0.7477913041529645,
+            "step": 16685
+        },
+        {
+            "loss": 13.9717,
+            "grad_norm": 1.919255018234253,
+            "learning_rate": 0.0005,
+            "epoch": 0.74801539504423,
+            "step": 16690
+        },
+        {
+            "loss": 13.8769,
+            "grad_norm": 1.7353612184524536,
+            "learning_rate": 0.0005,
+            "epoch": 0.7482394859354954,
+            "step": 16695
+        },
+        {
+            "loss": 13.9762,
+            "grad_norm": 1.7704685926437378,
+            "learning_rate": 0.0005,
+            "epoch": 0.7484635768267609,
+            "step": 16700
+        },
+        {
+            "loss": 13.9186,
+            "grad_norm": 1.7047014236450195,
+            "learning_rate": 0.0005,
+            "epoch": 0.7486876677180264,
+            "step": 16705
+        },
+        {
+            "loss": 13.8664,
+            "grad_norm": 1.841894268989563,
+            "learning_rate": 0.0005,
+            "epoch": 0.7489117586092919,
+            "step": 16710
+        },
+        {
+            "loss": 13.9868,
+            "grad_norm": 1.6897482872009277,
+            "learning_rate": 0.0005,
+            "epoch": 0.7491358495005574,
+            "step": 16715
+        },
+        {
+            "loss": 14.0523,
+            "grad_norm": 1.8365100622177124,
+            "learning_rate": 0.0005,
+            "epoch": 0.749359940391823,
+            "step": 16720
+        },
+        {
+            "loss": 13.9071,
+            "grad_norm": 1.8162260055541992,
+            "learning_rate": 0.0005,
+            "epoch": 0.7495840312830884,
+            "step": 16725
+        },
+        {
+            "loss": 13.9705,
+            "grad_norm": 1.6817643642425537,
+            "learning_rate": 0.0005,
+            "epoch": 0.7498081221743539,
+            "step": 16730
+        },
+        {
+            "loss": 13.9116,
+            "grad_norm": 1.930167317390442,
+            "learning_rate": 0.0005,
+            "epoch": 0.7500322130656194,
+            "step": 16735
+        },
+        {
+            "loss": 13.936,
+            "grad_norm": 1.8676118850708008,
+            "learning_rate": 0.0005,
+            "epoch": 0.7502563039568849,
+            "step": 16740
+        },
+        {
+            "loss": 13.9527,
+            "grad_norm": 1.9254913330078125,
+            "learning_rate": 0.0005,
+            "epoch": 0.7504803948481504,
+            "step": 16745
+        },
+        {
+            "loss": 14.0472,
+            "grad_norm": 1.6709167957305908,
+            "learning_rate": 0.0005,
+            "epoch": 0.750704485739416,
+            "step": 16750
+        },
+        {
+            "loss": 13.794,
+            "grad_norm": 1.8618227243423462,
+            "learning_rate": 0.0005,
+            "epoch": 0.7509285766306814,
+            "step": 16755
+        },
+        {
+            "loss": 13.8492,
+            "grad_norm": 1.8963994979858398,
+            "learning_rate": 0.0005,
+            "epoch": 0.7511526675219469,
+            "step": 16760
+        },
+        {
+            "loss": 13.9819,
+            "grad_norm": 1.883829951286316,
+            "learning_rate": 0.0005,
+            "epoch": 0.7513767584132124,
+            "step": 16765
+        },
+        {
+            "loss": 13.9072,
+            "grad_norm": 1.777325987815857,
+            "learning_rate": 0.0005,
+            "epoch": 0.7516008493044779,
+            "step": 16770
+        },
+        {
+            "loss": 13.9403,
+            "grad_norm": 1.7898032665252686,
+            "learning_rate": 0.0005,
+            "epoch": 0.7518249401957434,
+            "step": 16775
+        },
+        {
+            "loss": 13.9503,
+            "grad_norm": 1.8185434341430664,
+            "learning_rate": 0.0005,
+            "epoch": 0.7520490310870089,
+            "step": 16780
+        },
+        {
+            "loss": 13.9264,
+            "grad_norm": 1.8035706281661987,
+            "learning_rate": 0.0005,
+            "epoch": 0.7522731219782743,
+            "step": 16785
+        },
+        {
+            "loss": 14.0012,
+            "grad_norm": 1.9327406883239746,
+            "learning_rate": 0.0005,
+            "epoch": 0.7524972128695399,
+            "step": 16790
+        },
+        {
+            "loss": 13.9255,
+            "grad_norm": 1.8507956266403198,
+            "learning_rate": 0.0005,
+            "epoch": 0.7527213037608054,
+            "step": 16795
+        },
+        {
+            "loss": 13.7956,
+            "grad_norm": 1.8498870134353638,
+            "learning_rate": 0.0005,
+            "epoch": 0.7529453946520709,
+            "step": 16800
+        },
+        {
+            "loss": 13.9411,
+            "grad_norm": 1.7851393222808838,
+            "learning_rate": 0.0005,
+            "epoch": 0.7531694855433364,
+            "step": 16805
+        },
+        {
+            "loss": 13.9635,
+            "grad_norm": 1.8075121641159058,
+            "learning_rate": 0.0005,
+            "epoch": 0.7533935764346019,
+            "step": 16810
+        },
+        {
+            "loss": 13.882,
+            "grad_norm": 1.9553426504135132,
+            "learning_rate": 0.0005,
+            "epoch": 0.7536176673258673,
+            "step": 16815
+        },
+        {
+            "loss": 13.8641,
+            "grad_norm": 1.8933708667755127,
+            "learning_rate": 0.0005,
+            "epoch": 0.7538417582171328,
+            "step": 16820
+        },
+        {
+            "loss": 13.9657,
+            "grad_norm": 1.7268893718719482,
+            "learning_rate": 0.0005,
+            "epoch": 0.7540658491083984,
+            "step": 16825
+        },
+        {
+            "loss": 13.9142,
+            "grad_norm": 1.8427122831344604,
+            "learning_rate": 0.0005,
+            "epoch": 0.7542899399996639,
+            "step": 16830
+        },
+        {
+            "loss": 13.9643,
+            "grad_norm": 1.8097573518753052,
+            "learning_rate": 0.0005,
+            "epoch": 0.7545140308909294,
+            "step": 16835
+        },
+        {
+            "loss": 13.9884,
+            "grad_norm": 1.8019248247146606,
+            "learning_rate": 0.0005,
+            "epoch": 0.7547381217821949,
+            "step": 16840
+        },
+        {
+            "loss": 13.9075,
+            "grad_norm": 1.658731460571289,
+            "learning_rate": 0.0005,
+            "epoch": 0.7549622126734603,
+            "step": 16845
+        },
+        {
+            "loss": 13.9988,
+            "grad_norm": 1.731905460357666,
+            "learning_rate": 0.0005,
+            "epoch": 0.7551863035647258,
+            "step": 16850
+        },
+        {
+            "loss": 14.0389,
+            "grad_norm": 1.7171862125396729,
+            "learning_rate": 0.0005,
+            "epoch": 0.7554103944559913,
+            "step": 16855
+        },
+        {
+            "loss": 13.9988,
+            "grad_norm": 1.8597036600112915,
+            "learning_rate": 0.0005,
+            "epoch": 0.7556344853472569,
+            "step": 16860
+        },
+        {
+            "loss": 14.0282,
+            "grad_norm": 1.7419289350509644,
+            "learning_rate": 0.0005,
+            "epoch": 0.7558585762385224,
+            "step": 16865
+        },
+        {
+            "loss": 13.8449,
+            "grad_norm": 1.7694988250732422,
+            "learning_rate": 0.0005,
+            "epoch": 0.7560826671297879,
+            "step": 16870
+        },
+        {
+            "loss": 13.9721,
+            "grad_norm": 1.701286792755127,
+            "learning_rate": 0.0005,
+            "epoch": 0.7563067580210533,
+            "step": 16875
+        },
+        {
+            "loss": 13.9108,
+            "grad_norm": 2.0243606567382812,
+            "learning_rate": 0.0005,
+            "epoch": 0.7565308489123188,
+            "step": 16880
+        },
+        {
+            "loss": 13.9166,
+            "grad_norm": 1.786948800086975,
+            "learning_rate": 0.0005,
+            "epoch": 0.7567549398035843,
+            "step": 16885
+        },
+        {
+            "loss": 13.9138,
+            "grad_norm": 1.8174067735671997,
+            "learning_rate": 0.0005,
+            "epoch": 0.7569790306948498,
+            "step": 16890
+        },
+        {
+            "loss": 13.8914,
+            "grad_norm": 1.8659418821334839,
+            "learning_rate": 0.0005,
+            "epoch": 0.7572031215861154,
+            "step": 16895
+        },
+        {
+            "loss": 13.9287,
+            "grad_norm": 1.84630286693573,
+            "learning_rate": 0.0005,
+            "epoch": 0.7574272124773809,
+            "step": 16900
+        },
+        {
+            "loss": 13.9427,
+            "grad_norm": 1.8225772380828857,
+            "learning_rate": 0.0005,
+            "epoch": 0.7576513033686463,
+            "step": 16905
+        },
+        {
+            "loss": 13.9092,
+            "grad_norm": 1.7180335521697998,
+            "learning_rate": 0.0005,
+            "epoch": 0.7578753942599118,
+            "step": 16910
+        },
+        {
+            "loss": 14.0041,
+            "grad_norm": 1.6755090951919556,
+            "learning_rate": 0.0005,
+            "epoch": 0.7580994851511773,
+            "step": 16915
+        },
+        {
+            "loss": 13.8323,
+            "grad_norm": 1.6758276224136353,
+            "learning_rate": 0.0005,
+            "epoch": 0.7583235760424428,
+            "step": 16920
+        },
+        {
+            "loss": 13.9482,
+            "grad_norm": 1.7528473138809204,
+            "learning_rate": 0.0005,
+            "epoch": 0.7585476669337083,
+            "step": 16925
+        },
+        {
+            "loss": 13.9798,
+            "grad_norm": 1.7393977642059326,
+            "learning_rate": 0.0005,
+            "epoch": 0.7587717578249739,
+            "step": 16930
+        },
+        {
+            "loss": 13.8582,
+            "grad_norm": 1.8464726209640503,
+            "learning_rate": 0.0005,
+            "epoch": 0.7589958487162393,
+            "step": 16935
+        },
+        {
+            "loss": 14.0705,
+            "grad_norm": 1.881400465965271,
+            "learning_rate": 0.0005,
+            "epoch": 0.7592199396075048,
+            "step": 16940
+        },
+        {
+            "loss": 13.9858,
+            "grad_norm": 1.8284975290298462,
+            "learning_rate": 0.0005,
+            "epoch": 0.7594440304987703,
+            "step": 16945
+        },
+        {
+            "loss": 13.8379,
+            "grad_norm": 1.7476427555084229,
+            "learning_rate": 0.0005,
+            "epoch": 0.7596681213900358,
+            "step": 16950
+        },
+        {
+            "loss": 13.8957,
+            "grad_norm": 1.874758243560791,
+            "learning_rate": 0.0005,
+            "epoch": 0.7598922122813013,
+            "step": 16955
+        },
+        {
+            "loss": 13.9538,
+            "grad_norm": 2.0369491577148438,
+            "learning_rate": 0.0005,
+            "epoch": 0.7601163031725668,
+            "step": 16960
+        },
+        {
+            "loss": 13.9057,
+            "grad_norm": 1.8507211208343506,
+            "learning_rate": 0.0005,
+            "epoch": 0.7603403940638322,
+            "step": 16965
+        },
+        {
+            "loss": 13.9471,
+            "grad_norm": 2.078411340713501,
+            "learning_rate": 0.0005,
+            "epoch": 0.7605644849550978,
+            "step": 16970
+        },
+        {
+            "loss": 13.8973,
+            "grad_norm": 1.8963191509246826,
+            "learning_rate": 0.0005,
+            "epoch": 0.7607885758463633,
+            "step": 16975
+        },
+        {
+            "loss": 14.0286,
+            "grad_norm": 1.8882653713226318,
+            "learning_rate": 0.0005,
+            "epoch": 0.7610126667376288,
+            "step": 16980
+        },
+        {
+            "loss": 13.9606,
+            "grad_norm": 1.7255523204803467,
+            "learning_rate": 0.0005,
+            "epoch": 0.7612367576288943,
+            "step": 16985
+        },
+        {
+            "loss": 13.8637,
+            "grad_norm": 1.7835386991500854,
+            "learning_rate": 0.0005,
+            "epoch": 0.7614608485201598,
+            "step": 16990
+        },
+        {
+            "loss": 13.9112,
+            "grad_norm": 1.9392805099487305,
+            "learning_rate": 0.0005,
+            "epoch": 0.7616849394114252,
+            "step": 16995
+        },
+        {
+            "loss": 13.905,
+            "grad_norm": 2.0097835063934326,
+            "learning_rate": 0.0005,
+            "epoch": 0.7619090303026907,
+            "step": 17000
+        },
+        {
+            "eval_loss": 1.7384229898452759,
+            "eval_runtime": 18.5124,
+            "eval_samples_per_second": 885.026,
+            "eval_steps_per_second": 7.941,
+            "epoch": 0.7619090303026907,
+            "step": 17000
+        },
+        {
+            "loss": 13.8823,
+            "grad_norm": 2.398235321044922,
+            "learning_rate": 0.0005,
+            "epoch": 0.7621331211939563,
+            "step": 17005
+        },
+        {
+            "loss": 14.0402,
+            "grad_norm": 2.1302218437194824,
+            "learning_rate": 0.0005,
+            "epoch": 0.7623572120852218,
+            "step": 17010
+        },
+        {
+            "loss": 13.9419,
+            "grad_norm": 1.7162015438079834,
+            "learning_rate": 0.0005,
+            "epoch": 0.7625813029764873,
+            "step": 17015
+        },
+        {
+            "loss": 13.8423,
+            "grad_norm": 2.0047333240509033,
+            "learning_rate": 0.0005,
+            "epoch": 0.7628053938677528,
+            "step": 17020
+        },
+        {
+            "loss": 13.797,
+            "grad_norm": 1.8566488027572632,
+            "learning_rate": 0.0005,
+            "epoch": 0.7630294847590182,
+            "step": 17025
+        },
+        {
+            "loss": 13.8469,
+            "grad_norm": 1.8598730564117432,
+            "learning_rate": 0.0005,
+            "epoch": 0.7632535756502837,
+            "step": 17030
+        },
+        {
+            "loss": 13.9452,
+            "grad_norm": 1.9671058654785156,
+            "learning_rate": 0.0005,
+            "epoch": 0.7634776665415492,
+            "step": 17035
+        },
+        {
+            "loss": 13.9941,
+            "grad_norm": 1.9251089096069336,
+            "learning_rate": 0.0005,
+            "epoch": 0.7637017574328148,
+            "step": 17040
+        },
+        {
+            "loss": 13.914,
+            "grad_norm": 1.6617097854614258,
+            "learning_rate": 0.0005,
+            "epoch": 0.7639258483240803,
+            "step": 17045
+        },
+        {
+            "loss": 13.9591,
+            "grad_norm": 1.8101284503936768,
+            "learning_rate": 0.0005,
+            "epoch": 0.7641499392153458,
+            "step": 17050
+        },
+        {
+            "loss": 13.7848,
+            "grad_norm": 1.8096923828125,
+            "learning_rate": 0.0005,
+            "epoch": 0.7643740301066112,
+            "step": 17055
+        },
+        {
+            "loss": 13.9079,
+            "grad_norm": 1.896112084388733,
+            "learning_rate": 0.0005,
+            "epoch": 0.7645981209978767,
+            "step": 17060
+        },
+        {
+            "loss": 14.0023,
+            "grad_norm": 1.9171850681304932,
+            "learning_rate": 0.0005,
+            "epoch": 0.7648222118891422,
+            "step": 17065
+        },
+        {
+            "loss": 13.8576,
+            "grad_norm": 1.7272579669952393,
+            "learning_rate": 0.0005,
+            "epoch": 0.7650463027804077,
+            "step": 17070
+        },
+        {
+            "loss": 13.9216,
+            "grad_norm": 1.7884331941604614,
+            "learning_rate": 0.0005,
+            "epoch": 0.7652703936716733,
+            "step": 17075
+        },
+        {
+            "loss": 13.905,
+            "grad_norm": 1.927427887916565,
+            "learning_rate": 0.0005,
+            "epoch": 0.7654944845629388,
+            "step": 17080
+        },
+        {
+            "loss": 13.8944,
+            "grad_norm": 1.9965101480484009,
+            "learning_rate": 0.0005,
+            "epoch": 0.7657185754542042,
+            "step": 17085
+        },
+        {
+            "loss": 14.0026,
+            "grad_norm": 1.8458534479141235,
+            "learning_rate": 0.0005,
+            "epoch": 0.7659426663454697,
+            "step": 17090
+        },
+        {
+            "loss": 13.9527,
+            "grad_norm": 1.8137216567993164,
+            "learning_rate": 0.0005,
+            "epoch": 0.7661667572367352,
+            "step": 17095
+        },
+        {
+            "loss": 13.8867,
+            "grad_norm": 1.7597182989120483,
+            "learning_rate": 0.0005,
+            "epoch": 0.7663908481280007,
+            "step": 17100
+        },
+        {
+            "loss": 13.9618,
+            "grad_norm": 1.724302053451538,
+            "learning_rate": 0.0005,
+            "epoch": 0.7666149390192663,
+            "step": 17105
+        },
+        {
+            "loss": 13.8788,
+            "grad_norm": 1.7006725072860718,
+            "learning_rate": 0.0005,
+            "epoch": 0.7668390299105318,
+            "step": 17110
+        },
+        {
+            "loss": 13.9112,
+            "grad_norm": 1.6335428953170776,
+            "learning_rate": 0.0005,
+            "epoch": 0.7670631208017972,
+            "step": 17115
+        },
+        {
+            "loss": 13.9204,
+            "grad_norm": 1.7712724208831787,
+            "learning_rate": 0.0005,
+            "epoch": 0.7672872116930627,
+            "step": 17120
+        },
+        {
+            "loss": 13.9966,
+            "grad_norm": 1.7821861505508423,
+            "learning_rate": 0.0005,
+            "epoch": 0.7675113025843282,
+            "step": 17125
+        },
+        {
+            "loss": 13.8706,
+            "grad_norm": 1.9097272157669067,
+            "learning_rate": 0.0005,
+            "epoch": 0.7677353934755937,
+            "step": 17130
+        },
+        {
+            "loss": 13.8536,
+            "grad_norm": 2.0328292846679688,
+            "learning_rate": 0.0005,
+            "epoch": 0.7679594843668592,
+            "step": 17135
+        },
+        {
+            "loss": 13.9957,
+            "grad_norm": 2.003213882446289,
+            "learning_rate": 0.0005,
+            "epoch": 0.7681835752581246,
+            "step": 17140
+        },
+        {
+            "loss": 13.8826,
+            "grad_norm": 2.1363508701324463,
+            "learning_rate": 0.0005,
+            "epoch": 0.7684076661493902,
+            "step": 17145
+        },
+        {
+            "loss": 13.9027,
+            "grad_norm": 1.8297216892242432,
+            "learning_rate": 0.0005,
+            "epoch": 0.7686317570406557,
+            "step": 17150
+        },
+        {
+            "loss": 13.9279,
+            "grad_norm": 1.645095705986023,
+            "learning_rate": 0.0005,
+            "epoch": 0.7688558479319212,
+            "step": 17155
+        },
+        {
+            "loss": 13.8988,
+            "grad_norm": 1.7365893125534058,
+            "learning_rate": 0.0005,
+            "epoch": 0.7690799388231867,
+            "step": 17160
+        },
+        {
+            "loss": 14.1075,
+            "grad_norm": 1.8787424564361572,
+            "learning_rate": 0.0005,
+            "epoch": 0.7693040297144522,
+            "step": 17165
+        },
+        {
+            "loss": 13.8715,
+            "grad_norm": 1.850673794746399,
+            "learning_rate": 0.0005,
+            "epoch": 0.7695281206057176,
+            "step": 17170
+        },
+        {
+            "loss": 13.9795,
+            "grad_norm": 2.0193114280700684,
+            "learning_rate": 0.0005,
+            "epoch": 0.7697522114969831,
+            "step": 17175
+        },
+        {
+            "loss": 13.957,
+            "grad_norm": 1.9191477298736572,
+            "learning_rate": 0.0005,
+            "epoch": 0.7699763023882487,
+            "step": 17180
+        },
+        {
+            "loss": 13.9135,
+            "grad_norm": 1.8473769426345825,
+            "learning_rate": 0.0005,
+            "epoch": 0.7702003932795142,
+            "step": 17185
+        },
+        {
+            "loss": 13.9033,
+            "grad_norm": 1.8285760879516602,
+            "learning_rate": 0.0005,
+            "epoch": 0.7704244841707797,
+            "step": 17190
+        },
+        {
+            "loss": 13.8782,
+            "grad_norm": 1.740413784980774,
+            "learning_rate": 0.0005,
+            "epoch": 0.7706485750620452,
+            "step": 17195
+        },
+        {
+            "loss": 13.9807,
+            "grad_norm": 1.9219098091125488,
+            "learning_rate": 0.0005,
+            "epoch": 0.7708726659533106,
+            "step": 17200
+        },
+        {
+            "loss": 13.8142,
+            "grad_norm": 1.693150520324707,
+            "learning_rate": 0.0005,
+            "epoch": 0.7710967568445761,
+            "step": 17205
+        },
+        {
+            "loss": 13.9932,
+            "grad_norm": 1.8358036279678345,
+            "learning_rate": 0.0005,
+            "epoch": 0.7713208477358416,
+            "step": 17210
+        },
+        {
+            "loss": 14.0151,
+            "grad_norm": 2.059954881668091,
+            "learning_rate": 0.0005,
+            "epoch": 0.7715449386271072,
+            "step": 17215
+        },
+        {
+            "loss": 13.9524,
+            "grad_norm": 1.993303656578064,
+            "learning_rate": 0.0005,
+            "epoch": 0.7717690295183727,
+            "step": 17220
+        },
+        {
+            "loss": 13.928,
+            "grad_norm": 1.7009844779968262,
+            "learning_rate": 0.0005,
+            "epoch": 0.7719931204096382,
+            "step": 17225
+        },
+        {
+            "loss": 13.8545,
+            "grad_norm": 1.731141448020935,
+            "learning_rate": 0.0005,
+            "epoch": 0.7722172113009036,
+            "step": 17230
+        },
+        {
+            "loss": 13.9505,
+            "grad_norm": 1.7808451652526855,
+            "learning_rate": 0.0005,
+            "epoch": 0.7724413021921691,
+            "step": 17235
+        },
+        {
+            "loss": 13.8956,
+            "grad_norm": 1.6263585090637207,
+            "learning_rate": 0.0005,
+            "epoch": 0.7726653930834346,
+            "step": 17240
+        },
+        {
+            "loss": 14.0226,
+            "grad_norm": 1.7615787982940674,
+            "learning_rate": 0.0005,
+            "epoch": 0.7728894839747001,
+            "step": 17245
+        },
+        {
+            "loss": 13.9709,
+            "grad_norm": 1.766060471534729,
+            "learning_rate": 0.0005,
+            "epoch": 0.7731135748659657,
+            "step": 17250
+        },
+        {
+            "loss": 14.0055,
+            "grad_norm": 1.658046841621399,
+            "learning_rate": 0.0005,
+            "epoch": 0.7733376657572312,
+            "step": 17255
+        },
+        {
+            "loss": 13.8568,
+            "grad_norm": 1.9029816389083862,
+            "learning_rate": 0.0005,
+            "epoch": 0.7735617566484966,
+            "step": 17260
+        },
+        {
+            "loss": 13.8713,
+            "grad_norm": 1.806174397468567,
+            "learning_rate": 0.0005,
+            "epoch": 0.7737858475397621,
+            "step": 17265
+        },
+        {
+            "loss": 13.9703,
+            "grad_norm": 1.8879802227020264,
+            "learning_rate": 0.0005,
+            "epoch": 0.7740099384310276,
+            "step": 17270
+        },
+        {
+            "loss": 13.9304,
+            "grad_norm": 1.9935083389282227,
+            "learning_rate": 0.0005,
+            "epoch": 0.7742340293222931,
+            "step": 17275
+        },
+        {
+            "loss": 14.0471,
+            "grad_norm": 1.9430160522460938,
+            "learning_rate": 0.0005,
+            "epoch": 0.7744581202135586,
+            "step": 17280
+        },
+        {
+            "loss": 13.863,
+            "grad_norm": 1.7574583292007446,
+            "learning_rate": 0.0005,
+            "epoch": 0.7746822111048242,
+            "step": 17285
+        },
+        {
+            "loss": 13.8583,
+            "grad_norm": 1.865325927734375,
+            "learning_rate": 0.0005,
+            "epoch": 0.7749063019960896,
+            "step": 17290
+        },
+        {
+            "loss": 13.8901,
+            "grad_norm": 1.9842028617858887,
+            "learning_rate": 0.0005,
+            "epoch": 0.7751303928873551,
+            "step": 17295
+        },
+        {
+            "loss": 13.9087,
+            "grad_norm": 1.7361700534820557,
+            "learning_rate": 0.0005,
+            "epoch": 0.7753544837786206,
+            "step": 17300
+        },
+        {
+            "loss": 14.0314,
+            "grad_norm": 1.8161813020706177,
+            "learning_rate": 0.0005,
+            "epoch": 0.7755785746698861,
+            "step": 17305
+        },
+        {
+            "loss": 13.9428,
+            "grad_norm": 1.755927324295044,
+            "learning_rate": 0.0005,
+            "epoch": 0.7758026655611516,
+            "step": 17310
+        },
+        {
+            "loss": 13.8968,
+            "grad_norm": 1.747763991355896,
+            "learning_rate": 0.0005,
+            "epoch": 0.7760267564524171,
+            "step": 17315
+        },
+        {
+            "loss": 14.0306,
+            "grad_norm": 1.6892719268798828,
+            "learning_rate": 0.0005,
+            "epoch": 0.7762508473436825,
+            "step": 17320
+        },
+        {
+            "loss": 13.9098,
+            "grad_norm": 1.8000086545944214,
+            "learning_rate": 0.0005,
+            "epoch": 0.7764749382349481,
+            "step": 17325
+        },
+        {
+            "loss": 14.0461,
+            "grad_norm": 1.8460227251052856,
+            "learning_rate": 0.0005,
+            "epoch": 0.7766990291262136,
+            "step": 17330
+        },
+        {
+            "loss": 13.926,
+            "grad_norm": 1.7183973789215088,
+            "learning_rate": 0.0005,
+            "epoch": 0.7769231200174791,
+            "step": 17335
+        },
+        {
+            "loss": 13.9128,
+            "grad_norm": 1.7747199535369873,
+            "learning_rate": 0.0005,
+            "epoch": 0.7771472109087446,
+            "step": 17340
+        },
+        {
+            "loss": 13.9532,
+            "grad_norm": 1.8093260526657104,
+            "learning_rate": 0.0005,
+            "epoch": 0.7773713018000101,
+            "step": 17345
+        },
+        {
+            "loss": 13.8194,
+            "grad_norm": 1.9444884061813354,
+            "learning_rate": 0.0005,
+            "epoch": 0.7775953926912755,
+            "step": 17350
+        },
+        {
+            "loss": 13.908,
+            "grad_norm": 1.7580726146697998,
+            "learning_rate": 0.0005,
+            "epoch": 0.777819483582541,
+            "step": 17355
+        },
+        {
+            "loss": 13.9547,
+            "grad_norm": 1.7068579196929932,
+            "learning_rate": 0.0005,
+            "epoch": 0.7780435744738066,
+            "step": 17360
+        },
+        {
+            "loss": 13.8771,
+            "grad_norm": 1.8285844326019287,
+            "learning_rate": 0.0005,
+            "epoch": 0.7782676653650721,
+            "step": 17365
+        },
+        {
+            "loss": 13.9304,
+            "grad_norm": 1.7167176008224487,
+            "learning_rate": 0.0005,
+            "epoch": 0.7784917562563376,
+            "step": 17370
+        },
+        {
+            "loss": 13.9736,
+            "grad_norm": 1.820793867111206,
+            "learning_rate": 0.0005,
+            "epoch": 0.7787158471476031,
+            "step": 17375
+        },
+        {
+            "loss": 13.9138,
+            "grad_norm": 1.8576176166534424,
+            "learning_rate": 0.0005,
+            "epoch": 0.7789399380388685,
+            "step": 17380
+        },
+        {
+            "loss": 13.9551,
+            "grad_norm": 1.7626501321792603,
+            "learning_rate": 0.0005,
+            "epoch": 0.779164028930134,
+            "step": 17385
+        },
+        {
+            "loss": 13.9133,
+            "grad_norm": 1.788894534111023,
+            "learning_rate": 0.0005,
+            "epoch": 0.7793881198213995,
+            "step": 17390
+        },
+        {
+            "loss": 13.8893,
+            "grad_norm": 1.8483351469039917,
+            "learning_rate": 0.0005,
+            "epoch": 0.7796122107126651,
+            "step": 17395
+        },
+        {
+            "loss": 13.9881,
+            "grad_norm": 1.9322800636291504,
+            "learning_rate": 0.0005,
+            "epoch": 0.7798363016039306,
+            "step": 17400
+        },
+        {
+            "loss": 13.8728,
+            "grad_norm": 1.780640721321106,
+            "learning_rate": 0.0005,
+            "epoch": 0.7800603924951961,
+            "step": 17405
+        },
+        {
+            "loss": 13.857,
+            "grad_norm": 1.921074390411377,
+            "learning_rate": 0.0005,
+            "epoch": 0.7802844833864615,
+            "step": 17410
+        },
+        {
+            "loss": 13.9098,
+            "grad_norm": 1.8547496795654297,
+            "learning_rate": 0.0005,
+            "epoch": 0.780508574277727,
+            "step": 17415
+        },
+        {
+            "loss": 13.8675,
+            "grad_norm": 1.8414533138275146,
+            "learning_rate": 0.0005,
+            "epoch": 0.7807326651689925,
+            "step": 17420
+        },
+        {
+            "loss": 13.8619,
+            "grad_norm": 1.734214186668396,
+            "learning_rate": 0.0005,
+            "epoch": 0.780956756060258,
+            "step": 17425
+        },
+        {
+            "loss": 13.8024,
+            "grad_norm": 1.8890767097473145,
+            "learning_rate": 0.0005,
+            "epoch": 0.7811808469515236,
+            "step": 17430
+        },
+        {
+            "loss": 13.9844,
+            "grad_norm": 1.832506537437439,
+            "learning_rate": 0.0005,
+            "epoch": 0.7814049378427891,
+            "step": 17435
+        },
+        {
+            "loss": 13.9294,
+            "grad_norm": 1.803244709968567,
+            "learning_rate": 0.0005,
+            "epoch": 0.7816290287340545,
+            "step": 17440
+        },
+        {
+            "loss": 13.956,
+            "grad_norm": 1.704032063484192,
+            "learning_rate": 0.0005,
+            "epoch": 0.78185311962532,
+            "step": 17445
+        },
+        {
+            "loss": 13.9254,
+            "grad_norm": 1.7548668384552002,
+            "learning_rate": 0.0005,
+            "epoch": 0.7820772105165855,
+            "step": 17450
+        },
+        {
+            "loss": 14.0788,
+            "grad_norm": 1.7440739870071411,
+            "learning_rate": 0.0005,
+            "epoch": 0.782301301407851,
+            "step": 17455
+        },
+        {
+            "loss": 13.8141,
+            "grad_norm": 1.7532907724380493,
+            "learning_rate": 0.0005,
+            "epoch": 0.7825253922991166,
+            "step": 17460
+        },
+        {
+            "loss": 13.8557,
+            "grad_norm": 1.7621829509735107,
+            "learning_rate": 0.0005,
+            "epoch": 0.7827494831903821,
+            "step": 17465
+        },
+        {
+            "loss": 13.9685,
+            "grad_norm": 1.7075783014297485,
+            "learning_rate": 0.0005,
+            "epoch": 0.7829735740816475,
+            "step": 17470
+        },
+        {
+            "loss": 13.8901,
+            "grad_norm": 1.8711745738983154,
+            "learning_rate": 0.0005,
+            "epoch": 0.783197664972913,
+            "step": 17475
+        },
+        {
+            "loss": 13.9876,
+            "grad_norm": 2.0267229080200195,
+            "learning_rate": 0.0005,
+            "epoch": 0.7834217558641785,
+            "step": 17480
+        },
+        {
+            "loss": 13.9416,
+            "grad_norm": 1.8759691715240479,
+            "learning_rate": 0.0005,
+            "epoch": 0.783645846755444,
+            "step": 17485
+        },
+        {
+            "loss": 13.9271,
+            "grad_norm": 1.8488215208053589,
+            "learning_rate": 0.0005,
+            "epoch": 0.7838699376467095,
+            "step": 17490
+        },
+        {
+            "loss": 13.9616,
+            "grad_norm": 1.79073166847229,
+            "learning_rate": 0.0005,
+            "epoch": 0.784094028537975,
+            "step": 17495
+        },
+        {
+            "loss": 14.0167,
+            "grad_norm": 1.8655067682266235,
+            "learning_rate": 0.0005,
+            "epoch": 0.7843181194292405,
+            "step": 17500
+        },
+        {
+            "eval_loss": 1.742244005203247,
+            "eval_runtime": 18.6673,
+            "eval_samples_per_second": 877.685,
+            "eval_steps_per_second": 7.875,
+            "epoch": 0.7843181194292405,
+            "step": 17500
+        },
+        {
+            "loss": 13.8936,
+            "grad_norm": 1.7196091413497925,
+            "learning_rate": 0.0005,
+            "epoch": 0.784542210320506,
+            "step": 17505
+        },
+        {
+            "loss": 13.9612,
+            "grad_norm": 1.8762168884277344,
+            "learning_rate": 0.0005,
+            "epoch": 0.7847663012117715,
+            "step": 17510
+        },
+        {
+            "loss": 13.9842,
+            "grad_norm": 1.8332551717758179,
+            "learning_rate": 0.0005,
+            "epoch": 0.784990392103037,
+            "step": 17515
+        },
+        {
+            "loss": 14.0338,
+            "grad_norm": 1.7468173503875732,
+            "learning_rate": 0.0005,
+            "epoch": 0.7852144829943025,
+            "step": 17520
+        },
+        {
+            "loss": 13.8849,
+            "grad_norm": 1.7667865753173828,
+            "learning_rate": 0.0005,
+            "epoch": 0.785438573885568,
+            "step": 17525
+        },
+        {
+            "loss": 13.9859,
+            "grad_norm": 1.714019775390625,
+            "learning_rate": 0.0005,
+            "epoch": 0.7856626647768334,
+            "step": 17530
+        },
+        {
+            "loss": 14.0067,
+            "grad_norm": 1.8588460683822632,
+            "learning_rate": 0.0005,
+            "epoch": 0.785886755668099,
+            "step": 17535
+        },
+        {
+            "loss": 14.0311,
+            "grad_norm": 1.6515707969665527,
+            "learning_rate": 0.0005,
+            "epoch": 0.7861108465593645,
+            "step": 17540
+        },
+        {
+            "loss": 13.863,
+            "grad_norm": 1.7797318696975708,
+            "learning_rate": 0.0005,
+            "epoch": 0.78633493745063,
+            "step": 17545
+        },
+        {
+            "loss": 14.0213,
+            "grad_norm": 1.6977412700653076,
+            "learning_rate": 0.0005,
+            "epoch": 0.7865590283418955,
+            "step": 17550
+        },
+        {
+            "loss": 13.8579,
+            "grad_norm": 1.8260390758514404,
+            "learning_rate": 0.0005,
+            "epoch": 0.786783119233161,
+            "step": 17555
+        },
+        {
+            "loss": 13.9899,
+            "grad_norm": 1.8567028045654297,
+            "learning_rate": 0.0005,
+            "epoch": 0.7870072101244264,
+            "step": 17560
+        },
+        {
+            "loss": 14.0152,
+            "grad_norm": 1.9553159475326538,
+            "learning_rate": 0.0005,
+            "epoch": 0.7872313010156919,
+            "step": 17565
+        },
+        {
+            "loss": 14.0009,
+            "grad_norm": 1.808526873588562,
+            "learning_rate": 0.0005,
+            "epoch": 0.7874553919069575,
+            "step": 17570
+        },
+        {
+            "loss": 13.8205,
+            "grad_norm": 1.8745321035385132,
+            "learning_rate": 0.0005,
+            "epoch": 0.787679482798223,
+            "step": 17575
+        },
+        {
+            "loss": 13.8825,
+            "grad_norm": 1.6873557567596436,
+            "learning_rate": 0.0005,
+            "epoch": 0.7879035736894885,
+            "step": 17580
+        },
+        {
+            "loss": 14.0002,
+            "grad_norm": 1.663845181465149,
+            "learning_rate": 0.0005,
+            "epoch": 0.788127664580754,
+            "step": 17585
+        },
+        {
+            "loss": 13.9907,
+            "grad_norm": 1.7362158298492432,
+            "learning_rate": 0.0005,
+            "epoch": 0.7883517554720194,
+            "step": 17590
+        },
+        {
+            "loss": 13.9699,
+            "grad_norm": 1.9989452362060547,
+            "learning_rate": 0.0005,
+            "epoch": 0.7885758463632849,
+            "step": 17595
+        },
+        {
+            "loss": 13.9033,
+            "grad_norm": 1.7774451971054077,
+            "learning_rate": 0.0005,
+            "epoch": 0.7887999372545504,
+            "step": 17600
+        },
+        {
+            "loss": 13.9771,
+            "grad_norm": 1.8537983894348145,
+            "learning_rate": 0.0005,
+            "epoch": 0.789024028145816,
+            "step": 17605
+        },
+        {
+            "loss": 13.9926,
+            "grad_norm": 1.902446985244751,
+            "learning_rate": 0.0005,
+            "epoch": 0.7892481190370815,
+            "step": 17610
+        },
+        {
+            "loss": 14.0141,
+            "grad_norm": 1.8638439178466797,
+            "learning_rate": 0.0005,
+            "epoch": 0.789472209928347,
+            "step": 17615
+        },
+        {
+            "loss": 13.9426,
+            "grad_norm": 1.8101930618286133,
+            "learning_rate": 0.0005,
+            "epoch": 0.7896963008196124,
+            "step": 17620
+        },
+        {
+            "loss": 13.9717,
+            "grad_norm": 1.8008956909179688,
+            "learning_rate": 0.0005,
+            "epoch": 0.7899203917108779,
+            "step": 17625
+        },
+        {
+            "loss": 13.8962,
+            "grad_norm": 1.8648358583450317,
+            "learning_rate": 0.0005,
+            "epoch": 0.7901444826021434,
+            "step": 17630
+        },
+        {
+            "loss": 13.8996,
+            "grad_norm": 1.699042558670044,
+            "learning_rate": 0.0005,
+            "epoch": 0.7903685734934089,
+            "step": 17635
+        },
+        {
+            "loss": 13.8396,
+            "grad_norm": 1.7029451131820679,
+            "learning_rate": 0.0005,
+            "epoch": 0.7905926643846745,
+            "step": 17640
+        },
+        {
+            "loss": 13.9211,
+            "grad_norm": 1.6374444961547852,
+            "learning_rate": 0.0005,
+            "epoch": 0.79081675527594,
+            "step": 17645
+        },
+        {
+            "loss": 13.98,
+            "grad_norm": 1.8231736421585083,
+            "learning_rate": 0.0005,
+            "epoch": 0.7910408461672054,
+            "step": 17650
+        },
+        {
+            "loss": 13.8947,
+            "grad_norm": 1.7066539525985718,
+            "learning_rate": 0.0005,
+            "epoch": 0.7912649370584709,
+            "step": 17655
+        },
+        {
+            "loss": 13.9631,
+            "grad_norm": 1.9213404655456543,
+            "learning_rate": 0.0005,
+            "epoch": 0.7914890279497364,
+            "step": 17660
+        },
+        {
+            "loss": 13.8136,
+            "grad_norm": 1.8781342506408691,
+            "learning_rate": 0.0005,
+            "epoch": 0.7917131188410019,
+            "step": 17665
+        },
+        {
+            "loss": 13.9319,
+            "grad_norm": 1.76526939868927,
+            "learning_rate": 0.0005,
+            "epoch": 0.7919372097322674,
+            "step": 17670
+        },
+        {
+            "loss": 14.0505,
+            "grad_norm": 1.7963687181472778,
+            "learning_rate": 0.0005,
+            "epoch": 0.792161300623533,
+            "step": 17675
+        },
+        {
+            "loss": 13.9135,
+            "grad_norm": 1.9999289512634277,
+            "learning_rate": 0.0005,
+            "epoch": 0.7923853915147984,
+            "step": 17680
+        },
+        {
+            "loss": 13.9473,
+            "grad_norm": 1.8966845273971558,
+            "learning_rate": 0.0005,
+            "epoch": 0.7926094824060639,
+            "step": 17685
+        },
+        {
+            "loss": 13.9459,
+            "grad_norm": 1.9211297035217285,
+            "learning_rate": 0.0005,
+            "epoch": 0.7928335732973294,
+            "step": 17690
+        },
+        {
+            "loss": 13.8553,
+            "grad_norm": 1.8830465078353882,
+            "learning_rate": 0.0005,
+            "epoch": 0.7930576641885949,
+            "step": 17695
+        },
+        {
+            "loss": 13.9025,
+            "grad_norm": 1.9547890424728394,
+            "learning_rate": 0.0005,
+            "epoch": 0.7932817550798604,
+            "step": 17700
+        },
+        {
+            "loss": 14.1072,
+            "grad_norm": 1.8501747846603394,
+            "learning_rate": 0.0005,
+            "epoch": 0.7935058459711258,
+            "step": 17705
+        },
+        {
+            "loss": 13.7782,
+            "grad_norm": 1.9531677961349487,
+            "learning_rate": 0.0005,
+            "epoch": 0.7937299368623913,
+            "step": 17710
+        },
+        {
+            "loss": 13.9874,
+            "grad_norm": 2.038963794708252,
+            "learning_rate": 0.0005,
+            "epoch": 0.7939540277536569,
+            "step": 17715
+        },
+        {
+            "loss": 13.9117,
+            "grad_norm": 1.886931300163269,
+            "learning_rate": 0.0005,
+            "epoch": 0.7941781186449224,
+            "step": 17720
+        },
+        {
+            "loss": 13.9347,
+            "grad_norm": 1.8401106595993042,
+            "learning_rate": 0.0005,
+            "epoch": 0.7944022095361879,
+            "step": 17725
+        },
+        {
+            "loss": 13.9397,
+            "grad_norm": 1.7104369401931763,
+            "learning_rate": 0.0005,
+            "epoch": 0.7946263004274534,
+            "step": 17730
+        },
+        {
+            "loss": 13.8901,
+            "grad_norm": 1.6022361516952515,
+            "learning_rate": 0.0005,
+            "epoch": 0.7948503913187188,
+            "step": 17735
+        },
+        {
+            "loss": 13.9179,
+            "grad_norm": 1.8352298736572266,
+            "learning_rate": 0.0005,
+            "epoch": 0.7950744822099843,
+            "step": 17740
+        },
+        {
+            "loss": 13.973,
+            "grad_norm": 1.8135077953338623,
+            "learning_rate": 0.0005,
+            "epoch": 0.7952985731012499,
+            "step": 17745
+        },
+        {
+            "loss": 14.0032,
+            "grad_norm": 1.7439593076705933,
+            "learning_rate": 0.0005,
+            "epoch": 0.7955226639925154,
+            "step": 17750
+        },
+        {
+            "loss": 13.9162,
+            "grad_norm": 1.909914255142212,
+            "learning_rate": 0.0005,
+            "epoch": 0.7957467548837809,
+            "step": 17755
+        },
+        {
+            "loss": 13.9659,
+            "grad_norm": 1.9050699472427368,
+            "learning_rate": 0.0005,
+            "epoch": 0.7959708457750464,
+            "step": 17760
+        },
+        {
+            "loss": 14.0256,
+            "grad_norm": 1.806579828262329,
+            "learning_rate": 0.0005,
+            "epoch": 0.7961949366663118,
+            "step": 17765
+        },
+        {
+            "loss": 13.9308,
+            "grad_norm": 1.8770440816879272,
+            "learning_rate": 0.0005,
+            "epoch": 0.7964190275575773,
+            "step": 17770
+        },
+        {
+            "loss": 13.8989,
+            "grad_norm": 2.0780200958251953,
+            "learning_rate": 0.0005,
+            "epoch": 0.7966431184488428,
+            "step": 17775
+        },
+        {
+            "loss": 13.9108,
+            "grad_norm": 1.651223063468933,
+            "learning_rate": 0.0005,
+            "epoch": 0.7968672093401084,
+            "step": 17780
+        },
+        {
+            "loss": 13.8166,
+            "grad_norm": 1.7695311307907104,
+            "learning_rate": 0.0005,
+            "epoch": 0.7970913002313739,
+            "step": 17785
+        },
+        {
+            "loss": 13.8646,
+            "grad_norm": 1.7559232711791992,
+            "learning_rate": 0.0005,
+            "epoch": 0.7973153911226394,
+            "step": 17790
+        },
+        {
+            "loss": 13.8857,
+            "grad_norm": 1.7012014389038086,
+            "learning_rate": 0.0005,
+            "epoch": 0.7975394820139048,
+            "step": 17795
+        },
+        {
+            "loss": 13.9441,
+            "grad_norm": 1.9856010675430298,
+            "learning_rate": 0.0005,
+            "epoch": 0.7977635729051703,
+            "step": 17800
+        },
+        {
+            "loss": 13.9702,
+            "grad_norm": 1.7729270458221436,
+            "learning_rate": 0.0005,
+            "epoch": 0.7979876637964358,
+            "step": 17805
+        },
+        {
+            "loss": 13.9417,
+            "grad_norm": 1.7350739240646362,
+            "learning_rate": 0.0005,
+            "epoch": 0.7982117546877013,
+            "step": 17810
+        },
+        {
+            "loss": 13.8052,
+            "grad_norm": 1.6648602485656738,
+            "learning_rate": 0.0005,
+            "epoch": 0.7984358455789669,
+            "step": 17815
+        },
+        {
+            "loss": 14.0659,
+            "grad_norm": 1.7132405042648315,
+            "learning_rate": 0.0005,
+            "epoch": 0.7986599364702324,
+            "step": 17820
+        },
+        {
+            "loss": 13.947,
+            "grad_norm": 1.819995403289795,
+            "learning_rate": 0.0005,
+            "epoch": 0.7988840273614978,
+            "step": 17825
+        },
+        {
+            "loss": 13.8886,
+            "grad_norm": 1.8971381187438965,
+            "learning_rate": 0.0005,
+            "epoch": 0.7991081182527633,
+            "step": 17830
+        },
+        {
+            "loss": 13.8627,
+            "grad_norm": 1.7079675197601318,
+            "learning_rate": 0.0005,
+            "epoch": 0.7993322091440288,
+            "step": 17835
+        },
+        {
+            "loss": 13.9065,
+            "grad_norm": 1.9513356685638428,
+            "learning_rate": 0.0005,
+            "epoch": 0.7995563000352943,
+            "step": 17840
+        },
+        {
+            "loss": 13.9798,
+            "grad_norm": 2.0933167934417725,
+            "learning_rate": 0.0005,
+            "epoch": 0.7997803909265598,
+            "step": 17845
+        },
+        {
+            "loss": 13.9302,
+            "grad_norm": 2.068735122680664,
+            "learning_rate": 0.0005,
+            "epoch": 0.8000044818178254,
+            "step": 17850
+        },
+        {
+            "loss": 13.9432,
+            "grad_norm": 1.9295698404312134,
+            "learning_rate": 0.0005,
+            "epoch": 0.8002285727090908,
+            "step": 17855
+        },
+        {
+            "loss": 13.9133,
+            "grad_norm": 2.0122389793395996,
+            "learning_rate": 0.0005,
+            "epoch": 0.8004526636003563,
+            "step": 17860
+        },
+        {
+            "loss": 13.8614,
+            "grad_norm": 1.9103975296020508,
+            "learning_rate": 0.0005,
+            "epoch": 0.8006767544916218,
+            "step": 17865
+        },
+        {
+            "loss": 13.8453,
+            "grad_norm": 1.851839542388916,
+            "learning_rate": 0.0005,
+            "epoch": 0.8009008453828873,
+            "step": 17870
+        },
+        {
+            "loss": 13.8546,
+            "grad_norm": 1.8850983381271362,
+            "learning_rate": 0.0005,
+            "epoch": 0.8011249362741528,
+            "step": 17875
+        },
+        {
+            "loss": 13.9192,
+            "grad_norm": 1.7939026355743408,
+            "learning_rate": 0.0005,
+            "epoch": 0.8013490271654183,
+            "step": 17880
+        },
+        {
+            "loss": 14.0107,
+            "grad_norm": 1.7058433294296265,
+            "learning_rate": 0.0005,
+            "epoch": 0.8015731180566837,
+            "step": 17885
+        },
+        {
+            "loss": 13.9066,
+            "grad_norm": 1.8725535869598389,
+            "learning_rate": 0.0005,
+            "epoch": 0.8017972089479493,
+            "step": 17890
+        },
+        {
+            "loss": 13.8958,
+            "grad_norm": 1.797574520111084,
+            "learning_rate": 0.0005,
+            "epoch": 0.8020212998392148,
+            "step": 17895
+        },
+        {
+            "loss": 13.8328,
+            "grad_norm": 1.7859054803848267,
+            "learning_rate": 0.0005,
+            "epoch": 0.8022453907304803,
+            "step": 17900
+        },
+        {
+            "loss": 13.8831,
+            "grad_norm": 1.7225736379623413,
+            "learning_rate": 0.0005,
+            "epoch": 0.8024694816217458,
+            "step": 17905
+        },
+        {
+            "loss": 13.9417,
+            "grad_norm": 1.9291692972183228,
+            "learning_rate": 0.0005,
+            "epoch": 0.8026935725130113,
+            "step": 17910
+        },
+        {
+            "loss": 13.8104,
+            "grad_norm": 1.838352918624878,
+            "learning_rate": 0.0005,
+            "epoch": 0.8029176634042767,
+            "step": 17915
+        },
+        {
+            "loss": 13.9082,
+            "grad_norm": 1.8938319683074951,
+            "learning_rate": 0.0005,
+            "epoch": 0.8031417542955422,
+            "step": 17920
+        },
+        {
+            "loss": 13.95,
+            "grad_norm": 1.8627904653549194,
+            "learning_rate": 0.0005,
+            "epoch": 0.8033658451868078,
+            "step": 17925
+        },
+        {
+            "loss": 13.9393,
+            "grad_norm": 1.7981688976287842,
+            "learning_rate": 0.0005,
+            "epoch": 0.8035899360780733,
+            "step": 17930
+        },
+        {
+            "loss": 13.8658,
+            "grad_norm": 1.8402200937271118,
+            "learning_rate": 0.0005,
+            "epoch": 0.8038140269693388,
+            "step": 17935
+        },
+        {
+            "loss": 13.9451,
+            "grad_norm": 1.9201061725616455,
+            "learning_rate": 0.0005,
+            "epoch": 0.8040381178606043,
+            "step": 17940
+        },
+        {
+            "loss": 13.9314,
+            "grad_norm": 1.8516732454299927,
+            "learning_rate": 0.0005,
+            "epoch": 0.8042622087518697,
+            "step": 17945
+        },
+        {
+            "loss": 13.8518,
+            "grad_norm": 1.7574687004089355,
+            "learning_rate": 0.0005,
+            "epoch": 0.8044862996431352,
+            "step": 17950
+        },
+        {
+            "loss": 13.8738,
+            "grad_norm": 1.7064701318740845,
+            "learning_rate": 0.0005,
+            "epoch": 0.8047103905344007,
+            "step": 17955
+        },
+        {
+            "loss": 13.9399,
+            "grad_norm": 1.7522039413452148,
+            "learning_rate": 0.0005,
+            "epoch": 0.8049344814256663,
+            "step": 17960
+        },
+        {
+            "loss": 14.0378,
+            "grad_norm": 1.72758150100708,
+            "learning_rate": 0.0005,
+            "epoch": 0.8051585723169318,
+            "step": 17965
+        },
+        {
+            "loss": 13.9735,
+            "grad_norm": 1.7682212591171265,
+            "learning_rate": 0.0005,
+            "epoch": 0.8053826632081973,
+            "step": 17970
+        },
+        {
+            "loss": 13.9341,
+            "grad_norm": 1.6668306589126587,
+            "learning_rate": 0.0005,
+            "epoch": 0.8056067540994627,
+            "step": 17975
+        },
+        {
+            "loss": 13.892,
+            "grad_norm": 1.7351326942443848,
+            "learning_rate": 0.0005,
+            "epoch": 0.8058308449907282,
+            "step": 17980
+        },
+        {
+            "loss": 14.0166,
+            "grad_norm": 1.6612002849578857,
+            "learning_rate": 0.0005,
+            "epoch": 0.8060549358819937,
+            "step": 17985
+        },
+        {
+            "loss": 13.9435,
+            "grad_norm": 1.7721000909805298,
+            "learning_rate": 0.0005,
+            "epoch": 0.8062790267732592,
+            "step": 17990
+        },
+        {
+            "loss": 14.0486,
+            "grad_norm": 1.7399013042449951,
+            "learning_rate": 0.0005,
+            "epoch": 0.8065031176645248,
+            "step": 17995
+        },
+        {
+            "loss": 13.9363,
+            "grad_norm": 1.867493987083435,
+            "learning_rate": 0.0005,
+            "epoch": 0.8067272085557903,
+            "step": 18000
+        },
+        {
+            "eval_loss": 1.7346055507659912,
+            "eval_runtime": 18.7863,
+            "eval_samples_per_second": 872.126,
+            "eval_steps_per_second": 7.825,
+            "epoch": 0.8067272085557903,
+            "step": 18000
+        },
+        {
+            "loss": 13.9715,
+            "grad_norm": 1.8982458114624023,
+            "learning_rate": 0.0005,
+            "epoch": 0.8069512994470557,
+            "step": 18005
+        },
+        {
+            "loss": 13.9355,
+            "grad_norm": 2.0637786388397217,
+            "learning_rate": 0.0005,
+            "epoch": 0.8071753903383212,
+            "step": 18010
+        },
+        {
+            "loss": 13.9808,
+            "grad_norm": 1.6983004808425903,
+            "learning_rate": 0.0005,
+            "epoch": 0.8073994812295867,
+            "step": 18015
+        },
+        {
+            "loss": 13.8864,
+            "grad_norm": 1.8731592893600464,
+            "learning_rate": 0.0005,
+            "epoch": 0.8076235721208522,
+            "step": 18020
+        },
+        {
+            "loss": 13.831,
+            "grad_norm": 1.7289857864379883,
+            "learning_rate": 0.0005,
+            "epoch": 0.8078476630121177,
+            "step": 18025
+        },
+        {
+            "loss": 13.9294,
+            "grad_norm": 1.6939040422439575,
+            "learning_rate": 0.0005,
+            "epoch": 0.8080717539033833,
+            "step": 18030
+        },
+        {
+            "loss": 13.9463,
+            "grad_norm": 1.7209194898605347,
+            "learning_rate": 0.0005,
+            "epoch": 0.8082958447946487,
+            "step": 18035
+        },
+        {
+            "loss": 14.0243,
+            "grad_norm": 1.854943037033081,
+            "learning_rate": 0.0005,
+            "epoch": 0.8085199356859142,
+            "step": 18040
+        },
+        {
+            "loss": 13.9918,
+            "grad_norm": 1.75515878200531,
+            "learning_rate": 0.0005,
+            "epoch": 0.8087440265771797,
+            "step": 18045
+        },
+        {
+            "loss": 13.9171,
+            "grad_norm": 1.871372103691101,
+            "learning_rate": 0.0005,
+            "epoch": 0.8089681174684452,
+            "step": 18050
+        },
+        {
+            "loss": 13.9423,
+            "grad_norm": 1.8944830894470215,
+            "learning_rate": 0.0005,
+            "epoch": 0.8091922083597107,
+            "step": 18055
+        },
+        {
+            "loss": 13.8909,
+            "grad_norm": 1.8145884275436401,
+            "learning_rate": 0.0005,
+            "epoch": 0.8094162992509762,
+            "step": 18060
+        },
+        {
+            "loss": 13.9074,
+            "grad_norm": 2.0930395126342773,
+            "learning_rate": 0.0005,
+            "epoch": 0.8096403901422417,
+            "step": 18065
+        },
+        {
+            "loss": 13.8805,
+            "grad_norm": 1.8583229780197144,
+            "learning_rate": 0.0005,
+            "epoch": 0.8098644810335072,
+            "step": 18070
+        },
+        {
+            "loss": 13.9727,
+            "grad_norm": 1.7828443050384521,
+            "learning_rate": 0.0005,
+            "epoch": 0.8100885719247727,
+            "step": 18075
+        },
+        {
+            "loss": 13.9119,
+            "grad_norm": 1.7656959295272827,
+            "learning_rate": 0.0005,
+            "epoch": 0.8103126628160382,
+            "step": 18080
+        },
+        {
+            "loss": 13.8061,
+            "grad_norm": 1.8191685676574707,
+            "learning_rate": 0.0005,
+            "epoch": 0.8105367537073037,
+            "step": 18085
+        },
+        {
+            "loss": 13.9294,
+            "grad_norm": 1.8345978260040283,
+            "learning_rate": 0.0005,
+            "epoch": 0.8107608445985692,
+            "step": 18090
+        },
+        {
+            "loss": 13.9271,
+            "grad_norm": 1.7740345001220703,
+            "learning_rate": 0.0005,
+            "epoch": 0.8109849354898346,
+            "step": 18095
+        },
+        {
+            "loss": 13.9762,
+            "grad_norm": 1.8106119632720947,
+            "learning_rate": 0.0005,
+            "epoch": 0.8112090263811002,
+            "step": 18100
+        },
+        {
+            "loss": 13.9898,
+            "grad_norm": 1.719824194908142,
+            "learning_rate": 0.0005,
+            "epoch": 0.8114331172723657,
+            "step": 18105
+        },
+        {
+            "loss": 13.9451,
+            "grad_norm": 1.7734304666519165,
+            "learning_rate": 0.0005,
+            "epoch": 0.8116572081636312,
+            "step": 18110
+        },
+        {
+            "loss": 13.9099,
+            "grad_norm": 1.7129086256027222,
+            "learning_rate": 0.0005,
+            "epoch": 0.8118812990548967,
+            "step": 18115
+        },
+        {
+            "loss": 13.8419,
+            "grad_norm": 1.9217442274093628,
+            "learning_rate": 0.0005,
+            "epoch": 0.8121053899461622,
+            "step": 18120
+        },
+        {
+            "loss": 13.8746,
+            "grad_norm": 2.0497589111328125,
+            "learning_rate": 0.0005,
+            "epoch": 0.8123294808374276,
+            "step": 18125
+        },
+        {
+            "loss": 13.8953,
+            "grad_norm": 2.071010112762451,
+            "learning_rate": 0.0005,
+            "epoch": 0.8125535717286931,
+            "step": 18130
+        },
+        {
+            "loss": 13.9024,
+            "grad_norm": 1.7323050498962402,
+            "learning_rate": 0.0005,
+            "epoch": 0.8127776626199587,
+            "step": 18135
+        },
+        {
+            "loss": 13.7554,
+            "grad_norm": 1.8137826919555664,
+            "learning_rate": 0.0005,
+            "epoch": 0.8130017535112242,
+            "step": 18140
+        },
+        {
+            "loss": 13.9382,
+            "grad_norm": 1.8626395463943481,
+            "learning_rate": 0.0005,
+            "epoch": 0.8132258444024897,
+            "step": 18145
+        },
+        {
+            "loss": 13.9762,
+            "grad_norm": 1.7347149848937988,
+            "learning_rate": 0.0005,
+            "epoch": 0.8134499352937552,
+            "step": 18150
+        },
+        {
+            "loss": 13.9163,
+            "grad_norm": 1.817475438117981,
+            "learning_rate": 0.0005,
+            "epoch": 0.8136740261850206,
+            "step": 18155
+        },
+        {
+            "loss": 13.8897,
+            "grad_norm": 1.7329870462417603,
+            "learning_rate": 0.0005,
+            "epoch": 0.8138981170762861,
+            "step": 18160
+        },
+        {
+            "loss": 13.9202,
+            "grad_norm": 1.714704155921936,
+            "learning_rate": 0.0005,
+            "epoch": 0.8141222079675516,
+            "step": 18165
+        },
+        {
+            "loss": 13.974,
+            "grad_norm": 1.8908694982528687,
+            "learning_rate": 0.0005,
+            "epoch": 0.8143462988588172,
+            "step": 18170
+        },
+        {
+            "loss": 14.0251,
+            "grad_norm": 1.8410272598266602,
+            "learning_rate": 0.0005,
+            "epoch": 0.8145703897500827,
+            "step": 18175
+        },
+        {
+            "loss": 13.8982,
+            "grad_norm": 1.8369451761245728,
+            "learning_rate": 0.0005,
+            "epoch": 0.8147944806413482,
+            "step": 18180
+        },
+        {
+            "loss": 13.865,
+            "grad_norm": 1.8976902961730957,
+            "learning_rate": 0.0005,
+            "epoch": 0.8150185715326136,
+            "step": 18185
+        },
+        {
+            "loss": 13.9005,
+            "grad_norm": 1.7548612356185913,
+            "learning_rate": 0.0005,
+            "epoch": 0.8152426624238791,
+            "step": 18190
+        },
+        {
+            "loss": 13.8124,
+            "grad_norm": 1.7536848783493042,
+            "learning_rate": 0.0005,
+            "epoch": 0.8154667533151446,
+            "step": 18195
+        },
+        {
+            "loss": 13.9581,
+            "grad_norm": 1.722154140472412,
+            "learning_rate": 0.0005,
+            "epoch": 0.8156908442064101,
+            "step": 18200
+        },
+        {
+            "loss": 13.8407,
+            "grad_norm": 1.8374119997024536,
+            "learning_rate": 0.0005,
+            "epoch": 0.8159149350976757,
+            "step": 18205
+        },
+        {
+            "loss": 13.8782,
+            "grad_norm": 1.992284893989563,
+            "learning_rate": 0.0005,
+            "epoch": 0.8161390259889412,
+            "step": 18210
+        },
+        {
+            "loss": 13.8929,
+            "grad_norm": 1.832295298576355,
+            "learning_rate": 0.0005,
+            "epoch": 0.8163631168802066,
+            "step": 18215
+        },
+        {
+            "loss": 13.9147,
+            "grad_norm": 1.9362643957138062,
+            "learning_rate": 0.0005,
+            "epoch": 0.8165872077714721,
+            "step": 18220
+        },
+        {
+            "loss": 13.9553,
+            "grad_norm": 1.7259230613708496,
+            "learning_rate": 0.0005,
+            "epoch": 0.8168112986627376,
+            "step": 18225
+        },
+        {
+            "loss": 13.9219,
+            "grad_norm": 1.7937911748886108,
+            "learning_rate": 0.0005,
+            "epoch": 0.8170353895540031,
+            "step": 18230
+        },
+        {
+            "loss": 13.9057,
+            "grad_norm": 1.9405204057693481,
+            "learning_rate": 0.0005,
+            "epoch": 0.8172594804452686,
+            "step": 18235
+        },
+        {
+            "loss": 13.7556,
+            "grad_norm": 1.8398072719573975,
+            "learning_rate": 0.0005,
+            "epoch": 0.817483571336534,
+            "step": 18240
+        },
+        {
+            "loss": 13.9854,
+            "grad_norm": 1.8530958890914917,
+            "learning_rate": 0.0005,
+            "epoch": 0.8177076622277996,
+            "step": 18245
+        },
+        {
+            "loss": 14.0051,
+            "grad_norm": 1.9418649673461914,
+            "learning_rate": 0.0005,
+            "epoch": 0.8179317531190651,
+            "step": 18250
+        },
+        {
+            "loss": 13.8789,
+            "grad_norm": 1.7378404140472412,
+            "learning_rate": 0.0005,
+            "epoch": 0.8181558440103306,
+            "step": 18255
+        },
+        {
+            "loss": 13.952,
+            "grad_norm": 1.8054472208023071,
+            "learning_rate": 0.0005,
+            "epoch": 0.8183799349015961,
+            "step": 18260
+        },
+        {
+            "loss": 13.8416,
+            "grad_norm": 1.6757985353469849,
+            "learning_rate": 0.0005,
+            "epoch": 0.8186040257928616,
+            "step": 18265
+        },
+        {
+            "loss": 13.8685,
+            "grad_norm": 1.7990907430648804,
+            "learning_rate": 0.0005,
+            "epoch": 0.818828116684127,
+            "step": 18270
+        },
+        {
+            "loss": 13.9584,
+            "grad_norm": 1.8647428750991821,
+            "learning_rate": 0.0005,
+            "epoch": 0.8190522075753925,
+            "step": 18275
+        },
+        {
+            "loss": 13.8945,
+            "grad_norm": 1.6959542036056519,
+            "learning_rate": 0.0005,
+            "epoch": 0.8192762984666581,
+            "step": 18280
+        },
+        {
+            "loss": 13.9194,
+            "grad_norm": 1.8789013624191284,
+            "learning_rate": 0.0005,
+            "epoch": 0.8195003893579236,
+            "step": 18285
+        },
+        {
+            "loss": 13.9844,
+            "grad_norm": 1.7351081371307373,
+            "learning_rate": 0.0005,
+            "epoch": 0.8197244802491891,
+            "step": 18290
+        },
+        {
+            "loss": 13.8088,
+            "grad_norm": 1.7942126989364624,
+            "learning_rate": 0.0005,
+            "epoch": 0.8199485711404546,
+            "step": 18295
+        },
+        {
+            "loss": 13.8635,
+            "grad_norm": 1.7644922733306885,
+            "learning_rate": 0.0005,
+            "epoch": 0.82017266203172,
+            "step": 18300
+        },
+        {
+            "loss": 13.9958,
+            "grad_norm": 1.9505988359451294,
+            "learning_rate": 0.0005,
+            "epoch": 0.8203967529229855,
+            "step": 18305
+        },
+        {
+            "loss": 13.8548,
+            "grad_norm": 1.931020975112915,
+            "learning_rate": 0.0005,
+            "epoch": 0.820620843814251,
+            "step": 18310
+        },
+        {
+            "loss": 13.9201,
+            "grad_norm": 1.907620906829834,
+            "learning_rate": 0.0005,
+            "epoch": 0.8208449347055166,
+            "step": 18315
+        },
+        {
+            "loss": 13.9423,
+            "grad_norm": 1.9632011651992798,
+            "learning_rate": 0.0005,
+            "epoch": 0.8210690255967821,
+            "step": 18320
+        },
+        {
+            "loss": 13.9251,
+            "grad_norm": 1.6815967559814453,
+            "learning_rate": 0.0005,
+            "epoch": 0.8212931164880476,
+            "step": 18325
+        },
+        {
+            "loss": 13.8988,
+            "grad_norm": 1.7095072269439697,
+            "learning_rate": 0.0005,
+            "epoch": 0.821517207379313,
+            "step": 18330
+        },
+        {
+            "loss": 13.9442,
+            "grad_norm": 1.7073733806610107,
+            "learning_rate": 0.0005,
+            "epoch": 0.8217412982705785,
+            "step": 18335
+        },
+        {
+            "loss": 13.9667,
+            "grad_norm": 1.9493305683135986,
+            "learning_rate": 0.0005,
+            "epoch": 0.821965389161844,
+            "step": 18340
+        },
+        {
+            "loss": 13.8833,
+            "grad_norm": 1.689464807510376,
+            "learning_rate": 0.0005,
+            "epoch": 0.8221894800531095,
+            "step": 18345
+        },
+        {
+            "loss": 13.8829,
+            "grad_norm": 1.9471354484558105,
+            "learning_rate": 0.0005,
+            "epoch": 0.8224135709443751,
+            "step": 18350
+        },
+        {
+            "loss": 13.8999,
+            "grad_norm": 1.8226585388183594,
+            "learning_rate": 0.0005,
+            "epoch": 0.8226376618356406,
+            "step": 18355
+        },
+        {
+            "loss": 13.9831,
+            "grad_norm": 1.8304983377456665,
+            "learning_rate": 0.0005,
+            "epoch": 0.822861752726906,
+            "step": 18360
+        },
+        {
+            "loss": 14.0138,
+            "grad_norm": 1.7459100484848022,
+            "learning_rate": 0.0005,
+            "epoch": 0.8230858436181715,
+            "step": 18365
+        },
+        {
+            "loss": 13.9309,
+            "grad_norm": 1.6721384525299072,
+            "learning_rate": 0.0005,
+            "epoch": 0.823309934509437,
+            "step": 18370
+        },
+        {
+            "loss": 13.9495,
+            "grad_norm": 1.7700474262237549,
+            "learning_rate": 0.0005,
+            "epoch": 0.8235340254007025,
+            "step": 18375
+        },
+        {
+            "loss": 13.9657,
+            "grad_norm": 1.7054543495178223,
+            "learning_rate": 0.0005,
+            "epoch": 0.823758116291968,
+            "step": 18380
+        },
+        {
+            "loss": 14.0339,
+            "grad_norm": 1.7842026948928833,
+            "learning_rate": 0.0005,
+            "epoch": 0.8239822071832336,
+            "step": 18385
+        },
+        {
+            "loss": 13.9015,
+            "grad_norm": 1.9943057298660278,
+            "learning_rate": 0.0005,
+            "epoch": 0.824206298074499,
+            "step": 18390
+        },
+        {
+            "loss": 13.9177,
+            "grad_norm": 2.0690407752990723,
+            "learning_rate": 0.0005,
+            "epoch": 0.8244303889657645,
+            "step": 18395
+        },
+        {
+            "loss": 13.9146,
+            "grad_norm": 1.7490520477294922,
+            "learning_rate": 0.0005,
+            "epoch": 0.82465447985703,
+            "step": 18400
+        },
+        {
+            "loss": 13.8496,
+            "grad_norm": 1.9926691055297852,
+            "learning_rate": 0.0005,
+            "epoch": 0.8248785707482955,
+            "step": 18405
+        },
+        {
+            "loss": 13.9622,
+            "grad_norm": 1.9529262781143188,
+            "learning_rate": 0.0005,
+            "epoch": 0.825102661639561,
+            "step": 18410
+        },
+        {
+            "loss": 13.901,
+            "grad_norm": 1.7856429815292358,
+            "learning_rate": 0.0005,
+            "epoch": 0.8253267525308265,
+            "step": 18415
+        },
+        {
+            "loss": 13.8567,
+            "grad_norm": 1.7407591342926025,
+            "learning_rate": 0.0005,
+            "epoch": 0.825550843422092,
+            "step": 18420
+        },
+        {
+            "loss": 13.8574,
+            "grad_norm": 1.813948631286621,
+            "learning_rate": 0.0005,
+            "epoch": 0.8257749343133575,
+            "step": 18425
+        },
+        {
+            "loss": 13.9195,
+            "grad_norm": 1.7237603664398193,
+            "learning_rate": 0.0005,
+            "epoch": 0.825999025204623,
+            "step": 18430
+        },
+        {
+            "loss": 13.8594,
+            "grad_norm": 1.8184058666229248,
+            "learning_rate": 0.0005,
+            "epoch": 0.8262231160958885,
+            "step": 18435
+        },
+        {
+            "loss": 13.9363,
+            "grad_norm": 1.6922754049301147,
+            "learning_rate": 0.0005,
+            "epoch": 0.826447206987154,
+            "step": 18440
+        },
+        {
+            "loss": 13.9343,
+            "grad_norm": 1.7327067852020264,
+            "learning_rate": 0.0005,
+            "epoch": 0.8266712978784195,
+            "step": 18445
+        },
+        {
+            "loss": 13.8157,
+            "grad_norm": 1.8410130739212036,
+            "learning_rate": 0.0005,
+            "epoch": 0.8268953887696849,
+            "step": 18450
+        },
+        {
+            "loss": 13.8803,
+            "grad_norm": 1.9221383333206177,
+            "learning_rate": 0.0005,
+            "epoch": 0.8271194796609505,
+            "step": 18455
+        },
+        {
+            "loss": 13.9386,
+            "grad_norm": 1.9344953298568726,
+            "learning_rate": 0.0005,
+            "epoch": 0.827343570552216,
+            "step": 18460
+        },
+        {
+            "loss": 13.9093,
+            "grad_norm": 2.010633707046509,
+            "learning_rate": 0.0005,
+            "epoch": 0.8275676614434815,
+            "step": 18465
+        },
+        {
+            "loss": 13.865,
+            "grad_norm": 1.995612621307373,
+            "learning_rate": 0.0005,
+            "epoch": 0.827791752334747,
+            "step": 18470
+        },
+        {
+            "loss": 13.8581,
+            "grad_norm": 1.7131673097610474,
+            "learning_rate": 0.0005,
+            "epoch": 0.8280158432260125,
+            "step": 18475
+        },
+        {
+            "loss": 13.9404,
+            "grad_norm": 1.925093173980713,
+            "learning_rate": 0.0005,
+            "epoch": 0.8282399341172779,
+            "step": 18480
+        },
+        {
+            "loss": 13.927,
+            "grad_norm": 1.9550200700759888,
+            "learning_rate": 0.0005,
+            "epoch": 0.8284640250085434,
+            "step": 18485
+        },
+        {
+            "loss": 13.9704,
+            "grad_norm": 1.8336318731307983,
+            "learning_rate": 0.0005,
+            "epoch": 0.828688115899809,
+            "step": 18490
+        },
+        {
+            "loss": 13.8936,
+            "grad_norm": 1.737156867980957,
+            "learning_rate": 0.0005,
+            "epoch": 0.8289122067910745,
+            "step": 18495
+        },
+        {
+            "loss": 13.8499,
+            "grad_norm": 1.6031452417373657,
+            "learning_rate": 0.0005,
+            "epoch": 0.82913629768234,
+            "step": 18500
+        },
+        {
+            "eval_loss": 1.7327139377593994,
+            "eval_runtime": 18.9496,
+            "eval_samples_per_second": 864.608,
+            "eval_steps_per_second": 7.757,
+            "epoch": 0.82913629768234,
+            "step": 18500
+        },
+        {
+            "loss": 13.9124,
+            "grad_norm": 1.787706732749939,
+            "learning_rate": 0.0005,
+            "epoch": 0.8293603885736055,
+            "step": 18505
+        },
+        {
+            "loss": 13.8967,
+            "grad_norm": 1.7352012395858765,
+            "learning_rate": 0.0005,
+            "epoch": 0.8295844794648709,
+            "step": 18510
+        },
+        {
+            "loss": 13.9343,
+            "grad_norm": 1.8264018297195435,
+            "learning_rate": 0.0005,
+            "epoch": 0.8298085703561364,
+            "step": 18515
+        },
+        {
+            "loss": 13.9116,
+            "grad_norm": 1.9214221239089966,
+            "learning_rate": 0.0005,
+            "epoch": 0.8300326612474019,
+            "step": 18520
+        },
+        {
+            "loss": 13.8691,
+            "grad_norm": 1.6791073083877563,
+            "learning_rate": 0.0005,
+            "epoch": 0.8302567521386675,
+            "step": 18525
+        },
+        {
+            "loss": 13.9978,
+            "grad_norm": 1.7033040523529053,
+            "learning_rate": 0.0005,
+            "epoch": 0.830480843029933,
+            "step": 18530
+        },
+        {
+            "loss": 13.9676,
+            "grad_norm": 1.7302173376083374,
+            "learning_rate": 0.0005,
+            "epoch": 0.8307049339211985,
+            "step": 18535
+        },
+        {
+            "loss": 13.8587,
+            "grad_norm": 1.6527957916259766,
+            "learning_rate": 0.0005,
+            "epoch": 0.8309290248124639,
+            "step": 18540
+        },
+        {
+            "loss": 14.0066,
+            "grad_norm": 1.7439963817596436,
+            "learning_rate": 0.0005,
+            "epoch": 0.8311531157037294,
+            "step": 18545
+        },
+        {
+            "loss": 13.8333,
+            "grad_norm": 1.8020864725112915,
+            "learning_rate": 0.0005,
+            "epoch": 0.8313772065949949,
+            "step": 18550
+        },
+        {
+            "loss": 13.9397,
+            "grad_norm": 1.6987053155899048,
+            "learning_rate": 0.0005,
+            "epoch": 0.8316012974862604,
+            "step": 18555
+        },
+        {
+            "loss": 13.9715,
+            "grad_norm": 1.8397094011306763,
+            "learning_rate": 0.0005,
+            "epoch": 0.831825388377526,
+            "step": 18560
+        },
+        {
+            "loss": 13.8542,
+            "grad_norm": 1.8940035104751587,
+            "learning_rate": 0.0005,
+            "epoch": 0.8320494792687915,
+            "step": 18565
+        },
+        {
+            "loss": 13.7612,
+            "grad_norm": 1.703979730606079,
+            "learning_rate": 0.0005,
+            "epoch": 0.8322735701600569,
+            "step": 18570
+        },
+        {
+            "loss": 13.8202,
+            "grad_norm": 1.646026372909546,
+            "learning_rate": 0.0005,
+            "epoch": 0.8324976610513224,
+            "step": 18575
+        },
+        {
+            "loss": 13.9224,
+            "grad_norm": 1.7458852529525757,
+            "learning_rate": 0.0005,
+            "epoch": 0.8327217519425879,
+            "step": 18580
+        },
+        {
+            "loss": 13.9089,
+            "grad_norm": 1.803983449935913,
+            "learning_rate": 0.0005,
+            "epoch": 0.8329458428338534,
+            "step": 18585
+        },
+        {
+            "loss": 13.8724,
+            "grad_norm": 1.8974820375442505,
+            "learning_rate": 0.0005,
+            "epoch": 0.8331699337251189,
+            "step": 18590
+        },
+        {
+            "loss": 13.8951,
+            "grad_norm": 1.6343988180160522,
+            "learning_rate": 0.0005,
+            "epoch": 0.8333940246163845,
+            "step": 18595
+        },
+        {
+            "loss": 13.9771,
+            "grad_norm": 1.8503625392913818,
+            "learning_rate": 0.0005,
+            "epoch": 0.8336181155076499,
+            "step": 18600
+        },
+        {
+            "loss": 13.9672,
+            "grad_norm": 1.7192838191986084,
+            "learning_rate": 0.0005,
+            "epoch": 0.8338422063989154,
+            "step": 18605
+        },
+        {
+            "loss": 13.8696,
+            "grad_norm": 1.7600250244140625,
+            "learning_rate": 0.0005,
+            "epoch": 0.8340662972901809,
+            "step": 18610
+        },
+        {
+            "loss": 13.811,
+            "grad_norm": 1.8360655307769775,
+            "learning_rate": 0.0005,
+            "epoch": 0.8342903881814464,
+            "step": 18615
+        },
+        {
+            "loss": 13.8613,
+            "grad_norm": 1.7966500520706177,
+            "learning_rate": 0.0005,
+            "epoch": 0.8345144790727119,
+            "step": 18620
+        },
+        {
+            "loss": 13.9652,
+            "grad_norm": 1.9188029766082764,
+            "learning_rate": 0.0005,
+            "epoch": 0.8347385699639774,
+            "step": 18625
+        },
+        {
+            "loss": 13.7856,
+            "grad_norm": 1.8910226821899414,
+            "learning_rate": 0.0005,
+            "epoch": 0.8349626608552428,
+            "step": 18630
+        },
+        {
+            "loss": 13.8772,
+            "grad_norm": 1.678817629814148,
+            "learning_rate": 0.0005,
+            "epoch": 0.8351867517465084,
+            "step": 18635
+        },
+        {
+            "loss": 13.9314,
+            "grad_norm": 1.654451608657837,
+            "learning_rate": 0.0005,
+            "epoch": 0.8354108426377739,
+            "step": 18640
+        },
+        {
+            "loss": 13.9197,
+            "grad_norm": 1.702844262123108,
+            "learning_rate": 0.0005,
+            "epoch": 0.8356349335290394,
+            "step": 18645
+        },
+        {
+            "loss": 13.8353,
+            "grad_norm": 1.6565039157867432,
+            "learning_rate": 0.0005,
+            "epoch": 0.8358590244203049,
+            "step": 18650
+        },
+        {
+            "loss": 13.8804,
+            "grad_norm": 1.90591561794281,
+            "learning_rate": 0.0005,
+            "epoch": 0.8360831153115704,
+            "step": 18655
+        },
+        {
+            "loss": 13.9052,
+            "grad_norm": 2.0706241130828857,
+            "learning_rate": 0.0005,
+            "epoch": 0.8363072062028358,
+            "step": 18660
+        },
+        {
+            "loss": 13.9134,
+            "grad_norm": 1.9571928977966309,
+            "learning_rate": 0.0005,
+            "epoch": 0.8365312970941013,
+            "step": 18665
+        },
+        {
+            "loss": 13.9123,
+            "grad_norm": 2.055427074432373,
+            "learning_rate": 0.0005,
+            "epoch": 0.8367553879853669,
+            "step": 18670
+        },
+        {
+            "loss": 13.9184,
+            "grad_norm": 1.6851314306259155,
+            "learning_rate": 0.0005,
+            "epoch": 0.8369794788766324,
+            "step": 18675
+        },
+        {
+            "loss": 13.8479,
+            "grad_norm": 1.6859577894210815,
+            "learning_rate": 0.0005,
+            "epoch": 0.8372035697678979,
+            "step": 18680
+        },
+        {
+            "loss": 13.8261,
+            "grad_norm": 1.812279224395752,
+            "learning_rate": 0.0005,
+            "epoch": 0.8374276606591634,
+            "step": 18685
+        },
+        {
+            "loss": 13.9274,
+            "grad_norm": 1.7130305767059326,
+            "learning_rate": 0.0005,
+            "epoch": 0.8376517515504288,
+            "step": 18690
+        },
+        {
+            "loss": 13.8953,
+            "grad_norm": 1.7281595468521118,
+            "learning_rate": 0.0005,
+            "epoch": 0.8378758424416943,
+            "step": 18695
+        },
+        {
+            "loss": 13.8708,
+            "grad_norm": 1.7609429359436035,
+            "learning_rate": 0.0005,
+            "epoch": 0.8380999333329598,
+            "step": 18700
+        },
+        {
+            "loss": 13.9123,
+            "grad_norm": 1.6971237659454346,
+            "learning_rate": 0.0005,
+            "epoch": 0.8383240242242254,
+            "step": 18705
+        },
+        {
+            "loss": 13.8957,
+            "grad_norm": 1.8378429412841797,
+            "learning_rate": 0.0005,
+            "epoch": 0.8385481151154909,
+            "step": 18710
+        },
+        {
+            "loss": 13.8924,
+            "grad_norm": 1.6270524263381958,
+            "learning_rate": 0.0005,
+            "epoch": 0.8387722060067564,
+            "step": 18715
+        },
+        {
+            "loss": 13.9143,
+            "grad_norm": 1.8351507186889648,
+            "learning_rate": 0.0005,
+            "epoch": 0.8389962968980218,
+            "step": 18720
+        },
+        {
+            "loss": 13.9358,
+            "grad_norm": 1.8356674909591675,
+            "learning_rate": 0.0005,
+            "epoch": 0.8392203877892873,
+            "step": 18725
+        },
+        {
+            "loss": 13.8947,
+            "grad_norm": 1.7781685590744019,
+            "learning_rate": 0.0005,
+            "epoch": 0.8394444786805528,
+            "step": 18730
+        },
+        {
+            "loss": 13.9504,
+            "grad_norm": 1.6525814533233643,
+            "learning_rate": 0.0005,
+            "epoch": 0.8396685695718183,
+            "step": 18735
+        },
+        {
+            "loss": 13.9959,
+            "grad_norm": 1.7032777070999146,
+            "learning_rate": 0.0005,
+            "epoch": 0.8398926604630839,
+            "step": 18740
+        },
+        {
+            "loss": 13.9913,
+            "grad_norm": 1.6849335432052612,
+            "learning_rate": 0.0005,
+            "epoch": 0.8401167513543494,
+            "step": 18745
+        },
+        {
+            "loss": 13.9061,
+            "grad_norm": 1.7574923038482666,
+            "learning_rate": 0.0005,
+            "epoch": 0.8403408422456148,
+            "step": 18750
+        },
+        {
+            "loss": 13.9459,
+            "grad_norm": 1.8186347484588623,
+            "learning_rate": 0.0005,
+            "epoch": 0.8405649331368803,
+            "step": 18755
+        },
+        {
+            "loss": 13.8599,
+            "grad_norm": 1.8368334770202637,
+            "learning_rate": 0.0005,
+            "epoch": 0.8407890240281458,
+            "step": 18760
+        },
+        {
+            "loss": 13.9558,
+            "grad_norm": 1.780190348625183,
+            "learning_rate": 0.0005,
+            "epoch": 0.8410131149194113,
+            "step": 18765
+        },
+        {
+            "loss": 13.8318,
+            "grad_norm": 1.6295452117919922,
+            "learning_rate": 0.0005,
+            "epoch": 0.8412372058106768,
+            "step": 18770
+        },
+        {
+            "loss": 13.9218,
+            "grad_norm": 1.6828653812408447,
+            "learning_rate": 0.0005,
+            "epoch": 0.8414612967019424,
+            "step": 18775
+        },
+        {
+            "loss": 13.9613,
+            "grad_norm": 1.9260400533676147,
+            "learning_rate": 0.0005,
+            "epoch": 0.8416853875932078,
+            "step": 18780
+        },
+        {
+            "loss": 13.943,
+            "grad_norm": 1.6809277534484863,
+            "learning_rate": 0.0005,
+            "epoch": 0.8419094784844733,
+            "step": 18785
+        },
+        {
+            "loss": 13.984,
+            "grad_norm": 1.7654415369033813,
+            "learning_rate": 0.0005,
+            "epoch": 0.8421335693757388,
+            "step": 18790
+        },
+        {
+            "loss": 13.8248,
+            "grad_norm": 1.678613305091858,
+            "learning_rate": 0.0005,
+            "epoch": 0.8423576602670043,
+            "step": 18795
+        },
+        {
+            "loss": 13.9117,
+            "grad_norm": 1.7469117641448975,
+            "learning_rate": 0.0005,
+            "epoch": 0.8425817511582698,
+            "step": 18800
+        },
+        {
+            "loss": 13.9719,
+            "grad_norm": 1.945115566253662,
+            "learning_rate": 0.0005,
+            "epoch": 0.8428058420495352,
+            "step": 18805
+        },
+        {
+            "loss": 13.8406,
+            "grad_norm": 1.745692253112793,
+            "learning_rate": 0.0005,
+            "epoch": 0.8430299329408008,
+            "step": 18810
+        },
+        {
+            "loss": 13.9183,
+            "grad_norm": 1.9122262001037598,
+            "learning_rate": 0.0005,
+            "epoch": 0.8432540238320663,
+            "step": 18815
+        },
+        {
+            "loss": 13.9377,
+            "grad_norm": 1.7762372493743896,
+            "learning_rate": 0.0005,
+            "epoch": 0.8434781147233318,
+            "step": 18820
+        },
+        {
+            "loss": 13.8317,
+            "grad_norm": 1.7507115602493286,
+            "learning_rate": 0.0005,
+            "epoch": 0.8437022056145973,
+            "step": 18825
+        },
+        {
+            "loss": 13.9714,
+            "grad_norm": 1.6686385869979858,
+            "learning_rate": 0.0005,
+            "epoch": 0.8439262965058628,
+            "step": 18830
+        },
+        {
+            "loss": 13.8847,
+            "grad_norm": 1.852190375328064,
+            "learning_rate": 0.0005,
+            "epoch": 0.8441503873971282,
+            "step": 18835
+        },
+        {
+            "loss": 13.8739,
+            "grad_norm": 1.7109217643737793,
+            "learning_rate": 0.0005,
+            "epoch": 0.8443744782883937,
+            "step": 18840
+        },
+        {
+            "loss": 13.9082,
+            "grad_norm": 1.9363154172897339,
+            "learning_rate": 0.0005,
+            "epoch": 0.8445985691796593,
+            "step": 18845
+        },
+        {
+            "loss": 13.8858,
+            "grad_norm": 1.8679718971252441,
+            "learning_rate": 0.0005,
+            "epoch": 0.8448226600709248,
+            "step": 18850
+        },
+        {
+            "loss": 13.8217,
+            "grad_norm": 2.1826541423797607,
+            "learning_rate": 0.0005,
+            "epoch": 0.8450467509621903,
+            "step": 18855
+        },
+        {
+            "loss": 13.9356,
+            "grad_norm": 1.6823848485946655,
+            "learning_rate": 0.0005,
+            "epoch": 0.8452708418534558,
+            "step": 18860
+        },
+        {
+            "loss": 13.8828,
+            "grad_norm": 1.6551467180252075,
+            "learning_rate": 0.0005,
+            "epoch": 0.8454949327447212,
+            "step": 18865
+        },
+        {
+            "loss": 13.9097,
+            "grad_norm": 1.7100468873977661,
+            "learning_rate": 0.0005,
+            "epoch": 0.8457190236359867,
+            "step": 18870
+        },
+        {
+            "loss": 13.987,
+            "grad_norm": 1.6265673637390137,
+            "learning_rate": 0.0005,
+            "epoch": 0.8459431145272522,
+            "step": 18875
+        },
+        {
+            "loss": 13.9429,
+            "grad_norm": 1.8793754577636719,
+            "learning_rate": 0.0005,
+            "epoch": 0.8461672054185178,
+            "step": 18880
+        },
+        {
+            "loss": 13.897,
+            "grad_norm": 1.7657238245010376,
+            "learning_rate": 0.0005,
+            "epoch": 0.8463912963097833,
+            "step": 18885
+        },
+        {
+            "loss": 13.9154,
+            "grad_norm": 1.8032546043395996,
+            "learning_rate": 0.0005,
+            "epoch": 0.8466153872010488,
+            "step": 18890
+        },
+        {
+            "loss": 14.0057,
+            "grad_norm": 1.7182867527008057,
+            "learning_rate": 0.0005,
+            "epoch": 0.8468394780923142,
+            "step": 18895
+        },
+        {
+            "loss": 13.8333,
+            "grad_norm": 1.7613600492477417,
+            "learning_rate": 0.0005,
+            "epoch": 0.8470635689835797,
+            "step": 18900
+        },
+        {
+            "loss": 13.9154,
+            "grad_norm": 1.8472468852996826,
+            "learning_rate": 0.0005,
+            "epoch": 0.8472876598748452,
+            "step": 18905
+        },
+        {
+            "loss": 13.8747,
+            "grad_norm": 1.6938554048538208,
+            "learning_rate": 0.0005,
+            "epoch": 0.8475117507661107,
+            "step": 18910
+        },
+        {
+            "loss": 13.9094,
+            "grad_norm": 1.8218588829040527,
+            "learning_rate": 0.0005,
+            "epoch": 0.8477358416573763,
+            "step": 18915
+        },
+        {
+            "loss": 13.91,
+            "grad_norm": 1.8205779790878296,
+            "learning_rate": 0.0005,
+            "epoch": 0.8479599325486418,
+            "step": 18920
+        },
+        {
+            "loss": 14.0384,
+            "grad_norm": 1.8513416051864624,
+            "learning_rate": 0.0005,
+            "epoch": 0.8481840234399072,
+            "step": 18925
+        },
+        {
+            "loss": 13.8546,
+            "grad_norm": 1.8248463869094849,
+            "learning_rate": 0.0005,
+            "epoch": 0.8484081143311727,
+            "step": 18930
+        },
+        {
+            "loss": 13.9754,
+            "grad_norm": 2.1561028957366943,
+            "learning_rate": 0.0005,
+            "epoch": 0.8486322052224382,
+            "step": 18935
+        },
+        {
+            "loss": 13.9083,
+            "grad_norm": 1.9676803350448608,
+            "learning_rate": 0.0005,
+            "epoch": 0.8488562961137037,
+            "step": 18940
+        },
+        {
+            "loss": 13.9145,
+            "grad_norm": 1.7778428792953491,
+            "learning_rate": 0.0005,
+            "epoch": 0.8490803870049692,
+            "step": 18945
+        },
+        {
+            "loss": 13.9913,
+            "grad_norm": 1.7351329326629639,
+            "learning_rate": 0.0005,
+            "epoch": 0.8493044778962348,
+            "step": 18950
+        },
+        {
+            "loss": 13.84,
+            "grad_norm": 1.831153154373169,
+            "learning_rate": 0.0005,
+            "epoch": 0.8495285687875002,
+            "step": 18955
+        },
+        {
+            "loss": 13.8671,
+            "grad_norm": 1.6546090841293335,
+            "learning_rate": 0.0005,
+            "epoch": 0.8497526596787657,
+            "step": 18960
+        },
+        {
+            "loss": 13.9135,
+            "grad_norm": 1.8007551431655884,
+            "learning_rate": 0.0005,
+            "epoch": 0.8499767505700312,
+            "step": 18965
+        },
+        {
+            "loss": 13.9361,
+            "grad_norm": 2.103647232055664,
+            "learning_rate": 0.0005,
+            "epoch": 0.8502008414612967,
+            "step": 18970
+        },
+        {
+            "loss": 13.8892,
+            "grad_norm": 1.7419698238372803,
+            "learning_rate": 0.0005,
+            "epoch": 0.8504249323525622,
+            "step": 18975
+        },
+        {
+            "loss": 13.9764,
+            "grad_norm": 1.784730076789856,
+            "learning_rate": 0.0005,
+            "epoch": 0.8506490232438277,
+            "step": 18980
+        },
+        {
+            "loss": 13.9872,
+            "grad_norm": 1.8239227533340454,
+            "learning_rate": 0.0005,
+            "epoch": 0.8508731141350931,
+            "step": 18985
+        },
+        {
+            "loss": 13.8051,
+            "grad_norm": 1.6747732162475586,
+            "learning_rate": 0.0005,
+            "epoch": 0.8510972050263587,
+            "step": 18990
+        },
+        {
+            "loss": 13.8853,
+            "grad_norm": 1.812728762626648,
+            "learning_rate": 0.0005,
+            "epoch": 0.8513212959176242,
+            "step": 18995
+        },
+        {
+            "loss": 13.9022,
+            "grad_norm": 1.6415209770202637,
+            "learning_rate": 0.0005,
+            "epoch": 0.8515453868088897,
+            "step": 19000
+        },
+        {
+            "eval_loss": 1.7338833808898926,
+            "eval_runtime": 18.642,
+            "eval_samples_per_second": 878.875,
+            "eval_steps_per_second": 7.885,
+            "epoch": 0.8515453868088897,
+            "step": 19000
+        },
+        {
+            "loss": 13.9902,
+            "grad_norm": 1.8604235649108887,
+            "learning_rate": 0.0005,
+            "epoch": 0.8517694777001552,
+            "step": 19005
+        },
+        {
+            "loss": 13.9708,
+            "grad_norm": 2.1418986320495605,
+            "learning_rate": 0.0005,
+            "epoch": 0.8519935685914207,
+            "step": 19010
+        },
+        {
+            "loss": 13.8404,
+            "grad_norm": 1.7601839303970337,
+            "learning_rate": 0.0005,
+            "epoch": 0.8522176594826861,
+            "step": 19015
+        },
+        {
+            "loss": 13.8326,
+            "grad_norm": 1.786251187324524,
+            "learning_rate": 0.0005,
+            "epoch": 0.8524417503739516,
+            "step": 19020
+        },
+        {
+            "loss": 13.9209,
+            "grad_norm": 1.8217843770980835,
+            "learning_rate": 0.0005,
+            "epoch": 0.8526658412652172,
+            "step": 19025
+        },
+        {
+            "loss": 13.8807,
+            "grad_norm": 1.7725396156311035,
+            "learning_rate": 0.0005,
+            "epoch": 0.8528899321564827,
+            "step": 19030
+        },
+        {
+            "loss": 13.9871,
+            "grad_norm": 1.8573862314224243,
+            "learning_rate": 0.0005,
+            "epoch": 0.8531140230477482,
+            "step": 19035
+        },
+        {
+            "loss": 13.8264,
+            "grad_norm": 1.7490694522857666,
+            "learning_rate": 0.0005,
+            "epoch": 0.8533381139390137,
+            "step": 19040
+        },
+        {
+            "loss": 13.8547,
+            "grad_norm": 1.8021537065505981,
+            "learning_rate": 0.0005,
+            "epoch": 0.8535622048302791,
+            "step": 19045
+        },
+        {
+            "loss": 13.8538,
+            "grad_norm": 2.0671534538269043,
+            "learning_rate": 0.0005,
+            "epoch": 0.8537862957215446,
+            "step": 19050
+        },
+        {
+            "loss": 13.9328,
+            "grad_norm": 1.8159527778625488,
+            "learning_rate": 0.0005,
+            "epoch": 0.8540103866128101,
+            "step": 19055
+        },
+        {
+            "loss": 13.8714,
+            "grad_norm": 1.8175792694091797,
+            "learning_rate": 0.0005,
+            "epoch": 0.8542344775040757,
+            "step": 19060
+        },
+        {
+            "loss": 13.8578,
+            "grad_norm": 1.8844847679138184,
+            "learning_rate": 0.0005,
+            "epoch": 0.8544585683953412,
+            "step": 19065
+        },
+        {
+            "loss": 13.9442,
+            "grad_norm": 1.9673024415969849,
+            "learning_rate": 0.0005,
+            "epoch": 0.8546826592866067,
+            "step": 19070
+        },
+        {
+            "loss": 13.871,
+            "grad_norm": 1.8300015926361084,
+            "learning_rate": 0.0005,
+            "epoch": 0.8549067501778721,
+            "step": 19075
+        },
+        {
+            "loss": 13.913,
+            "grad_norm": 1.7731380462646484,
+            "learning_rate": 0.0005,
+            "epoch": 0.8551308410691376,
+            "step": 19080
+        },
+        {
+            "loss": 14.018,
+            "grad_norm": 1.9178194999694824,
+            "learning_rate": 0.0005,
+            "epoch": 0.8553549319604031,
+            "step": 19085
+        },
+        {
+            "loss": 13.7574,
+            "grad_norm": 1.7870759963989258,
+            "learning_rate": 0.0005,
+            "epoch": 0.8555790228516686,
+            "step": 19090
+        },
+        {
+            "loss": 13.7963,
+            "grad_norm": 1.7498259544372559,
+            "learning_rate": 0.0005,
+            "epoch": 0.8558031137429342,
+            "step": 19095
+        },
+        {
+            "loss": 13.8313,
+            "grad_norm": 1.940000295639038,
+            "learning_rate": 0.0005,
+            "epoch": 0.8560272046341997,
+            "step": 19100
+        },
+        {
+            "loss": 13.858,
+            "grad_norm": 2.03861403465271,
+            "learning_rate": 0.0005,
+            "epoch": 0.8562512955254651,
+            "step": 19105
+        },
+        {
+            "loss": 13.9348,
+            "grad_norm": 1.7815747261047363,
+            "learning_rate": 0.0005,
+            "epoch": 0.8564753864167306,
+            "step": 19110
+        },
+        {
+            "loss": 13.8962,
+            "grad_norm": 1.8152567148208618,
+            "learning_rate": 0.0005,
+            "epoch": 0.8566994773079961,
+            "step": 19115
+        },
+        {
+            "loss": 13.8382,
+            "grad_norm": 1.733628511428833,
+            "learning_rate": 0.0005,
+            "epoch": 0.8569235681992616,
+            "step": 19120
+        },
+        {
+            "loss": 13.7807,
+            "grad_norm": 1.7825106382369995,
+            "learning_rate": 0.0005,
+            "epoch": 0.8571476590905271,
+            "step": 19125
+        },
+        {
+            "loss": 13.9753,
+            "grad_norm": 1.7949985265731812,
+            "learning_rate": 0.0005,
+            "epoch": 0.8573717499817927,
+            "step": 19130
+        },
+        {
+            "loss": 13.9514,
+            "grad_norm": 1.7800675630569458,
+            "learning_rate": 0.0005,
+            "epoch": 0.8575958408730581,
+            "step": 19135
+        },
+        {
+            "loss": 13.8505,
+            "grad_norm": 1.756445288658142,
+            "learning_rate": 0.0005,
+            "epoch": 0.8578199317643236,
+            "step": 19140
+        },
+        {
+            "loss": 13.9917,
+            "grad_norm": 1.7357224225997925,
+            "learning_rate": 0.0005,
+            "epoch": 0.8580440226555891,
+            "step": 19145
+        },
+        {
+            "loss": 13.9566,
+            "grad_norm": 1.7281850576400757,
+            "learning_rate": 0.0005,
+            "epoch": 0.8582681135468546,
+            "step": 19150
+        },
+        {
+            "loss": 13.9108,
+            "grad_norm": 1.8354952335357666,
+            "learning_rate": 0.0005,
+            "epoch": 0.8584922044381201,
+            "step": 19155
+        },
+        {
+            "loss": 13.8971,
+            "grad_norm": 1.9671531915664673,
+            "learning_rate": 0.0005,
+            "epoch": 0.8587162953293856,
+            "step": 19160
+        },
+        {
+            "loss": 13.9922,
+            "grad_norm": 1.797380805015564,
+            "learning_rate": 0.0005,
+            "epoch": 0.858940386220651,
+            "step": 19165
+        },
+        {
+            "loss": 13.8633,
+            "grad_norm": 1.87890625,
+            "learning_rate": 0.0005,
+            "epoch": 0.8591644771119166,
+            "step": 19170
+        },
+        {
+            "loss": 14.0032,
+            "grad_norm": 1.718583345413208,
+            "learning_rate": 0.0005,
+            "epoch": 0.8593885680031821,
+            "step": 19175
+        },
+        {
+            "loss": 13.8237,
+            "grad_norm": 1.7219116687774658,
+            "learning_rate": 0.0005,
+            "epoch": 0.8596126588944476,
+            "step": 19180
+        },
+        {
+            "loss": 13.8988,
+            "grad_norm": 1.7475054264068604,
+            "learning_rate": 0.0005,
+            "epoch": 0.8598367497857131,
+            "step": 19185
+        },
+        {
+            "loss": 13.8977,
+            "grad_norm": 1.7914988994598389,
+            "learning_rate": 0.0005,
+            "epoch": 0.8600608406769786,
+            "step": 19190
+        },
+        {
+            "loss": 13.982,
+            "grad_norm": 1.9456027746200562,
+            "learning_rate": 0.0005,
+            "epoch": 0.860284931568244,
+            "step": 19195
+        },
+        {
+            "loss": 13.9596,
+            "grad_norm": 1.8214281797409058,
+            "learning_rate": 0.0005,
+            "epoch": 0.8605090224595096,
+            "step": 19200
+        },
+        {
+            "loss": 13.9707,
+            "grad_norm": 1.917678713798523,
+            "learning_rate": 0.0005,
+            "epoch": 0.8607331133507751,
+            "step": 19205
+        },
+        {
+            "loss": 13.9555,
+            "grad_norm": 1.9066157341003418,
+            "learning_rate": 0.0005,
+            "epoch": 0.8609572042420406,
+            "step": 19210
+        },
+        {
+            "loss": 13.8718,
+            "grad_norm": 2.0217795372009277,
+            "learning_rate": 0.0005,
+            "epoch": 0.8611812951333061,
+            "step": 19215
+        },
+        {
+            "loss": 13.9169,
+            "grad_norm": 1.7404314279556274,
+            "learning_rate": 0.0005,
+            "epoch": 0.8614053860245716,
+            "step": 19220
+        },
+        {
+            "loss": 13.9322,
+            "grad_norm": 1.8170093297958374,
+            "learning_rate": 0.0005,
+            "epoch": 0.861629476915837,
+            "step": 19225
+        },
+        {
+            "loss": 13.8852,
+            "grad_norm": 1.9152315855026245,
+            "learning_rate": 0.0005,
+            "epoch": 0.8618535678071025,
+            "step": 19230
+        },
+        {
+            "loss": 13.9438,
+            "grad_norm": 1.834559679031372,
+            "learning_rate": 0.0005,
+            "epoch": 0.862077658698368,
+            "step": 19235
+        },
+        {
+            "loss": 13.9153,
+            "grad_norm": 1.789071798324585,
+            "learning_rate": 0.0005,
+            "epoch": 0.8623017495896336,
+            "step": 19240
+        },
+        {
+            "loss": 13.9149,
+            "grad_norm": 1.931227684020996,
+            "learning_rate": 0.0005,
+            "epoch": 0.8625258404808991,
+            "step": 19245
+        },
+        {
+            "loss": 13.8789,
+            "grad_norm": 1.8586195707321167,
+            "learning_rate": 0.0005,
+            "epoch": 0.8627499313721646,
+            "step": 19250
+        },
+        {
+            "loss": 13.9125,
+            "grad_norm": 1.851481556892395,
+            "learning_rate": 0.0005,
+            "epoch": 0.86297402226343,
+            "step": 19255
+        },
+        {
+            "loss": 13.9039,
+            "grad_norm": 1.7920666933059692,
+            "learning_rate": 0.0005,
+            "epoch": 0.8631981131546955,
+            "step": 19260
+        },
+        {
+            "loss": 13.8492,
+            "grad_norm": 1.775800108909607,
+            "learning_rate": 0.0005,
+            "epoch": 0.863422204045961,
+            "step": 19265
+        },
+        {
+            "loss": 13.9419,
+            "grad_norm": 2.010636329650879,
+            "learning_rate": 0.0005,
+            "epoch": 0.8636462949372266,
+            "step": 19270
+        },
+        {
+            "loss": 13.8676,
+            "grad_norm": 1.778244972229004,
+            "learning_rate": 0.0005,
+            "epoch": 0.8638703858284921,
+            "step": 19275
+        },
+        {
+            "loss": 13.8888,
+            "grad_norm": 1.7896620035171509,
+            "learning_rate": 0.0005,
+            "epoch": 0.8640944767197576,
+            "step": 19280
+        },
+        {
+            "loss": 13.7832,
+            "grad_norm": 1.8602795600891113,
+            "learning_rate": 0.0005,
+            "epoch": 0.864318567611023,
+            "step": 19285
+        },
+        {
+            "loss": 13.9504,
+            "grad_norm": 1.7610522508621216,
+            "learning_rate": 0.0005,
+            "epoch": 0.8645426585022885,
+            "step": 19290
+        },
+        {
+            "loss": 13.8595,
+            "grad_norm": 1.7807914018630981,
+            "learning_rate": 0.0005,
+            "epoch": 0.864766749393554,
+            "step": 19295
+        },
+        {
+            "loss": 13.8986,
+            "grad_norm": 1.550917148590088,
+            "learning_rate": 0.0005,
+            "epoch": 0.8649908402848195,
+            "step": 19300
+        },
+        {
+            "loss": 13.9378,
+            "grad_norm": 1.7979971170425415,
+            "learning_rate": 0.0005,
+            "epoch": 0.8652149311760851,
+            "step": 19305
+        },
+        {
+            "loss": 13.936,
+            "grad_norm": 1.7553660869598389,
+            "learning_rate": 0.0005,
+            "epoch": 0.8654390220673506,
+            "step": 19310
+        },
+        {
+            "loss": 13.7785,
+            "grad_norm": 1.725150465965271,
+            "learning_rate": 0.0005,
+            "epoch": 0.865663112958616,
+            "step": 19315
+        },
+        {
+            "loss": 13.8794,
+            "grad_norm": 2.143718957901001,
+            "learning_rate": 0.0005,
+            "epoch": 0.8658872038498815,
+            "step": 19320
+        },
+        {
+            "loss": 13.8788,
+            "grad_norm": 1.7977149486541748,
+            "learning_rate": 0.0005,
+            "epoch": 0.866111294741147,
+            "step": 19325
+        },
+        {
+            "loss": 13.8802,
+            "grad_norm": 1.9429805278778076,
+            "learning_rate": 0.0005,
+            "epoch": 0.8663353856324125,
+            "step": 19330
+        },
+        {
+            "loss": 13.831,
+            "grad_norm": 1.764933705329895,
+            "learning_rate": 0.0005,
+            "epoch": 0.866559476523678,
+            "step": 19335
+        },
+        {
+            "loss": 13.8147,
+            "grad_norm": 1.8815680742263794,
+            "learning_rate": 0.0005,
+            "epoch": 0.8667835674149436,
+            "step": 19340
+        },
+        {
+            "loss": 13.9175,
+            "grad_norm": 1.7224299907684326,
+            "learning_rate": 0.0005,
+            "epoch": 0.867007658306209,
+            "step": 19345
+        },
+        {
+            "loss": 13.86,
+            "grad_norm": 1.8962249755859375,
+            "learning_rate": 0.0005,
+            "epoch": 0.8672317491974745,
+            "step": 19350
+        },
+        {
+            "loss": 13.9039,
+            "grad_norm": 1.7022427320480347,
+            "learning_rate": 0.0005,
+            "epoch": 0.86745584008874,
+            "step": 19355
+        },
+        {
+            "loss": 13.886,
+            "grad_norm": 1.8437137603759766,
+            "learning_rate": 0.0005,
+            "epoch": 0.8676799309800055,
+            "step": 19360
+        },
+        {
+            "loss": 13.9262,
+            "grad_norm": 1.783321499824524,
+            "learning_rate": 0.0005,
+            "epoch": 0.867904021871271,
+            "step": 19365
+        },
+        {
+            "loss": 13.883,
+            "grad_norm": 1.7349745035171509,
+            "learning_rate": 0.0005,
+            "epoch": 0.8681281127625364,
+            "step": 19370
+        },
+        {
+            "loss": 13.8675,
+            "grad_norm": 1.6686931848526,
+            "learning_rate": 0.0005,
+            "epoch": 0.868352203653802,
+            "step": 19375
+        },
+        {
+            "loss": 13.7891,
+            "grad_norm": 1.6508697271347046,
+            "learning_rate": 0.0005,
+            "epoch": 0.8685762945450675,
+            "step": 19380
+        },
+        {
+            "loss": 13.9467,
+            "grad_norm": 1.7480441331863403,
+            "learning_rate": 0.0005,
+            "epoch": 0.868800385436333,
+            "step": 19385
+        },
+        {
+            "loss": 13.9946,
+            "grad_norm": 1.7600085735321045,
+            "learning_rate": 0.0005,
+            "epoch": 0.8690244763275985,
+            "step": 19390
+        },
+        {
+            "loss": 13.8532,
+            "grad_norm": 1.7637922763824463,
+            "learning_rate": 0.0005,
+            "epoch": 0.869248567218864,
+            "step": 19395
+        },
+        {
+            "loss": 13.7949,
+            "grad_norm": 1.7422178983688354,
+            "learning_rate": 0.0005,
+            "epoch": 0.8694726581101294,
+            "step": 19400
+        },
+        {
+            "loss": 13.8172,
+            "grad_norm": 1.7967084646224976,
+            "learning_rate": 0.0005,
+            "epoch": 0.8696967490013949,
+            "step": 19405
+        },
+        {
+            "loss": 13.7484,
+            "grad_norm": 1.632609486579895,
+            "learning_rate": 0.0005,
+            "epoch": 0.8699208398926604,
+            "step": 19410
+        },
+        {
+            "loss": 13.9417,
+            "grad_norm": 1.8991572856903076,
+            "learning_rate": 0.0005,
+            "epoch": 0.870144930783926,
+            "step": 19415
+        },
+        {
+            "loss": 13.9278,
+            "grad_norm": 1.9792591333389282,
+            "learning_rate": 0.0005,
+            "epoch": 0.8703690216751915,
+            "step": 19420
+        },
+        {
+            "loss": 13.9294,
+            "grad_norm": 1.761389970779419,
+            "learning_rate": 0.0005,
+            "epoch": 0.870593112566457,
+            "step": 19425
+        },
+        {
+            "loss": 13.9177,
+            "grad_norm": 1.7680258750915527,
+            "learning_rate": 0.0005,
+            "epoch": 0.8708172034577224,
+            "step": 19430
+        },
+        {
+            "loss": 13.9728,
+            "grad_norm": 2.0201072692871094,
+            "learning_rate": 0.0005,
+            "epoch": 0.8710412943489879,
+            "step": 19435
+        },
+        {
+            "loss": 13.8466,
+            "grad_norm": 1.7837570905685425,
+            "learning_rate": 0.0005,
+            "epoch": 0.8712653852402534,
+            "step": 19440
+        },
+        {
+            "loss": 13.8657,
+            "grad_norm": 1.7108758687973022,
+            "learning_rate": 0.0005,
+            "epoch": 0.871489476131519,
+            "step": 19445
+        },
+        {
+            "loss": 13.916,
+            "grad_norm": 1.9229105710983276,
+            "learning_rate": 0.0005,
+            "epoch": 0.8717135670227845,
+            "step": 19450
+        },
+        {
+            "loss": 13.8466,
+            "grad_norm": 1.9333982467651367,
+            "learning_rate": 0.0005,
+            "epoch": 0.87193765791405,
+            "step": 19455
+        },
+        {
+            "loss": 13.8918,
+            "grad_norm": 1.7065330743789673,
+            "learning_rate": 0.0005,
+            "epoch": 0.8721617488053154,
+            "step": 19460
+        },
+        {
+            "loss": 13.809,
+            "grad_norm": 2.0091190338134766,
+            "learning_rate": 0.0005,
+            "epoch": 0.8723858396965809,
+            "step": 19465
+        },
+        {
+            "loss": 13.9354,
+            "grad_norm": 1.8350383043289185,
+            "learning_rate": 0.0005,
+            "epoch": 0.8726099305878464,
+            "step": 19470
+        },
+        {
+            "loss": 13.8716,
+            "grad_norm": 1.9058568477630615,
+            "learning_rate": 0.0005,
+            "epoch": 0.8728340214791119,
+            "step": 19475
+        },
+        {
+            "loss": 13.8056,
+            "grad_norm": 1.6378625631332397,
+            "learning_rate": 0.0005,
+            "epoch": 0.8730581123703774,
+            "step": 19480
+        },
+        {
+            "loss": 13.8535,
+            "grad_norm": 1.866127610206604,
+            "learning_rate": 0.0005,
+            "epoch": 0.873282203261643,
+            "step": 19485
+        },
+        {
+            "loss": 13.9903,
+            "grad_norm": 1.8506743907928467,
+            "learning_rate": 0.0005,
+            "epoch": 0.8735062941529084,
+            "step": 19490
+        },
+        {
+            "loss": 13.9016,
+            "grad_norm": 1.6948919296264648,
+            "learning_rate": 0.0005,
+            "epoch": 0.8737303850441739,
+            "step": 19495
+        },
+        {
+            "loss": 13.9009,
+            "grad_norm": 1.6821815967559814,
+            "learning_rate": 0.0005,
+            "epoch": 0.8739544759354394,
+            "step": 19500
+        },
+        {
+            "eval_loss": 1.7357780933380127,
+            "eval_runtime": 18.3496,
+            "eval_samples_per_second": 892.88,
+            "eval_steps_per_second": 8.011,
+            "epoch": 0.8739544759354394,
+            "step": 19500
+        },
+        {
+            "loss": 13.8466,
+            "grad_norm": 1.8314194679260254,
+            "learning_rate": 0.0005,
+            "epoch": 0.8741785668267049,
+            "step": 19505
+        },
+        {
+            "loss": 13.8573,
+            "grad_norm": 1.7998930215835571,
+            "learning_rate": 0.0005,
+            "epoch": 0.8744026577179704,
+            "step": 19510
+        },
+        {
+            "loss": 13.8579,
+            "grad_norm": 1.8054667711257935,
+            "learning_rate": 0.0005,
+            "epoch": 0.874626748609236,
+            "step": 19515
+        },
+        {
+            "loss": 13.8764,
+            "grad_norm": 1.7818348407745361,
+            "learning_rate": 0.0005,
+            "epoch": 0.8748508395005014,
+            "step": 19520
+        },
+        {
+            "loss": 13.887,
+            "grad_norm": 1.737918734550476,
+            "learning_rate": 0.0005,
+            "epoch": 0.8750749303917669,
+            "step": 19525
+        },
+        {
+            "loss": 13.9379,
+            "grad_norm": 1.8601570129394531,
+            "learning_rate": 0.0005,
+            "epoch": 0.8752990212830324,
+            "step": 19530
+        },
+        {
+            "loss": 13.8531,
+            "grad_norm": 1.8514798879623413,
+            "learning_rate": 0.0005,
+            "epoch": 0.8755231121742979,
+            "step": 19535
+        },
+        {
+            "loss": 13.9075,
+            "grad_norm": 1.7769105434417725,
+            "learning_rate": 0.0005,
+            "epoch": 0.8757472030655634,
+            "step": 19540
+        },
+        {
+            "loss": 13.867,
+            "grad_norm": 1.6772818565368652,
+            "learning_rate": 0.0005,
+            "epoch": 0.8759712939568289,
+            "step": 19545
+        },
+        {
+            "loss": 13.9573,
+            "grad_norm": 1.7507994174957275,
+            "learning_rate": 0.0005,
+            "epoch": 0.8761953848480943,
+            "step": 19550
+        },
+        {
+            "loss": 13.8608,
+            "grad_norm": 1.710281491279602,
+            "learning_rate": 0.0005,
+            "epoch": 0.8764194757393599,
+            "step": 19555
+        },
+        {
+            "loss": 13.9944,
+            "grad_norm": 1.6837866306304932,
+            "learning_rate": 0.0005,
+            "epoch": 0.8766435666306254,
+            "step": 19560
+        },
+        {
+            "loss": 13.8669,
+            "grad_norm": 1.816420078277588,
+            "learning_rate": 0.0005,
+            "epoch": 0.8768676575218909,
+            "step": 19565
+        },
+        {
+            "loss": 13.8127,
+            "grad_norm": 1.7850531339645386,
+            "learning_rate": 0.0005,
+            "epoch": 0.8770917484131564,
+            "step": 19570
+        },
+        {
+            "loss": 13.9502,
+            "grad_norm": 2.1283504962921143,
+            "learning_rate": 0.0005,
+            "epoch": 0.8773158393044219,
+            "step": 19575
+        },
+        {
+            "loss": 13.859,
+            "grad_norm": 1.8324092626571655,
+            "learning_rate": 0.0005,
+            "epoch": 0.8775399301956873,
+            "step": 19580
+        },
+        {
+            "loss": 13.8817,
+            "grad_norm": 1.8633873462677002,
+            "learning_rate": 0.0005,
+            "epoch": 0.8777640210869528,
+            "step": 19585
+        },
+        {
+            "loss": 13.9691,
+            "grad_norm": 1.7599881887435913,
+            "learning_rate": 0.0005,
+            "epoch": 0.8779881119782184,
+            "step": 19590
+        },
+        {
+            "loss": 13.8084,
+            "grad_norm": 1.763671636581421,
+            "learning_rate": 0.0005,
+            "epoch": 0.8782122028694839,
+            "step": 19595
+        },
+        {
+            "loss": 13.9648,
+            "grad_norm": 1.6990044116973877,
+            "learning_rate": 0.0005,
+            "epoch": 0.8784362937607494,
+            "step": 19600
+        },
+        {
+            "loss": 13.8874,
+            "grad_norm": 1.7792152166366577,
+            "learning_rate": 0.0005,
+            "epoch": 0.8786603846520149,
+            "step": 19605
+        },
+        {
+            "loss": 13.9478,
+            "grad_norm": 2.006032705307007,
+            "learning_rate": 0.0005,
+            "epoch": 0.8788844755432803,
+            "step": 19610
+        },
+        {
+            "loss": 13.8684,
+            "grad_norm": 1.5835766792297363,
+            "learning_rate": 0.0005,
+            "epoch": 0.8791085664345458,
+            "step": 19615
+        },
+        {
+            "loss": 13.977,
+            "grad_norm": 1.6947407722473145,
+            "learning_rate": 0.0005,
+            "epoch": 0.8793326573258113,
+            "step": 19620
+        },
+        {
+            "loss": 13.8611,
+            "grad_norm": 1.629675269126892,
+            "learning_rate": 0.0005,
+            "epoch": 0.8795567482170769,
+            "step": 19625
+        },
+        {
+            "loss": 13.9137,
+            "grad_norm": 1.6017608642578125,
+            "learning_rate": 0.0005,
+            "epoch": 0.8797808391083424,
+            "step": 19630
+        },
+        {
+            "loss": 13.8787,
+            "grad_norm": 1.7836179733276367,
+            "learning_rate": 0.0005,
+            "epoch": 0.8800049299996079,
+            "step": 19635
+        },
+        {
+            "loss": 13.8423,
+            "grad_norm": 1.722022294998169,
+            "learning_rate": 0.0005,
+            "epoch": 0.8802290208908733,
+            "step": 19640
+        },
+        {
+            "loss": 13.9003,
+            "grad_norm": 1.904645562171936,
+            "learning_rate": 0.0005,
+            "epoch": 0.8804531117821388,
+            "step": 19645
+        },
+        {
+            "loss": 13.8872,
+            "grad_norm": 1.8378710746765137,
+            "learning_rate": 0.0005,
+            "epoch": 0.8806772026734043,
+            "step": 19650
+        },
+        {
+            "loss": 13.876,
+            "grad_norm": 1.9562255144119263,
+            "learning_rate": 0.0005,
+            "epoch": 0.8809012935646698,
+            "step": 19655
+        },
+        {
+            "loss": 13.957,
+            "grad_norm": 1.8990522623062134,
+            "learning_rate": 0.0005,
+            "epoch": 0.8811253844559354,
+            "step": 19660
+        },
+        {
+            "loss": 13.8762,
+            "grad_norm": 1.7711070775985718,
+            "learning_rate": 0.0005,
+            "epoch": 0.8813494753472009,
+            "step": 19665
+        },
+        {
+            "loss": 13.7974,
+            "grad_norm": 1.7412478923797607,
+            "learning_rate": 0.0005,
+            "epoch": 0.8815735662384663,
+            "step": 19670
+        },
+        {
+            "loss": 13.894,
+            "grad_norm": 1.8522121906280518,
+            "learning_rate": 0.0005,
+            "epoch": 0.8817976571297318,
+            "step": 19675
+        },
+        {
+            "loss": 13.8729,
+            "grad_norm": 1.6800007820129395,
+            "learning_rate": 0.0005,
+            "epoch": 0.8820217480209973,
+            "step": 19680
+        },
+        {
+            "loss": 13.8689,
+            "grad_norm": 1.7280434370040894,
+            "learning_rate": 0.0005,
+            "epoch": 0.8822458389122628,
+            "step": 19685
+        },
+        {
+            "loss": 13.8875,
+            "grad_norm": 1.8460952043533325,
+            "learning_rate": 0.0005,
+            "epoch": 0.8824699298035283,
+            "step": 19690
+        },
+        {
+            "loss": 14.0027,
+            "grad_norm": 1.9587526321411133,
+            "learning_rate": 0.0005,
+            "epoch": 0.8826940206947939,
+            "step": 19695
+        },
+        {
+            "loss": 13.7925,
+            "grad_norm": 1.8540096282958984,
+            "learning_rate": 0.0005,
+            "epoch": 0.8829181115860593,
+            "step": 19700
+        },
+        {
+            "loss": 13.8576,
+            "grad_norm": 1.955402135848999,
+            "learning_rate": 0.0005,
+            "epoch": 0.8831422024773248,
+            "step": 19705
+        },
+        {
+            "loss": 13.8973,
+            "grad_norm": 1.7868638038635254,
+            "learning_rate": 0.0005,
+            "epoch": 0.8833662933685903,
+            "step": 19710
+        },
+        {
+            "loss": 13.9396,
+            "grad_norm": 1.7307571172714233,
+            "learning_rate": 0.0005,
+            "epoch": 0.8835903842598558,
+            "step": 19715
+        },
+        {
+            "loss": 13.8281,
+            "grad_norm": 1.8804699182510376,
+            "learning_rate": 0.0005,
+            "epoch": 0.8838144751511213,
+            "step": 19720
+        },
+        {
+            "loss": 13.8287,
+            "grad_norm": 1.7727694511413574,
+            "learning_rate": 0.0005,
+            "epoch": 0.8840385660423868,
+            "step": 19725
+        },
+        {
+            "loss": 13.9586,
+            "grad_norm": 1.6995322704315186,
+            "learning_rate": 0.0005,
+            "epoch": 0.8842626569336522,
+            "step": 19730
+        },
+        {
+            "loss": 13.9179,
+            "grad_norm": 1.7488749027252197,
+            "learning_rate": 0.0005,
+            "epoch": 0.8844867478249178,
+            "step": 19735
+        },
+        {
+            "loss": 13.8479,
+            "grad_norm": 1.7972700595855713,
+            "learning_rate": 0.0005,
+            "epoch": 0.8847108387161833,
+            "step": 19740
+        },
+        {
+            "loss": 13.7467,
+            "grad_norm": 1.8058732748031616,
+            "learning_rate": 0.0005,
+            "epoch": 0.8849349296074488,
+            "step": 19745
+        },
+        {
+            "loss": 13.8238,
+            "grad_norm": 1.9566853046417236,
+            "learning_rate": 0.0005,
+            "epoch": 0.8851590204987143,
+            "step": 19750
+        },
+        {
+            "loss": 13.8789,
+            "grad_norm": 1.8107908964157104,
+            "learning_rate": 0.0005,
+            "epoch": 0.8853831113899798,
+            "step": 19755
+        },
+        {
+            "loss": 13.9075,
+            "grad_norm": 1.7433689832687378,
+            "learning_rate": 0.0005,
+            "epoch": 0.8856072022812452,
+            "step": 19760
+        },
+        {
+            "loss": 13.8934,
+            "grad_norm": 1.682373285293579,
+            "learning_rate": 0.0005,
+            "epoch": 0.8858312931725107,
+            "step": 19765
+        },
+        {
+            "loss": 13.8291,
+            "grad_norm": 1.7980401515960693,
+            "learning_rate": 0.0005,
+            "epoch": 0.8860553840637763,
+            "step": 19770
+        },
+        {
+            "loss": 13.8356,
+            "grad_norm": 1.777256965637207,
+            "learning_rate": 0.0005,
+            "epoch": 0.8862794749550418,
+            "step": 19775
+        },
+        {
+            "loss": 13.9096,
+            "grad_norm": 1.7960213422775269,
+            "learning_rate": 0.0005,
+            "epoch": 0.8865035658463073,
+            "step": 19780
+        },
+        {
+            "loss": 13.9526,
+            "grad_norm": 1.7837116718292236,
+            "learning_rate": 0.0005,
+            "epoch": 0.8867276567375728,
+            "step": 19785
+        },
+        {
+            "loss": 13.8384,
+            "grad_norm": 1.8213714361190796,
+            "learning_rate": 0.0005,
+            "epoch": 0.8869517476288382,
+            "step": 19790
+        },
+        {
+            "loss": 13.8886,
+            "grad_norm": 1.6740883588790894,
+            "learning_rate": 0.0005,
+            "epoch": 0.8871758385201037,
+            "step": 19795
+        },
+        {
+            "loss": 13.8911,
+            "grad_norm": 1.9716994762420654,
+            "learning_rate": 0.0005,
+            "epoch": 0.8873999294113692,
+            "step": 19800
+        },
+        {
+            "loss": 13.963,
+            "grad_norm": 1.802721381187439,
+            "learning_rate": 0.0005,
+            "epoch": 0.8876240203026348,
+            "step": 19805
+        },
+        {
+            "loss": 14.0038,
+            "grad_norm": 2.034996509552002,
+            "learning_rate": 0.0005,
+            "epoch": 0.8878481111939003,
+            "step": 19810
+        },
+        {
+            "loss": 13.7372,
+            "grad_norm": 1.8164931535720825,
+            "learning_rate": 0.0005,
+            "epoch": 0.8880722020851658,
+            "step": 19815
+        },
+        {
+            "loss": 13.9128,
+            "grad_norm": 1.8093912601470947,
+            "learning_rate": 0.0005,
+            "epoch": 0.8882962929764312,
+            "step": 19820
+        },
+        {
+            "loss": 13.8272,
+            "grad_norm": 1.724300503730774,
+            "learning_rate": 0.0005,
+            "epoch": 0.8885203838676967,
+            "step": 19825
+        },
+        {
+            "loss": 13.9372,
+            "grad_norm": 1.9092543125152588,
+            "learning_rate": 0.0005,
+            "epoch": 0.8887444747589622,
+            "step": 19830
+        },
+        {
+            "loss": 13.867,
+            "grad_norm": 1.949688196182251,
+            "learning_rate": 0.0005,
+            "epoch": 0.8889685656502277,
+            "step": 19835
+        },
+        {
+            "loss": 13.9037,
+            "grad_norm": 1.7329602241516113,
+            "learning_rate": 0.0005,
+            "epoch": 0.8891926565414933,
+            "step": 19840
+        },
+        {
+            "loss": 13.8886,
+            "grad_norm": 1.8665833473205566,
+            "learning_rate": 0.0005,
+            "epoch": 0.8894167474327588,
+            "step": 19845
+        },
+        {
+            "loss": 13.8636,
+            "grad_norm": 1.697461724281311,
+            "learning_rate": 0.0005,
+            "epoch": 0.8896408383240242,
+            "step": 19850
+        },
+        {
+            "loss": 13.8087,
+            "grad_norm": 1.8883639574050903,
+            "learning_rate": 0.0005,
+            "epoch": 0.8898649292152897,
+            "step": 19855
+        },
+        {
+            "loss": 13.9671,
+            "grad_norm": 1.902814269065857,
+            "learning_rate": 0.0005,
+            "epoch": 0.8900890201065552,
+            "step": 19860
+        },
+        {
+            "loss": 13.8051,
+            "grad_norm": 1.8618606328964233,
+            "learning_rate": 0.0005,
+            "epoch": 0.8903131109978207,
+            "step": 19865
+        },
+        {
+            "loss": 13.9258,
+            "grad_norm": 1.9716463088989258,
+            "learning_rate": 0.0005,
+            "epoch": 0.8905372018890862,
+            "step": 19870
+        },
+        {
+            "loss": 13.8961,
+            "grad_norm": 1.9195139408111572,
+            "learning_rate": 0.0005,
+            "epoch": 0.8907612927803518,
+            "step": 19875
+        },
+        {
+            "loss": 13.8617,
+            "grad_norm": 2.006978750228882,
+            "learning_rate": 0.0005,
+            "epoch": 0.8909853836716172,
+            "step": 19880
+        },
+        {
+            "loss": 13.8008,
+            "grad_norm": 1.9325141906738281,
+            "learning_rate": 0.0005,
+            "epoch": 0.8912094745628827,
+            "step": 19885
+        },
+        {
+            "loss": 13.9499,
+            "grad_norm": 2.0363314151763916,
+            "learning_rate": 0.0005,
+            "epoch": 0.8914335654541482,
+            "step": 19890
+        },
+        {
+            "loss": 13.9364,
+            "grad_norm": 1.7569103240966797,
+            "learning_rate": 0.0005,
+            "epoch": 0.8916576563454137,
+            "step": 19895
+        },
+        {
+            "loss": 13.8901,
+            "grad_norm": 1.6593209505081177,
+            "learning_rate": 0.0005,
+            "epoch": 0.8918817472366792,
+            "step": 19900
+        },
+        {
+            "loss": 13.879,
+            "grad_norm": 1.7446542978286743,
+            "learning_rate": 0.0005,
+            "epoch": 0.8921058381279446,
+            "step": 19905
+        },
+        {
+            "loss": 13.8354,
+            "grad_norm": 1.8931866884231567,
+            "learning_rate": 0.0005,
+            "epoch": 0.8923299290192102,
+            "step": 19910
+        },
+        {
+            "loss": 13.7994,
+            "grad_norm": 1.8957358598709106,
+            "learning_rate": 0.0005,
+            "epoch": 0.8925540199104757,
+            "step": 19915
+        },
+        {
+            "loss": 13.7895,
+            "grad_norm": 1.6988424062728882,
+            "learning_rate": 0.0005,
+            "epoch": 0.8927781108017412,
+            "step": 19920
+        },
+        {
+            "loss": 13.8562,
+            "grad_norm": 1.9120550155639648,
+            "learning_rate": 0.0005,
+            "epoch": 0.8930022016930067,
+            "step": 19925
+        },
+        {
+            "loss": 13.9294,
+            "grad_norm": 1.809348702430725,
+            "learning_rate": 0.0005,
+            "epoch": 0.8932262925842722,
+            "step": 19930
+        },
+        {
+            "loss": 13.8896,
+            "grad_norm": 1.9009032249450684,
+            "learning_rate": 0.0005,
+            "epoch": 0.8934503834755376,
+            "step": 19935
+        },
+        {
+            "loss": 13.8517,
+            "grad_norm": 1.8128573894500732,
+            "learning_rate": 0.0005,
+            "epoch": 0.8936744743668031,
+            "step": 19940
+        },
+        {
+            "loss": 13.8585,
+            "grad_norm": 1.8379244804382324,
+            "learning_rate": 0.0005,
+            "epoch": 0.8938985652580687,
+            "step": 19945
+        },
+        {
+            "loss": 13.8022,
+            "grad_norm": 1.6590509414672852,
+            "learning_rate": 0.0005,
+            "epoch": 0.8941226561493342,
+            "step": 19950
+        },
+        {
+            "loss": 13.8558,
+            "grad_norm": 1.6452158689498901,
+            "learning_rate": 0.0005,
+            "epoch": 0.8943467470405997,
+            "step": 19955
+        },
+        {
+            "loss": 13.914,
+            "grad_norm": 1.834581971168518,
+            "learning_rate": 0.0005,
+            "epoch": 0.8945708379318652,
+            "step": 19960
+        },
+        {
+            "loss": 13.9026,
+            "grad_norm": 1.8982086181640625,
+            "learning_rate": 0.0005,
+            "epoch": 0.8947949288231306,
+            "step": 19965
+        },
+        {
+            "loss": 13.9778,
+            "grad_norm": 1.9215788841247559,
+            "learning_rate": 0.0005,
+            "epoch": 0.8950190197143961,
+            "step": 19970
+        },
+        {
+            "loss": 13.8827,
+            "grad_norm": 1.979630470275879,
+            "learning_rate": 0.0005,
+            "epoch": 0.8952431106056616,
+            "step": 19975
+        },
+        {
+            "loss": 13.9766,
+            "grad_norm": 1.9976118803024292,
+            "learning_rate": 0.0005,
+            "epoch": 0.8954672014969272,
+            "step": 19980
+        },
+        {
+            "loss": 13.8556,
+            "grad_norm": 1.787476658821106,
+            "learning_rate": 0.0005,
+            "epoch": 0.8956912923881927,
+            "step": 19985
+        },
+        {
+            "loss": 13.8811,
+            "grad_norm": 1.778098464012146,
+            "learning_rate": 0.0005,
+            "epoch": 0.8959153832794582,
+            "step": 19990
+        },
+        {
+            "loss": 13.9134,
+            "grad_norm": 2.0431041717529297,
+            "learning_rate": 0.0005,
+            "epoch": 0.8961394741707236,
+            "step": 19995
+        },
+        {
+            "loss": 13.8592,
+            "grad_norm": 1.6326873302459717,
+            "learning_rate": 0.0005,
+            "epoch": 0.8963635650619891,
+            "step": 20000
+        },
+        {
+            "eval_loss": 1.7306761741638184,
+            "eval_runtime": 18.5601,
+            "eval_samples_per_second": 882.752,
+            "eval_steps_per_second": 7.92,
+            "epoch": 0.8963635650619891,
+            "step": 20000
+        },
+        {
+            "loss": 13.8888,
+            "grad_norm": 1.5741008520126343,
+            "learning_rate": 0.0005,
+            "epoch": 0.8965876559532546,
+            "step": 20005
+        },
+        {
+            "loss": 13.813,
+            "grad_norm": 1.871200442314148,
+            "learning_rate": 0.0005,
+            "epoch": 0.8968117468445201,
+            "step": 20010
+        },
+        {
+            "loss": 13.879,
+            "grad_norm": 2.068232536315918,
+            "learning_rate": 0.0005,
+            "epoch": 0.8970358377357857,
+            "step": 20015
+        },
+        {
+            "loss": 14.0274,
+            "grad_norm": 1.8000730276107788,
+            "learning_rate": 0.0005,
+            "epoch": 0.8972599286270512,
+            "step": 20020
+        },
+        {
+            "loss": 13.8916,
+            "grad_norm": 1.913434624671936,
+            "learning_rate": 0.0005,
+            "epoch": 0.8974840195183166,
+            "step": 20025
+        },
+        {
+            "loss": 13.8162,
+            "grad_norm": 1.6789735555648804,
+            "learning_rate": 0.0005,
+            "epoch": 0.8977081104095821,
+            "step": 20030
+        },
+        {
+            "loss": 13.8043,
+            "grad_norm": 1.673842191696167,
+            "learning_rate": 0.0005,
+            "epoch": 0.8979322013008476,
+            "step": 20035
+        },
+        {
+            "loss": 13.9092,
+            "grad_norm": 1.6753332614898682,
+            "learning_rate": 0.0005,
+            "epoch": 0.8981562921921131,
+            "step": 20040
+        },
+        {
+            "loss": 13.8253,
+            "grad_norm": 1.6415342092514038,
+            "learning_rate": 0.0005,
+            "epoch": 0.8983803830833786,
+            "step": 20045
+        },
+        {
+            "loss": 13.8676,
+            "grad_norm": 1.6949872970581055,
+            "learning_rate": 0.0005,
+            "epoch": 0.8986044739746442,
+            "step": 20050
+        },
+        {
+            "loss": 13.8147,
+            "grad_norm": 1.6152931451797485,
+            "learning_rate": 0.0005,
+            "epoch": 0.8988285648659096,
+            "step": 20055
+        },
+        {
+            "loss": 13.9099,
+            "grad_norm": 1.5995293855667114,
+            "learning_rate": 0.0005,
+            "epoch": 0.8990526557571751,
+            "step": 20060
+        },
+        {
+            "loss": 13.9379,
+            "grad_norm": 1.7555081844329834,
+            "learning_rate": 0.0005,
+            "epoch": 0.8992767466484406,
+            "step": 20065
+        },
+        {
+            "loss": 13.8903,
+            "grad_norm": 1.837789535522461,
+            "learning_rate": 0.0005,
+            "epoch": 0.8995008375397061,
+            "step": 20070
+        },
+        {
+            "loss": 13.9749,
+            "grad_norm": 1.7679381370544434,
+            "learning_rate": 0.0005,
+            "epoch": 0.8997249284309716,
+            "step": 20075
+        },
+        {
+            "loss": 13.992,
+            "grad_norm": 1.783341884613037,
+            "learning_rate": 0.0005,
+            "epoch": 0.8999490193222371,
+            "step": 20080
+        },
+        {
+            "loss": 13.8361,
+            "grad_norm": 1.754961371421814,
+            "learning_rate": 0.0004999977712403221,
+            "epoch": 0.9001731102135025,
+            "step": 20085
+        },
+        {
+            "loss": 13.9087,
+            "grad_norm": 1.792695164680481,
+            "learning_rate": 0.000499984151186201,
+            "epoch": 0.9003972011047681,
+            "step": 20090
+        },
+        {
+            "loss": 13.8374,
+            "grad_norm": 1.7169972658157349,
+            "learning_rate": 0.0004999581499515344,
+            "epoch": 0.9006212919960336,
+            "step": 20095
+        },
+        {
+            "loss": 13.9435,
+            "grad_norm": 1.7526806592941284,
+            "learning_rate": 0.0004999197688241076,
+            "epoch": 0.9008453828872991,
+            "step": 20100
+        },
+        {
+            "loss": 13.8647,
+            "grad_norm": 1.7318717241287231,
+            "learning_rate": 0.0004998690097048561,
+            "epoch": 0.9010694737785646,
+            "step": 20105
+        },
+        {
+            "loss": 13.9183,
+            "grad_norm": 1.895723581314087,
+            "learning_rate": 0.0004998058751077704,
+            "epoch": 0.9012935646698301,
+            "step": 20110
+        },
+        {
+            "loss": 13.9082,
+            "grad_norm": 1.8175562620162964,
+            "learning_rate": 0.0004997303681597721,
+            "epoch": 0.9015176555610955,
+            "step": 20115
+        },
+        {
+            "loss": 13.8523,
+            "grad_norm": 1.7849726676940918,
+            "learning_rate": 0.000499642492600559,
+            "epoch": 0.901741746452361,
+            "step": 20120
+        },
+        {
+            "loss": 13.8472,
+            "grad_norm": 1.8542882204055786,
+            "learning_rate": 0.0004995422527824195,
+            "epoch": 0.9019658373436266,
+            "step": 20125
+        },
+        {
+            "loss": 13.9467,
+            "grad_norm": 1.7311253547668457,
+            "learning_rate": 0.0004994296536700177,
+            "epoch": 0.9021899282348921,
+            "step": 20130
+        },
+        {
+            "loss": 13.9977,
+            "grad_norm": 1.6828018426895142,
+            "learning_rate": 0.0004993047008401468,
+            "epoch": 0.9024140191261576,
+            "step": 20135
+        },
+        {
+            "loss": 13.8359,
+            "grad_norm": 1.7294998168945312,
+            "learning_rate": 0.0004991674004814531,
+            "epoch": 0.9026381100174231,
+            "step": 20140
+        },
+        {
+            "loss": 13.9296,
+            "grad_norm": 1.7701478004455566,
+            "learning_rate": 0.0004990177593941303,
+            "epoch": 0.9028622009086885,
+            "step": 20145
+        },
+        {
+            "loss": 13.8692,
+            "grad_norm": 1.8345767259597778,
+            "learning_rate": 0.000498855784989581,
+            "epoch": 0.903086291799954,
+            "step": 20150
+        },
+        {
+            "loss": 13.8737,
+            "grad_norm": 1.6233254671096802,
+            "learning_rate": 0.0004986814852900517,
+            "epoch": 0.9033103826912195,
+            "step": 20155
+        },
+        {
+            "loss": 13.8891,
+            "grad_norm": 1.671451449394226,
+            "learning_rate": 0.0004984948689282333,
+            "epoch": 0.9035344735824851,
+            "step": 20160
+        },
+        {
+            "loss": 13.8339,
+            "grad_norm": 1.6069334745407104,
+            "learning_rate": 0.0004982959451468356,
+            "epoch": 0.9037585644737506,
+            "step": 20165
+        },
+        {
+            "loss": 13.898,
+            "grad_norm": 1.6981641054153442,
+            "learning_rate": 0.0004980847237981281,
+            "epoch": 0.9039826553650161,
+            "step": 20170
+        },
+        {
+            "loss": 13.842,
+            "grad_norm": 1.7139631509780884,
+            "learning_rate": 0.0004978612153434526,
+            "epoch": 0.9042067462562815,
+            "step": 20175
+        },
+        {
+            "loss": 13.9396,
+            "grad_norm": 1.597270131111145,
+            "learning_rate": 0.000497625430852705,
+            "epoch": 0.904430837147547,
+            "step": 20180
+        },
+        {
+            "loss": 13.8921,
+            "grad_norm": 1.760495901107788,
+            "learning_rate": 0.000497377382003787,
+            "epoch": 0.9046549280388125,
+            "step": 20185
+        },
+        {
+            "loss": 13.8138,
+            "grad_norm": 1.900946021080017,
+            "learning_rate": 0.0004971170810820279,
+            "epoch": 0.904879018930078,
+            "step": 20190
+        },
+        {
+            "loss": 13.9918,
+            "grad_norm": 1.8058969974517822,
+            "learning_rate": 0.0004968445409795756,
+            "epoch": 0.9051031098213436,
+            "step": 20195
+        },
+        {
+            "loss": 13.8749,
+            "grad_norm": 1.6940749883651733,
+            "learning_rate": 0.0004965597751947589,
+            "epoch": 0.9053272007126091,
+            "step": 20200
+        },
+        {
+            "loss": 13.9152,
+            "grad_norm": 1.9464482069015503,
+            "learning_rate": 0.0004962627978314181,
+            "epoch": 0.9055512916038745,
+            "step": 20205
+        },
+        {
+            "loss": 13.9305,
+            "grad_norm": 1.7196178436279297,
+            "learning_rate": 0.0004959536235982073,
+            "epoch": 0.90577538249514,
+            "step": 20210
+        },
+        {
+            "loss": 13.9398,
+            "grad_norm": 1.7908014059066772,
+            "learning_rate": 0.000495632267807865,
+            "epoch": 0.9059994733864055,
+            "step": 20215
+        },
+        {
+            "loss": 13.9284,
+            "grad_norm": 1.6106112003326416,
+            "learning_rate": 0.0004952987463764568,
+            "epoch": 0.906223564277671,
+            "step": 20220
+        },
+        {
+            "loss": 13.8344,
+            "grad_norm": 1.6919456720352173,
+            "learning_rate": 0.0004949530758225857,
+            "epoch": 0.9064476551689366,
+            "step": 20225
+        },
+        {
+            "loss": 13.9635,
+            "grad_norm": 1.7031246423721313,
+            "learning_rate": 0.0004945952732665755,
+            "epoch": 0.9066717460602021,
+            "step": 20230
+        },
+        {
+            "loss": 13.7976,
+            "grad_norm": 1.6084965467453003,
+            "learning_rate": 0.0004942253564296218,
+            "epoch": 0.9068958369514675,
+            "step": 20235
+        },
+        {
+            "loss": 13.8327,
+            "grad_norm": 1.846540093421936,
+            "learning_rate": 0.0004938433436329145,
+            "epoch": 0.907119927842733,
+            "step": 20240
+        },
+        {
+            "loss": 13.8758,
+            "grad_norm": 1.7569811344146729,
+            "learning_rate": 0.0004934492537967308,
+            "epoch": 0.9073440187339985,
+            "step": 20245
+        },
+        {
+            "loss": 13.9523,
+            "grad_norm": 1.7285062074661255,
+            "learning_rate": 0.0004930431064394977,
+            "epoch": 0.907568109625264,
+            "step": 20250
+        },
+        {
+            "loss": 13.9325,
+            "grad_norm": 1.881080985069275,
+            "learning_rate": 0.0004926249216768255,
+            "epoch": 0.9077922005165295,
+            "step": 20255
+        },
+        {
+            "loss": 13.9285,
+            "grad_norm": 1.8552517890930176,
+            "learning_rate": 0.0004921947202205112,
+            "epoch": 0.908016291407795,
+            "step": 20260
+        },
+        {
+            "loss": 13.9245,
+            "grad_norm": 1.9606412649154663,
+            "learning_rate": 0.0004917525233775137,
+            "epoch": 0.9082403822990605,
+            "step": 20265
+        },
+        {
+            "loss": 13.8919,
+            "grad_norm": 1.70395827293396,
+            "learning_rate": 0.0004912983530488966,
+            "epoch": 0.908464473190326,
+            "step": 20270
+        },
+        {
+            "loss": 13.8468,
+            "grad_norm": 1.7387700080871582,
+            "learning_rate": 0.0004908322317287456,
+            "epoch": 0.9086885640815915,
+            "step": 20275
+        },
+        {
+            "loss": 13.8912,
+            "grad_norm": 1.8163692951202393,
+            "learning_rate": 0.0004903541825030532,
+            "epoch": 0.908912654972857,
+            "step": 20280
+        },
+        {
+            "loss": 13.8181,
+            "grad_norm": 1.732064127922058,
+            "learning_rate": 0.0004898642290485751,
+            "epoch": 0.9091367458641225,
+            "step": 20285
+        },
+        {
+            "loss": 13.8886,
+            "grad_norm": 1.7527867555618286,
+            "learning_rate": 0.0004893623956316589,
+            "epoch": 0.909360836755388,
+            "step": 20290
+        },
+        {
+            "loss": 14.0081,
+            "grad_norm": 1.914961338043213,
+            "learning_rate": 0.0004888487071070405,
+            "epoch": 0.9095849276466534,
+            "step": 20295
+        },
+        {
+            "loss": 13.8787,
+            "grad_norm": 1.7557860612869263,
+            "learning_rate": 0.0004883231889166143,
+            "epoch": 0.909809018537919,
+            "step": 20300
+        },
+        {
+            "loss": 13.8582,
+            "grad_norm": 1.670168399810791,
+            "learning_rate": 0.00048778586708817277,
+            "epoch": 0.9100331094291845,
+            "step": 20305
+        },
+        {
+            "loss": 13.8668,
+            "grad_norm": 1.6830042600631714,
+            "learning_rate": 0.00048723676823411727,
+            "epoch": 0.91025720032045,
+            "step": 20310
+        },
+        {
+            "loss": 13.8074,
+            "grad_norm": 1.8048717975616455,
+            "learning_rate": 0.00048667591955014013,
+            "epoch": 0.9104812912117155,
+            "step": 20315
+        },
+        {
+            "loss": 13.9116,
+            "grad_norm": 1.7091537714004517,
+            "learning_rate": 0.0004861033488138774,
+            "epoch": 0.910705382102981,
+            "step": 20320
+        },
+        {
+            "loss": 13.9951,
+            "grad_norm": 1.7124671936035156,
+            "learning_rate": 0.00048551908438353375,
+            "epoch": 0.9109294729942464,
+            "step": 20325
+        },
+        {
+            "loss": 13.8583,
+            "grad_norm": 1.6607993841171265,
+            "learning_rate": 0.0004849231551964771,
+            "epoch": 0.9111535638855119,
+            "step": 20330
+        },
+        {
+            "loss": 13.8799,
+            "grad_norm": 1.7443516254425049,
+            "learning_rate": 0.00048431559076780607,
+            "epoch": 0.9113776547767775,
+            "step": 20335
+        },
+        {
+            "loss": 13.8942,
+            "grad_norm": 1.738348126411438,
+            "learning_rate": 0.0004836964211888878,
+            "epoch": 0.911601745668043,
+            "step": 20340
+        },
+        {
+            "loss": 13.791,
+            "grad_norm": 1.6687777042388916,
+            "learning_rate": 0.0004830656771258677,
+            "epoch": 0.9118258365593085,
+            "step": 20345
+        },
+        {
+            "loss": 13.852,
+            "grad_norm": 1.673667550086975,
+            "learning_rate": 0.00048242338981815085,
+            "epoch": 0.912049927450574,
+            "step": 20350
+        },
+        {
+            "loss": 13.8364,
+            "grad_norm": 1.7567943334579468,
+            "learning_rate": 0.00048176959107685435,
+            "epoch": 0.9122740183418394,
+            "step": 20355
+        },
+        {
+            "loss": 13.8419,
+            "grad_norm": 1.7680989503860474,
+            "learning_rate": 0.000481104313283232,
+            "epoch": 0.9124981092331049,
+            "step": 20360
+        },
+        {
+            "loss": 13.7604,
+            "grad_norm": 1.6283504962921143,
+            "learning_rate": 0.0004804275893870704,
+            "epoch": 0.9127222001243704,
+            "step": 20365
+        },
+        {
+            "loss": 13.8656,
+            "grad_norm": 1.7220295667648315,
+            "learning_rate": 0.00047973945290505766,
+            "epoch": 0.912946291015636,
+            "step": 20370
+        },
+        {
+            "loss": 13.8799,
+            "grad_norm": 1.7215722799301147,
+            "learning_rate": 0.00047903993791912226,
+            "epoch": 0.9131703819069015,
+            "step": 20375
+        },
+        {
+            "loss": 13.8281,
+            "grad_norm": 1.8215144872665405,
+            "learning_rate": 0.000478329079074746,
+            "epoch": 0.913394472798167,
+            "step": 20380
+        },
+        {
+            "loss": 13.8491,
+            "grad_norm": 1.7994507551193237,
+            "learning_rate": 0.0004776069115792475,
+            "epoch": 0.9136185636894324,
+            "step": 20385
+        },
+        {
+            "loss": 13.8475,
+            "grad_norm": 1.7810310125350952,
+            "learning_rate": 0.000476873471200039,
+            "epoch": 0.9138426545806979,
+            "step": 20390
+        },
+        {
+            "loss": 13.8628,
+            "grad_norm": 1.8750712871551514,
+            "learning_rate": 0.00047612879426285425,
+            "epoch": 0.9140667454719634,
+            "step": 20395
+        },
+        {
+            "loss": 14.005,
+            "grad_norm": 1.9540858268737793,
+            "learning_rate": 0.00047537291764995006,
+            "epoch": 0.9142908363632289,
+            "step": 20400
+        },
+        {
+            "loss": 13.766,
+            "grad_norm": 1.9318475723266602,
+            "learning_rate": 0.0004746058787982788,
+            "epoch": 0.9145149272544945,
+            "step": 20405
+        },
+        {
+            "loss": 13.8293,
+            "grad_norm": 1.6670743227005005,
+            "learning_rate": 0.00047382771569763485,
+            "epoch": 0.91473901814576,
+            "step": 20410
+        },
+        {
+            "loss": 13.9367,
+            "grad_norm": 1.6447280645370483,
+            "learning_rate": 0.000473038466888773,
+            "epoch": 0.9149631090370254,
+            "step": 20415
+        },
+        {
+            "loss": 13.784,
+            "grad_norm": 1.9621039628982544,
+            "learning_rate": 0.0004722381714614994,
+            "epoch": 0.9151871999282909,
+            "step": 20420
+        },
+        {
+            "loss": 13.9344,
+            "grad_norm": 1.5999281406402588,
+            "learning_rate": 0.00047142686905273537,
+            "epoch": 0.9154112908195564,
+            "step": 20425
+        },
+        {
+            "loss": 13.8596,
+            "grad_norm": 1.761801838874817,
+            "learning_rate": 0.0004706045998445548,
+            "epoch": 0.9156353817108219,
+            "step": 20430
+        },
+        {
+            "loss": 13.7922,
+            "grad_norm": 1.7019723653793335,
+            "learning_rate": 0.0004697714045621935,
+            "epoch": 0.9158594726020874,
+            "step": 20435
+        },
+        {
+            "loss": 13.7741,
+            "grad_norm": 1.755383014678955,
+            "learning_rate": 0.0004689273244720325,
+            "epoch": 0.916083563493353,
+            "step": 20440
+        },
+        {
+            "loss": 13.9154,
+            "grad_norm": 1.742811918258667,
+            "learning_rate": 0.000468072401379554,
+            "epoch": 0.9163076543846184,
+            "step": 20445
+        },
+        {
+            "loss": 13.8688,
+            "grad_norm": 1.6975293159484863,
+            "learning_rate": 0.000467206677627271,
+            "epoch": 0.9165317452758839,
+            "step": 20450
+        },
+        {
+            "loss": 13.817,
+            "grad_norm": 1.7448455095291138,
+            "learning_rate": 0.00046633019609262997,
+            "epoch": 0.9167558361671494,
+            "step": 20455
+        },
+        {
+            "loss": 13.8818,
+            "grad_norm": 1.705538034439087,
+            "learning_rate": 0.00046544300018588745,
+            "epoch": 0.9169799270584149,
+            "step": 20460
+        },
+        {
+            "loss": 13.8366,
+            "grad_norm": 1.7605187892913818,
+            "learning_rate": 0.00046454513384795986,
+            "epoch": 0.9172040179496804,
+            "step": 20465
+        },
+        {
+            "loss": 13.932,
+            "grad_norm": 1.904154896736145,
+            "learning_rate": 0.0004636366415482474,
+            "epoch": 0.9174281088409458,
+            "step": 20470
+        },
+        {
+            "loss": 13.7937,
+            "grad_norm": 1.7890197038650513,
+            "learning_rate": 0.00046271756828243117,
+            "epoch": 0.9176521997322113,
+            "step": 20475
+        },
+        {
+            "loss": 13.8559,
+            "grad_norm": 1.8883311748504639,
+            "learning_rate": 0.0004617879595702452,
+            "epoch": 0.9178762906234769,
+            "step": 20480
+        },
+        {
+            "loss": 14.0243,
+            "grad_norm": 1.725429654121399,
+            "learning_rate": 0.00046084786145322143,
+            "epoch": 0.9181003815147424,
+            "step": 20485
+        },
+        {
+            "loss": 13.8569,
+            "grad_norm": 1.7402434349060059,
+            "learning_rate": 0.00045989732049240976,
+            "epoch": 0.9183244724060079,
+            "step": 20490
+        },
+        {
+            "loss": 13.85,
+            "grad_norm": 1.7885849475860596,
+            "learning_rate": 0.0004589363837660716,
+            "epoch": 0.9185485632972734,
+            "step": 20495
+        },
+        {
+            "loss": 13.8368,
+            "grad_norm": 1.859695315361023,
+            "learning_rate": 0.0004579650988673487,
+            "epoch": 0.9187726541885388,
+            "step": 20500
+        },
+        {
+            "eval_loss": 1.7257344722747803,
+            "eval_runtime": 18.4835,
+            "eval_samples_per_second": 886.41,
+            "eval_steps_per_second": 7.953,
+            "epoch": 0.9187726541885388,
+            "step": 20500
+        },
+        {
+            "loss": 13.8213,
+            "grad_norm": 1.6608607769012451,
+            "learning_rate": 0.0004569835139019054,
+            "epoch": 0.9189967450798043,
+            "step": 20505
+        },
+        {
+            "loss": 13.9715,
+            "grad_norm": 1.6903189420700073,
+            "learning_rate": 0.0004559916774855464,
+            "epoch": 0.9192208359710698,
+            "step": 20510
+        },
+        {
+            "loss": 13.8817,
+            "grad_norm": 1.82123601436615,
+            "learning_rate": 0.0004549896387418089,
+            "epoch": 0.9194449268623354,
+            "step": 20515
+        },
+        {
+            "loss": 13.7962,
+            "grad_norm": 1.8259592056274414,
+            "learning_rate": 0.0004539774472995296,
+            "epoch": 0.9196690177536009,
+            "step": 20520
+        },
+        {
+            "loss": 13.783,
+            "grad_norm": 1.7011497020721436,
+            "learning_rate": 0.0004529551532903865,
+            "epoch": 0.9198931086448664,
+            "step": 20525
+        },
+        {
+            "loss": 13.9082,
+            "grad_norm": 1.7713634967803955,
+            "learning_rate": 0.00045192280734641623,
+            "epoch": 0.9201171995361318,
+            "step": 20530
+        },
+        {
+            "loss": 13.8889,
+            "grad_norm": 1.750898003578186,
+            "learning_rate": 0.00045088046059750634,
+            "epoch": 0.9203412904273973,
+            "step": 20535
+        },
+        {
+            "loss": 13.7703,
+            "grad_norm": 1.7162286043167114,
+            "learning_rate": 0.0004498281646688627,
+            "epoch": 0.9205653813186628,
+            "step": 20540
+        },
+        {
+            "loss": 13.9167,
+            "grad_norm": 1.7430411577224731,
+            "learning_rate": 0.00044876597167845276,
+            "epoch": 0.9207894722099284,
+            "step": 20545
+        },
+        {
+            "loss": 13.921,
+            "grad_norm": 1.7012724876403809,
+            "learning_rate": 0.0004476939342344246,
+            "epoch": 0.9210135631011939,
+            "step": 20550
+        },
+        {
+            "loss": 13.935,
+            "grad_norm": 1.8371593952178955,
+            "learning_rate": 0.00044661210543250077,
+            "epoch": 0.9212376539924594,
+            "step": 20555
+        },
+        {
+            "loss": 13.8992,
+            "grad_norm": 2.012096881866455,
+            "learning_rate": 0.00044552053885334875,
+            "epoch": 0.9214617448837248,
+            "step": 20560
+        },
+        {
+            "loss": 13.7958,
+            "grad_norm": 1.8178032636642456,
+            "learning_rate": 0.0004444192885599276,
+            "epoch": 0.9216858357749903,
+            "step": 20565
+        },
+        {
+            "loss": 13.7693,
+            "grad_norm": 1.6493083238601685,
+            "learning_rate": 0.00044330840909480984,
+            "epoch": 0.9219099266662558,
+            "step": 20570
+        },
+        {
+            "loss": 13.8159,
+            "grad_norm": 1.667523980140686,
+            "learning_rate": 0.0004421879554774803,
+            "epoch": 0.9221340175575213,
+            "step": 20575
+        },
+        {
+            "loss": 13.8814,
+            "grad_norm": 1.6737457513809204,
+            "learning_rate": 0.0004410579832016112,
+            "epoch": 0.9223581084487869,
+            "step": 20580
+        },
+        {
+            "loss": 13.8307,
+            "grad_norm": 1.687098741531372,
+            "learning_rate": 0.0004399185482323134,
+            "epoch": 0.9225821993400524,
+            "step": 20585
+        },
+        {
+            "loss": 13.722,
+            "grad_norm": 1.6777094602584839,
+            "learning_rate": 0.00043876970700336496,
+            "epoch": 0.9228062902313178,
+            "step": 20590
+        },
+        {
+            "loss": 13.8917,
+            "grad_norm": 1.747768521308899,
+            "learning_rate": 0.00043761151641441565,
+            "epoch": 0.9230303811225833,
+            "step": 20595
+        },
+        {
+            "loss": 13.8852,
+            "grad_norm": 1.6435922384262085,
+            "learning_rate": 0.00043644403382816913,
+            "epoch": 0.9232544720138488,
+            "step": 20600
+        },
+        {
+            "loss": 13.8821,
+            "grad_norm": 1.6521140336990356,
+            "learning_rate": 0.00043526731706754196,
+            "epoch": 0.9234785629051143,
+            "step": 20605
+        },
+        {
+            "loss": 13.8213,
+            "grad_norm": 1.8001853227615356,
+            "learning_rate": 0.0004340814244127993,
+            "epoch": 0.9237026537963798,
+            "step": 20610
+        },
+        {
+            "loss": 13.8501,
+            "grad_norm": 1.6893656253814697,
+            "learning_rate": 0.00043288641459866915,
+            "epoch": 0.9239267446876454,
+            "step": 20615
+        },
+        {
+            "loss": 13.7712,
+            "grad_norm": 1.6274510622024536,
+            "learning_rate": 0.00043168234681143246,
+            "epoch": 0.9241508355789108,
+            "step": 20620
+        },
+        {
+            "loss": 13.8701,
+            "grad_norm": 1.8652204275131226,
+            "learning_rate": 0.0004304692806859927,
+            "epoch": 0.9243749264701763,
+            "step": 20625
+        },
+        {
+            "loss": 13.8061,
+            "grad_norm": 1.9463926553726196,
+            "learning_rate": 0.00042924727630292125,
+            "epoch": 0.9245990173614418,
+            "step": 20630
+        },
+        {
+            "loss": 13.8314,
+            "grad_norm": 1.8704006671905518,
+            "learning_rate": 0.0004280163941854828,
+            "epoch": 0.9248231082527073,
+            "step": 20635
+        },
+        {
+            "loss": 13.7852,
+            "grad_norm": 1.7580029964447021,
+            "learning_rate": 0.00042677669529663686,
+            "epoch": 0.9250471991439728,
+            "step": 20640
+        },
+        {
+            "loss": 13.8812,
+            "grad_norm": 1.6548473834991455,
+            "learning_rate": 0.00042552824103601916,
+            "epoch": 0.9252712900352383,
+            "step": 20645
+        },
+        {
+            "loss": 13.7267,
+            "grad_norm": 1.6016569137573242,
+            "learning_rate": 0.0004242710932368998,
+            "epoch": 0.9254953809265037,
+            "step": 20650
+        },
+        {
+            "loss": 13.7553,
+            "grad_norm": 1.7345774173736572,
+            "learning_rate": 0.0004230053141631216,
+            "epoch": 0.9257194718177693,
+            "step": 20655
+        },
+        {
+            "loss": 13.8659,
+            "grad_norm": 1.749651551246643,
+            "learning_rate": 0.00042173096650601594,
+            "epoch": 0.9259435627090348,
+            "step": 20660
+        },
+        {
+            "loss": 13.7555,
+            "grad_norm": 1.6441956758499146,
+            "learning_rate": 0.0004204481133812977,
+            "epoch": 0.9261676536003003,
+            "step": 20665
+        },
+        {
+            "loss": 13.908,
+            "grad_norm": 1.6786205768585205,
+            "learning_rate": 0.00041915681832593936,
+            "epoch": 0.9263917444915658,
+            "step": 20670
+        },
+        {
+            "loss": 13.8288,
+            "grad_norm": 1.71892249584198,
+            "learning_rate": 0.00041785714529502427,
+            "epoch": 0.9266158353828313,
+            "step": 20675
+        },
+        {
+            "loss": 13.7958,
+            "grad_norm": 1.9180169105529785,
+            "learning_rate": 0.000416549158658579,
+            "epoch": 0.9268399262740967,
+            "step": 20680
+        },
+        {
+            "loss": 13.8891,
+            "grad_norm": 1.7008872032165527,
+            "learning_rate": 0.0004152329231983852,
+            "epoch": 0.9270640171653622,
+            "step": 20685
+        },
+        {
+            "loss": 13.8643,
+            "grad_norm": 1.6335214376449585,
+            "learning_rate": 0.0004139085041047711,
+            "epoch": 0.9272881080566278,
+            "step": 20690
+        },
+        {
+            "loss": 13.6942,
+            "grad_norm": 1.766997218132019,
+            "learning_rate": 0.00041257596697338286,
+            "epoch": 0.9275121989478933,
+            "step": 20695
+        },
+        {
+            "loss": 13.7377,
+            "grad_norm": 1.6384992599487305,
+            "learning_rate": 0.00041123537780193554,
+            "epoch": 0.9277362898391588,
+            "step": 20700
+        },
+        {
+            "loss": 13.8826,
+            "grad_norm": 1.7075258493423462,
+            "learning_rate": 0.0004098868029869447,
+            "epoch": 0.9279603807304243,
+            "step": 20705
+        },
+        {
+            "loss": 13.7023,
+            "grad_norm": 1.6997811794281006,
+            "learning_rate": 0.00040853030932043775,
+            "epoch": 0.9281844716216897,
+            "step": 20710
+        },
+        {
+            "loss": 13.7944,
+            "grad_norm": 1.677700161933899,
+            "learning_rate": 0.0004071659639866457,
+            "epoch": 0.9284085625129552,
+            "step": 20715
+        },
+        {
+            "loss": 13.8049,
+            "grad_norm": 1.6065220832824707,
+            "learning_rate": 0.0004057938345586761,
+            "epoch": 0.9286326534042207,
+            "step": 20720
+        },
+        {
+            "loss": 13.842,
+            "grad_norm": 1.6548614501953125,
+            "learning_rate": 0.0004044139889951659,
+            "epoch": 0.9288567442954863,
+            "step": 20725
+        },
+        {
+            "loss": 13.7809,
+            "grad_norm": 1.7440521717071533,
+            "learning_rate": 0.00040302649563691575,
+            "epoch": 0.9290808351867518,
+            "step": 20730
+        },
+        {
+            "loss": 13.8484,
+            "grad_norm": 1.6194367408752441,
+            "learning_rate": 0.00040163142320350523,
+            "epoch": 0.9293049260780173,
+            "step": 20735
+        },
+        {
+            "loss": 13.8731,
+            "grad_norm": 1.6101369857788086,
+            "learning_rate": 0.0004002288407898893,
+            "epoch": 0.9295290169692827,
+            "step": 20740
+        },
+        {
+            "loss": 13.8792,
+            "grad_norm": 1.6936492919921875,
+            "learning_rate": 0.0003988188178629763,
+            "epoch": 0.9297531078605482,
+            "step": 20745
+        },
+        {
+            "loss": 13.7914,
+            "grad_norm": 1.6488006114959717,
+            "learning_rate": 0.00039740142425818715,
+            "epoch": 0.9299771987518137,
+            "step": 20750
+        },
+        {
+            "loss": 13.8232,
+            "grad_norm": 1.5743663311004639,
+            "learning_rate": 0.0003959767301759967,
+            "epoch": 0.9302012896430792,
+            "step": 20755
+        },
+        {
+            "loss": 13.7971,
+            "grad_norm": 1.6571708917617798,
+            "learning_rate": 0.00039454480617845676,
+            "epoch": 0.9304253805343448,
+            "step": 20760
+        },
+        {
+            "loss": 13.8978,
+            "grad_norm": 1.8308016061782837,
+            "learning_rate": 0.0003931057231857017,
+            "epoch": 0.9306494714256103,
+            "step": 20765
+        },
+        {
+            "loss": 13.8654,
+            "grad_norm": 1.6224706172943115,
+            "learning_rate": 0.0003916595524724353,
+            "epoch": 0.9308735623168757,
+            "step": 20770
+        },
+        {
+            "loss": 13.8414,
+            "grad_norm": 1.542962670326233,
+            "learning_rate": 0.00039020636566440114,
+            "epoch": 0.9310976532081412,
+            "step": 20775
+        },
+        {
+            "loss": 13.8781,
+            "grad_norm": 1.9705100059509277,
+            "learning_rate": 0.0003887462347348349,
+            "epoch": 0.9313217440994067,
+            "step": 20780
+        },
+        {
+            "loss": 13.7762,
+            "grad_norm": 1.6651804447174072,
+            "learning_rate": 0.00038727923200089975,
+            "epoch": 0.9315458349906722,
+            "step": 20785
+        },
+        {
+            "loss": 13.7654,
+            "grad_norm": 1.8163079023361206,
+            "learning_rate": 0.0003858054301201047,
+            "epoch": 0.9317699258819377,
+            "step": 20790
+        },
+        {
+            "loss": 13.8023,
+            "grad_norm": 1.91973078250885,
+            "learning_rate": 0.000384324902086706,
+            "epoch": 0.9319940167732033,
+            "step": 20795
+        },
+        {
+            "loss": 13.8323,
+            "grad_norm": 1.7480084896087646,
+            "learning_rate": 0.0003828377212280917,
+            "epoch": 0.9322181076644687,
+            "step": 20800
+        },
+        {
+            "loss": 13.7816,
+            "grad_norm": 1.6364456415176392,
+            "learning_rate": 0.0003813439612011501,
+            "epoch": 0.9324421985557342,
+            "step": 20805
+        },
+        {
+            "loss": 13.6706,
+            "grad_norm": 1.5766396522521973,
+            "learning_rate": 0.0003798436959886219,
+            "epoch": 0.9326662894469997,
+            "step": 20810
+        },
+        {
+            "loss": 13.7765,
+            "grad_norm": 1.6076284646987915,
+            "learning_rate": 0.00037833699989543544,
+            "epoch": 0.9328903803382652,
+            "step": 20815
+        },
+        {
+            "loss": 13.8702,
+            "grad_norm": 1.6237750053405762,
+            "learning_rate": 0.00037682394754502685,
+            "epoch": 0.9331144712295307,
+            "step": 20820
+        },
+        {
+            "loss": 13.822,
+            "grad_norm": 1.6188440322875977,
+            "learning_rate": 0.0003753046138756442,
+            "epoch": 0.9333385621207962,
+            "step": 20825
+        },
+        {
+            "loss": 13.8648,
+            "grad_norm": 1.74038827419281,
+            "learning_rate": 0.0003737790741366358,
+            "epoch": 0.9335626530120616,
+            "step": 20830
+        },
+        {
+            "loss": 13.7871,
+            "grad_norm": 1.7488499879837036,
+            "learning_rate": 0.0003722474038847235,
+            "epoch": 0.9337867439033272,
+            "step": 20835
+        },
+        {
+            "loss": 13.7595,
+            "grad_norm": 1.6546235084533691,
+            "learning_rate": 0.0003707096789802599,
+            "epoch": 0.9340108347945927,
+            "step": 20840
+        },
+        {
+            "loss": 13.8413,
+            "grad_norm": 1.6696339845657349,
+            "learning_rate": 0.00036916597558347215,
+            "epoch": 0.9342349256858582,
+            "step": 20845
+        },
+        {
+            "loss": 13.7561,
+            "grad_norm": 1.7855514287948608,
+            "learning_rate": 0.00036761637015068893,
+            "epoch": 0.9344590165771237,
+            "step": 20850
+        },
+        {
+            "loss": 13.7537,
+            "grad_norm": 1.6255345344543457,
+            "learning_rate": 0.0003660609394305543,
+            "epoch": 0.9346831074683892,
+            "step": 20855
+        },
+        {
+            "loss": 13.9035,
+            "grad_norm": 1.6804163455963135,
+            "learning_rate": 0.00036449976046022643,
+            "epoch": 0.9349071983596546,
+            "step": 20860
+        },
+        {
+            "loss": 13.8235,
+            "grad_norm": 1.679900884628296,
+            "learning_rate": 0.00036293291056156175,
+            "epoch": 0.9351312892509202,
+            "step": 20865
+        },
+        {
+            "loss": 13.819,
+            "grad_norm": 1.7183867692947388,
+            "learning_rate": 0.00036136046733728613,
+            "epoch": 0.9353553801421857,
+            "step": 20870
+        },
+        {
+            "loss": 13.8095,
+            "grad_norm": 1.6135212182998657,
+            "learning_rate": 0.00035978250866715034,
+            "epoch": 0.9355794710334512,
+            "step": 20875
+        },
+        {
+            "loss": 13.7488,
+            "grad_norm": 1.6178854703903198,
+            "learning_rate": 0.00035819911270407374,
+            "epoch": 0.9358035619247167,
+            "step": 20880
+        },
+        {
+            "loss": 13.8237,
+            "grad_norm": 1.6501045227050781,
+            "learning_rate": 0.0003566103578702731,
+            "epoch": 0.9360276528159822,
+            "step": 20885
+        },
+        {
+            "loss": 13.6911,
+            "grad_norm": 1.5672601461410522,
+            "learning_rate": 0.00035501632285337873,
+            "epoch": 0.9362517437072476,
+            "step": 20890
+        },
+        {
+            "loss": 13.835,
+            "grad_norm": 1.591254711151123,
+            "learning_rate": 0.00035341708660253685,
+            "epoch": 0.9364758345985131,
+            "step": 20895
+        },
+        {
+            "loss": 13.8382,
+            "grad_norm": 1.7729851007461548,
+            "learning_rate": 0.00035181272832449984,
+            "epoch": 0.9366999254897787,
+            "step": 20900
+        },
+        {
+            "loss": 13.6892,
+            "grad_norm": 1.6097650527954102,
+            "learning_rate": 0.0003502033274797031,
+            "epoch": 0.9369240163810442,
+            "step": 20905
+        },
+        {
+            "loss": 13.7056,
+            "grad_norm": 1.6839957237243652,
+            "learning_rate": 0.00034858896377832965,
+            "epoch": 0.9371481072723097,
+            "step": 20910
+        },
+        {
+            "loss": 13.7331,
+            "grad_norm": 1.6045210361480713,
+            "learning_rate": 0.00034696971717636217,
+            "epoch": 0.9373721981635752,
+            "step": 20915
+        },
+        {
+            "loss": 13.681,
+            "grad_norm": 1.7205440998077393,
+            "learning_rate": 0.0003453456678716227,
+            "epoch": 0.9375962890548406,
+            "step": 20920
+        },
+        {
+            "loss": 13.7979,
+            "grad_norm": 1.5592955350875854,
+            "learning_rate": 0.0003437168962998014,
+            "epoch": 0.9378203799461061,
+            "step": 20925
+        },
+        {
+            "loss": 13.7619,
+            "grad_norm": 1.6273744106292725,
+            "learning_rate": 0.00034208348313047185,
+            "epoch": 0.9380444708373716,
+            "step": 20930
+        },
+        {
+            "loss": 13.8481,
+            "grad_norm": 1.618740200996399,
+            "learning_rate": 0.0003404455092630959,
+            "epoch": 0.9382685617286372,
+            "step": 20935
+        },
+        {
+            "loss": 13.7558,
+            "grad_norm": 1.6477937698364258,
+            "learning_rate": 0.00033880305582301764,
+            "epoch": 0.9384926526199027,
+            "step": 20940
+        },
+        {
+            "loss": 13.867,
+            "grad_norm": 1.6027733087539673,
+            "learning_rate": 0.0003371562041574439,
+            "epoch": 0.9387167435111682,
+            "step": 20945
+        },
+        {
+            "loss": 13.8095,
+            "grad_norm": 1.5962384939193726,
+            "learning_rate": 0.0003355050358314172,
+            "epoch": 0.9389408344024336,
+            "step": 20950
+        },
+        {
+            "loss": 13.8599,
+            "grad_norm": 1.7121936082839966,
+            "learning_rate": 0.0003338496326237743,
+            "epoch": 0.9391649252936991,
+            "step": 20955
+        },
+        {
+            "loss": 13.8941,
+            "grad_norm": 1.6914751529693604,
+            "learning_rate": 0.0003321900765230969,
+            "epoch": 0.9393890161849646,
+            "step": 20960
+        },
+        {
+            "loss": 13.8342,
+            "grad_norm": 1.5547442436218262,
+            "learning_rate": 0.00033052644972365056,
+            "epoch": 0.9396131070762301,
+            "step": 20965
+        },
+        {
+            "loss": 13.8076,
+            "grad_norm": 1.5029220581054688,
+            "learning_rate": 0.0003288588346213139,
+            "epoch": 0.9398371979674957,
+            "step": 20970
+        },
+        {
+            "loss": 13.8405,
+            "grad_norm": 1.5579488277435303,
+            "learning_rate": 0.00032718731380949754,
+            "epoch": 0.9400612888587612,
+            "step": 20975
+        },
+        {
+            "loss": 13.7782,
+            "grad_norm": 1.647713303565979,
+            "learning_rate": 0.0003255119700750535,
+            "epoch": 0.9402853797500266,
+            "step": 20980
+        },
+        {
+            "loss": 13.7087,
+            "grad_norm": 1.569794774055481,
+            "learning_rate": 0.0003238328863941753,
+            "epoch": 0.9405094706412921,
+            "step": 20985
+        },
+        {
+            "loss": 13.7996,
+            "grad_norm": 1.584929347038269,
+            "learning_rate": 0.0003221501459282877,
+            "epoch": 0.9407335615325576,
+            "step": 20990
+        },
+        {
+            "loss": 13.8525,
+            "grad_norm": 1.613590955734253,
+            "learning_rate": 0.0003204638320199282,
+            "epoch": 0.9409576524238231,
+            "step": 20995
+        },
+        {
+            "loss": 13.6583,
+            "grad_norm": 1.6018569469451904,
+            "learning_rate": 0.0003187740281886195,
+            "epoch": 0.9411817433150886,
+            "step": 21000
+        },
+        {
+            "eval_loss": 1.7149229049682617,
+            "eval_runtime": 18.6557,
+            "eval_samples_per_second": 878.231,
+            "eval_steps_per_second": 7.88,
+            "epoch": 0.9411817433150886,
+            "step": 21000
+        },
+        {
+            "loss": 13.7715,
+            "grad_norm": 1.6226269006729126,
+            "learning_rate": 0.0003170808181267326,
+            "epoch": 0.941405834206354,
+            "step": 21005
+        },
+        {
+            "loss": 13.742,
+            "grad_norm": 1.617499589920044,
+            "learning_rate": 0.0003153842856953417,
+            "epoch": 0.9416299250976196,
+            "step": 21010
+        },
+        {
+            "loss": 13.7314,
+            "grad_norm": 1.5880169868469238,
+            "learning_rate": 0.000313684514920071,
+            "epoch": 0.9418540159888851,
+            "step": 21015
+        },
+        {
+            "loss": 13.7776,
+            "grad_norm": 1.879854679107666,
+            "learning_rate": 0.0003119815899869329,
+            "epoch": 0.9420781068801506,
+            "step": 21020
+        },
+        {
+            "loss": 13.7681,
+            "grad_norm": 1.631612777709961,
+            "learning_rate": 0.0003102755952381586,
+            "epoch": 0.9423021977714161,
+            "step": 21025
+        },
+        {
+            "loss": 13.8383,
+            "grad_norm": 1.8044886589050293,
+            "learning_rate": 0.00030856661516802055,
+            "epoch": 0.9425262886626816,
+            "step": 21030
+        },
+        {
+            "loss": 13.7984,
+            "grad_norm": 1.5461212396621704,
+            "learning_rate": 0.0003068547344186478,
+            "epoch": 0.942750379553947,
+            "step": 21035
+        },
+        {
+            "loss": 13.6981,
+            "grad_norm": 1.5587557554244995,
+            "learning_rate": 0.00030514003777583397,
+            "epoch": 0.9429744704452125,
+            "step": 21040
+        },
+        {
+            "loss": 13.6394,
+            "grad_norm": 1.6465073823928833,
+            "learning_rate": 0.0003034226101648377,
+            "epoch": 0.9431985613364781,
+            "step": 21045
+        },
+        {
+            "loss": 13.7656,
+            "grad_norm": 1.6942849159240723,
+            "learning_rate": 0.00030170253664617687,
+            "epoch": 0.9434226522277436,
+            "step": 21050
+        },
+        {
+            "loss": 13.7423,
+            "grad_norm": 1.5614516735076904,
+            "learning_rate": 0.0002999799024114151,
+            "epoch": 0.9436467431190091,
+            "step": 21055
+        },
+        {
+            "loss": 13.7156,
+            "grad_norm": 1.4851198196411133,
+            "learning_rate": 0.0002982547927789434,
+            "epoch": 0.9438708340102746,
+            "step": 21060
+        },
+        {
+            "loss": 13.7704,
+            "grad_norm": 1.475362777709961,
+            "learning_rate": 0.00029652729318975333,
+            "epoch": 0.94409492490154,
+            "step": 21065
+        },
+        {
+            "loss": 13.7994,
+            "grad_norm": 1.5233135223388672,
+            "learning_rate": 0.00029479748920320634,
+            "epoch": 0.9443190157928055,
+            "step": 21070
+        },
+        {
+            "loss": 13.5917,
+            "grad_norm": 1.5046474933624268,
+            "learning_rate": 0.0002930654664927955,
+            "epoch": 0.944543106684071,
+            "step": 21075
+        },
+        {
+            "loss": 13.5741,
+            "grad_norm": 1.5400524139404297,
+            "learning_rate": 0.00029133131084190265,
+            "epoch": 0.9447671975753366,
+            "step": 21080
+        },
+        {
+            "loss": 13.6754,
+            "grad_norm": 1.6073517799377441,
+            "learning_rate": 0.0002895951081395496,
+            "epoch": 0.9449912884666021,
+            "step": 21085
+        },
+        {
+            "loss": 13.6572,
+            "grad_norm": 1.5722095966339111,
+            "learning_rate": 0.0002878569443761442,
+            "epoch": 0.9452153793578676,
+            "step": 21090
+        },
+        {
+            "loss": 13.7381,
+            "grad_norm": 1.5152595043182373,
+            "learning_rate": 0.00028611690563922144,
+            "epoch": 0.945439470249133,
+            "step": 21095
+        },
+        {
+            "loss": 13.7101,
+            "grad_norm": 1.5220506191253662,
+            "learning_rate": 0.0002843750781091798,
+            "epoch": 0.9456635611403985,
+            "step": 21100
+        },
+        {
+            "loss": 13.7555,
+            "grad_norm": 1.5469439029693604,
+            "learning_rate": 0.000282631548055013,
+            "epoch": 0.945887652031664,
+            "step": 21105
+        },
+        {
+            "loss": 13.7246,
+            "grad_norm": 1.5397659540176392,
+            "learning_rate": 0.0002808864018300367,
+            "epoch": 0.9461117429229295,
+            "step": 21110
+        },
+        {
+            "loss": 13.6633,
+            "grad_norm": 1.6063482761383057,
+            "learning_rate": 0.00027913972586761246,
+            "epoch": 0.9463358338141951,
+            "step": 21115
+        },
+        {
+            "loss": 13.749,
+            "grad_norm": 1.5963630676269531,
+            "learning_rate": 0.00027739160667686634,
+            "epoch": 0.9465599247054606,
+            "step": 21120
+        },
+        {
+            "loss": 13.6779,
+            "grad_norm": 1.4886605739593506,
+            "learning_rate": 0.00027564213083840433,
+            "epoch": 0.946784015596726,
+            "step": 21125
+        },
+        {
+            "loss": 13.7391,
+            "grad_norm": 1.5576682090759277,
+            "learning_rate": 0.0002738913850000246,
+            "epoch": 0.9470081064879915,
+            "step": 21130
+        },
+        {
+            "loss": 13.6428,
+            "grad_norm": 1.5561497211456299,
+            "learning_rate": 0.00027213945587242506,
+            "epoch": 0.947232197379257,
+            "step": 21135
+        },
+        {
+            "loss": 13.7662,
+            "grad_norm": 1.5810502767562866,
+            "learning_rate": 0.0002703864302249102,
+            "epoch": 0.9474562882705225,
+            "step": 21140
+        },
+        {
+            "loss": 13.682,
+            "grad_norm": 1.6845345497131348,
+            "learning_rate": 0.0002686323948810921,
+            "epoch": 0.947680379161788,
+            "step": 21145
+        },
+        {
+            "loss": 13.6574,
+            "grad_norm": 1.6456736326217651,
+            "learning_rate": 0.0002668774367145913,
+            "epoch": 0.9479044700530536,
+            "step": 21150
+        },
+        {
+            "loss": 13.6922,
+            "grad_norm": 1.7522460222244263,
+            "learning_rate": 0.00026512164264473387,
+            "epoch": 0.948128560944319,
+            "step": 21155
+        },
+        {
+            "loss": 13.7328,
+            "grad_norm": 1.5039067268371582,
+            "learning_rate": 0.0002633650996322461,
+            "epoch": 0.9483526518355845,
+            "step": 21160
+        },
+        {
+            "loss": 13.7239,
+            "grad_norm": 1.5195696353912354,
+            "learning_rate": 0.00026160789467494786,
+            "epoch": 0.94857674272685,
+            "step": 21165
+        },
+        {
+            "loss": 13.6447,
+            "grad_norm": 1.589791178703308,
+            "learning_rate": 0.0002598501148034439,
+            "epoch": 0.9488008336181155,
+            "step": 21170
+        },
+        {
+            "loss": 13.667,
+            "grad_norm": 1.6418699026107788,
+            "learning_rate": 0.00025809184707681316,
+            "epoch": 0.949024924509381,
+            "step": 21175
+        },
+        {
+            "loss": 13.6324,
+            "grad_norm": 1.6431262493133545,
+            "learning_rate": 0.000256333178578297,
+            "epoch": 0.9492490154006465,
+            "step": 21180
+        },
+        {
+            "loss": 13.6902,
+            "grad_norm": 1.6390429735183716,
+            "learning_rate": 0.00025457419641098614,
+            "epoch": 0.949473106291912,
+            "step": 21185
+        },
+        {
+            "loss": 13.7573,
+            "grad_norm": 1.4962824583053589,
+            "learning_rate": 0.0002528149876935065,
+            "epoch": 0.9496971971831775,
+            "step": 21190
+        },
+        {
+            "loss": 13.599,
+            "grad_norm": 1.4873199462890625,
+            "learning_rate": 0.0002510556395557048,
+            "epoch": 0.949921288074443,
+            "step": 21195
+        },
+        {
+            "loss": 13.6939,
+            "grad_norm": 1.5821828842163086,
+            "learning_rate": 0.0002492962391343329,
+            "epoch": 0.9501453789657085,
+            "step": 21200
+        },
+        {
+            "loss": 13.7601,
+            "grad_norm": 1.5797697305679321,
+            "learning_rate": 0.00024753687356873213,
+            "epoch": 0.950369469856974,
+            "step": 21205
+        },
+        {
+            "loss": 13.6709,
+            "grad_norm": 1.4244056940078735,
+            "learning_rate": 0.0002457776299965173,
+            "epoch": 0.9505935607482395,
+            "step": 21210
+        },
+        {
+            "loss": 13.7313,
+            "grad_norm": 1.5052160024642944,
+            "learning_rate": 0.00024401859554926125,
+            "epoch": 0.9508176516395049,
+            "step": 21215
+        },
+        {
+            "loss": 13.7406,
+            "grad_norm": 1.6649264097213745,
+            "learning_rate": 0.0002422598573481797,
+            "epoch": 0.9510417425307705,
+            "step": 21220
+        },
+        {
+            "loss": 13.7893,
+            "grad_norm": 1.5528390407562256,
+            "learning_rate": 0.0002405015024998152,
+            "epoch": 0.951265833422036,
+            "step": 21225
+        },
+        {
+            "loss": 13.6475,
+            "grad_norm": 1.5336627960205078,
+            "learning_rate": 0.0002387436180917243,
+            "epoch": 0.9514899243133015,
+            "step": 21230
+        },
+        {
+            "loss": 13.6648,
+            "grad_norm": 1.5527150630950928,
+            "learning_rate": 0.00023698629118816335,
+            "epoch": 0.951714015204567,
+            "step": 21235
+        },
+        {
+            "loss": 13.7438,
+            "grad_norm": 1.579766035079956,
+            "learning_rate": 0.0002352296088257767,
+            "epoch": 0.9519381060958325,
+            "step": 21240
+        },
+        {
+            "loss": 13.7382,
+            "grad_norm": 1.4811670780181885,
+            "learning_rate": 0.00023347365800928602,
+            "epoch": 0.9521621969870979,
+            "step": 21245
+        },
+        {
+            "loss": 13.7377,
+            "grad_norm": 1.6110016107559204,
+            "learning_rate": 0.00023171852570718097,
+            "epoch": 0.9523862878783634,
+            "step": 21250
+        },
+        {
+            "loss": 13.6928,
+            "grad_norm": 1.587537169456482,
+            "learning_rate": 0.00022996429884741227,
+            "epoch": 0.952610378769629,
+            "step": 21255
+        },
+        {
+            "loss": 13.5688,
+            "grad_norm": 1.54538893699646,
+            "learning_rate": 0.00022821106431308543,
+            "epoch": 0.9528344696608945,
+            "step": 21260
+        },
+        {
+            "loss": 13.6446,
+            "grad_norm": 1.3947179317474365,
+            "learning_rate": 0.00022645890893815878,
+            "epoch": 0.95305856055216,
+            "step": 21265
+        },
+        {
+            "loss": 13.7477,
+            "grad_norm": 1.6194840669631958,
+            "learning_rate": 0.000224707919503142,
+            "epoch": 0.9532826514434255,
+            "step": 21270
+        },
+        {
+            "loss": 13.6585,
+            "grad_norm": 1.5126726627349854,
+            "learning_rate": 0.00022295818273079798,
+            "epoch": 0.9535067423346909,
+            "step": 21275
+        },
+        {
+            "loss": 13.5939,
+            "grad_norm": 1.5080811977386475,
+            "learning_rate": 0.00022120978528184833,
+            "epoch": 0.9537308332259564,
+            "step": 21280
+        },
+        {
+            "loss": 13.6283,
+            "grad_norm": 1.4491814374923706,
+            "learning_rate": 0.00021946281375068058,
+            "epoch": 0.9539549241172219,
+            "step": 21285
+        },
+        {
+            "loss": 13.6394,
+            "grad_norm": 1.4810166358947754,
+            "learning_rate": 0.0002177173546610597,
+            "epoch": 0.9541790150084875,
+            "step": 21290
+        },
+        {
+            "loss": 13.7695,
+            "grad_norm": 1.5009487867355347,
+            "learning_rate": 0.00021597349446184248,
+            "epoch": 0.954403105899753,
+            "step": 21295
+        },
+        {
+            "loss": 13.6825,
+            "grad_norm": 1.4756357669830322,
+            "learning_rate": 0.00021423131952269653,
+            "epoch": 0.9546271967910185,
+            "step": 21300
+        },
+        {
+            "loss": 13.7127,
+            "grad_norm": 1.5060594081878662,
+            "learning_rate": 0.00021249091612982155,
+            "epoch": 0.9548512876822839,
+            "step": 21305
+        },
+        {
+            "loss": 13.7083,
+            "grad_norm": 1.4578135013580322,
+            "learning_rate": 0.00021075237048167678,
+            "epoch": 0.9550753785735494,
+            "step": 21310
+        },
+        {
+            "loss": 13.6741,
+            "grad_norm": 1.4751098155975342,
+            "learning_rate": 0.00020901576868471126,
+            "epoch": 0.9552994694648149,
+            "step": 21315
+        },
+        {
+            "loss": 13.738,
+            "grad_norm": 1.4860180616378784,
+            "learning_rate": 0.00020728119674909894,
+            "epoch": 0.9555235603560804,
+            "step": 21320
+        },
+        {
+            "loss": 13.6476,
+            "grad_norm": 1.6425741910934448,
+            "learning_rate": 0.0002055487405844795,
+            "epoch": 0.955747651247346,
+            "step": 21325
+        },
+        {
+            "loss": 13.8511,
+            "grad_norm": 1.4640700817108154,
+            "learning_rate": 0.00020381848599570275,
+            "epoch": 0.9559717421386115,
+            "step": 21330
+        },
+        {
+            "loss": 13.6041,
+            "grad_norm": 1.4632081985473633,
+            "learning_rate": 0.0002020905186785791,
+            "epoch": 0.9561958330298769,
+            "step": 21335
+        },
+        {
+            "loss": 13.4984,
+            "grad_norm": 1.47840416431427,
+            "learning_rate": 0.0002003649242156355,
+            "epoch": 0.9564199239211424,
+            "step": 21340
+        },
+        {
+            "loss": 13.7759,
+            "grad_norm": 1.5127596855163574,
+            "learning_rate": 0.0001986417880718764,
+            "epoch": 0.9566440148124079,
+            "step": 21345
+        },
+        {
+            "loss": 13.725,
+            "grad_norm": 1.437754511833191,
+            "learning_rate": 0.00019692119559055102,
+            "epoch": 0.9568681057036734,
+            "step": 21350
+        },
+        {
+            "loss": 13.6686,
+            "grad_norm": 1.4675992727279663,
+            "learning_rate": 0.00019520323198892622,
+            "epoch": 0.9570921965949389,
+            "step": 21355
+        },
+        {
+            "loss": 13.7135,
+            "grad_norm": 1.4326221942901611,
+            "learning_rate": 0.00019348798235406628,
+            "epoch": 0.9573162874862045,
+            "step": 21360
+        },
+        {
+            "loss": 13.7331,
+            "grad_norm": 1.481142282485962,
+            "learning_rate": 0.0001917755316386185,
+            "epoch": 0.9575403783774699,
+            "step": 21365
+        },
+        {
+            "loss": 13.6136,
+            "grad_norm": 1.4526582956314087,
+            "learning_rate": 0.00019006596465660547,
+            "epoch": 0.9577644692687354,
+            "step": 21370
+        },
+        {
+            "loss": 13.6919,
+            "grad_norm": 1.4179489612579346,
+            "learning_rate": 0.00018835936607922483,
+            "epoch": 0.9579885601600009,
+            "step": 21375
+        },
+        {
+            "loss": 13.6267,
+            "grad_norm": 1.5630279779434204,
+            "learning_rate": 0.0001866558204306556,
+            "epoch": 0.9582126510512664,
+            "step": 21380
+        },
+        {
+            "loss": 13.6937,
+            "grad_norm": 1.4708082675933838,
+            "learning_rate": 0.00018495541208387128,
+            "epoch": 0.9584367419425319,
+            "step": 21385
+        },
+        {
+            "loss": 13.6937,
+            "grad_norm": 1.444834589958191,
+            "learning_rate": 0.00018325822525646208,
+            "epoch": 0.9586608328337974,
+            "step": 21390
+        },
+        {
+            "loss": 13.5621,
+            "grad_norm": 1.498726487159729,
+            "learning_rate": 0.000181564344006463,
+            "epoch": 0.9588849237250628,
+            "step": 21395
+        },
+        {
+            "loss": 13.8514,
+            "grad_norm": 1.4610778093338013,
+            "learning_rate": 0.0001798738522281907,
+            "epoch": 0.9591090146163284,
+            "step": 21400
+        },
+        {
+            "loss": 13.6962,
+            "grad_norm": 1.419760823249817,
+            "learning_rate": 0.00017818683364808884,
+            "epoch": 0.9593331055075939,
+            "step": 21405
+        },
+        {
+            "loss": 13.6112,
+            "grad_norm": 1.4531667232513428,
+            "learning_rate": 0.00017650337182058086,
+            "epoch": 0.9595571963988594,
+            "step": 21410
+        },
+        {
+            "loss": 13.6452,
+            "grad_norm": 1.4697725772857666,
+            "learning_rate": 0.00017482355012393176,
+            "epoch": 0.9597812872901249,
+            "step": 21415
+        },
+        {
+            "loss": 13.72,
+            "grad_norm": 1.4820373058319092,
+            "learning_rate": 0.0001731474517561188,
+            "epoch": 0.9600053781813904,
+            "step": 21420
+        },
+        {
+            "loss": 13.5874,
+            "grad_norm": 1.436105728149414,
+            "learning_rate": 0.00017147515973071076,
+            "epoch": 0.9602294690726558,
+            "step": 21425
+        },
+        {
+            "loss": 13.6086,
+            "grad_norm": 1.4315346479415894,
+            "learning_rate": 0.00016980675687275614,
+            "epoch": 0.9604535599639213,
+            "step": 21430
+        },
+        {
+            "loss": 13.669,
+            "grad_norm": 1.3922876119613647,
+            "learning_rate": 0.00016814232581468158,
+            "epoch": 0.9606776508551869,
+            "step": 21435
+        },
+        {
+            "loss": 13.4939,
+            "grad_norm": 1.5503712892532349,
+            "learning_rate": 0.00016648194899219885,
+            "epoch": 0.9609017417464524,
+            "step": 21440
+        },
+        {
+            "loss": 13.5681,
+            "grad_norm": 1.4173859357833862,
+            "learning_rate": 0.0001648257086402221,
+            "epoch": 0.9611258326377179,
+            "step": 21445
+        },
+        {
+            "loss": 13.5037,
+            "grad_norm": 1.3971552848815918,
+            "learning_rate": 0.00016317368678879496,
+            "epoch": 0.9613499235289834,
+            "step": 21450
+        },
+        {
+            "loss": 13.6957,
+            "grad_norm": 1.4970334768295288,
+            "learning_rate": 0.00016152596525902764,
+            "epoch": 0.9615740144202488,
+            "step": 21455
+        },
+        {
+            "loss": 13.5936,
+            "grad_norm": 1.4297453165054321,
+            "learning_rate": 0.0001598826256590449,
+            "epoch": 0.9617981053115143,
+            "step": 21460
+        },
+        {
+            "loss": 13.5697,
+            "grad_norm": 1.440354585647583,
+            "learning_rate": 0.0001582437493799434,
+            "epoch": 0.9620221962027798,
+            "step": 21465
+        },
+        {
+            "loss": 13.6957,
+            "grad_norm": 1.4379942417144775,
+            "learning_rate": 0.0001566094175917616,
+            "epoch": 0.9622462870940454,
+            "step": 21470
+        },
+        {
+            "loss": 13.5226,
+            "grad_norm": 1.4286212921142578,
+            "learning_rate": 0.00015497971123945873,
+            "epoch": 0.9624703779853109,
+            "step": 21475
+        },
+        {
+            "loss": 13.6297,
+            "grad_norm": 1.476987361907959,
+            "learning_rate": 0.00015335471103890603,
+            "epoch": 0.9626944688765764,
+            "step": 21480
+        },
+        {
+            "loss": 13.5253,
+            "grad_norm": 1.4980093240737915,
+            "learning_rate": 0.00015173449747288932,
+            "epoch": 0.9629185597678418,
+            "step": 21485
+        },
+        {
+            "loss": 13.5963,
+            "grad_norm": 1.3949865102767944,
+            "learning_rate": 0.00015011915078712252,
+            "epoch": 0.9631426506591073,
+            "step": 21490
+        },
+        {
+            "loss": 13.6885,
+            "grad_norm": 1.4428666830062866,
+            "learning_rate": 0.00014850875098627324,
+            "epoch": 0.9633667415503728,
+            "step": 21495
+        },
+        {
+            "loss": 13.586,
+            "grad_norm": 1.4490755796432495,
+            "learning_rate": 0.00014690337783000075,
+            "epoch": 0.9635908324416383,
+            "step": 21500
+        },
+        {
+            "eval_loss": 1.6970657110214233,
+            "eval_runtime": 18.8072,
+            "eval_samples_per_second": 871.155,
+            "eval_steps_per_second": 7.816,
+            "epoch": 0.9635908324416383,
+            "step": 21500
+        },
+        {
+            "loss": 13.6201,
+            "grad_norm": 1.3927302360534668,
+            "learning_rate": 0.00014530311082900526,
+            "epoch": 0.9638149233329039,
+            "step": 21505
+        },
+        {
+            "loss": 13.587,
+            "grad_norm": 1.371099829673767,
+            "learning_rate": 0.0001437080292410899,
+            "epoch": 0.9640390142241694,
+            "step": 21510
+        },
+        {
+            "loss": 13.7055,
+            "grad_norm": 1.3734136819839478,
+            "learning_rate": 0.00014211821206723535,
+            "epoch": 0.9642631051154348,
+            "step": 21515
+        },
+        {
+            "loss": 13.5536,
+            "grad_norm": 1.366150140762329,
+            "learning_rate": 0.00014053373804768742,
+            "epoch": 0.9644871960067003,
+            "step": 21520
+        },
+        {
+            "loss": 13.6378,
+            "grad_norm": 1.3889002799987793,
+            "learning_rate": 0.00013895468565805656,
+            "epoch": 0.9647112868979658,
+            "step": 21525
+        },
+        {
+            "loss": 13.5321,
+            "grad_norm": 1.3824708461761475,
+            "learning_rate": 0.00013738113310543176,
+            "epoch": 0.9649353777892313,
+            "step": 21530
+        },
+        {
+            "loss": 13.5276,
+            "grad_norm": 1.4974249601364136,
+            "learning_rate": 0.00013581315832450662,
+            "epoch": 0.9651594686804968,
+            "step": 21535
+        },
+        {
+            "loss": 13.5834,
+            "grad_norm": 1.4725018739700317,
+            "learning_rate": 0.00013425083897371983,
+            "epoch": 0.9653835595717624,
+            "step": 21540
+        },
+        {
+            "loss": 13.6833,
+            "grad_norm": 1.446146845817566,
+            "learning_rate": 0.00013269425243140853,
+            "epoch": 0.9656076504630278,
+            "step": 21545
+        },
+        {
+            "loss": 13.6362,
+            "grad_norm": 1.3499586582183838,
+            "learning_rate": 0.0001311434757919762,
+            "epoch": 0.9658317413542933,
+            "step": 21550
+        },
+        {
+            "loss": 13.6207,
+            "grad_norm": 1.40580153465271,
+            "learning_rate": 0.00012959858586207435,
+            "epoch": 0.9660558322455588,
+            "step": 21555
+        },
+        {
+            "loss": 13.5173,
+            "grad_norm": 1.4460829496383667,
+            "learning_rate": 0.00012805965915679807,
+            "epoch": 0.9662799231368243,
+            "step": 21560
+        },
+        {
+            "loss": 13.638,
+            "grad_norm": 1.4111204147338867,
+            "learning_rate": 0.0001265267718958971,
+            "epoch": 0.9665040140280898,
+            "step": 21565
+        },
+        {
+            "loss": 13.7012,
+            "grad_norm": 1.3578362464904785,
+            "learning_rate": 0.00012500000000000006,
+            "epoch": 0.9667281049193552,
+            "step": 21570
+        },
+        {
+            "loss": 13.5422,
+            "grad_norm": 1.5003434419631958,
+            "learning_rate": 0.00012347941908685464,
+            "epoch": 0.9669521958106208,
+            "step": 21575
+        },
+        {
+            "loss": 13.5665,
+            "grad_norm": 1.3598604202270508,
+            "learning_rate": 0.00012196510446758268,
+            "epoch": 0.9671762867018863,
+            "step": 21580
+        },
+        {
+            "loss": 13.6541,
+            "grad_norm": 1.4082996845245361,
+            "learning_rate": 0.0001204571311429496,
+            "epoch": 0.9674003775931518,
+            "step": 21585
+        },
+        {
+            "loss": 13.5797,
+            "grad_norm": 1.3907629251480103,
+            "learning_rate": 0.00011895557379965005,
+            "epoch": 0.9676244684844173,
+            "step": 21590
+        },
+        {
+            "loss": 13.5672,
+            "grad_norm": 1.365025281906128,
+            "learning_rate": 0.00011746050680660903,
+            "epoch": 0.9678485593756828,
+            "step": 21595
+        },
+        {
+            "loss": 13.5871,
+            "grad_norm": 1.3568958044052124,
+            "learning_rate": 0.00011597200421129844,
+            "epoch": 0.9680726502669482,
+            "step": 21600
+        },
+        {
+            "loss": 13.5817,
+            "grad_norm": 1.4446165561676025,
+            "learning_rate": 0.00011449013973606907,
+            "epoch": 0.9682967411582137,
+            "step": 21605
+        },
+        {
+            "loss": 13.4769,
+            "grad_norm": 1.3863264322280884,
+            "learning_rate": 0.00011301498677450037,
+            "epoch": 0.9685208320494793,
+            "step": 21610
+        },
+        {
+            "loss": 13.5077,
+            "grad_norm": 1.3644050359725952,
+            "learning_rate": 0.00011154661838776472,
+            "epoch": 0.9687449229407448,
+            "step": 21615
+        },
+        {
+            "loss": 13.5713,
+            "grad_norm": 1.4149904251098633,
+            "learning_rate": 0.00011008510730100893,
+            "epoch": 0.9689690138320103,
+            "step": 21620
+        },
+        {
+            "loss": 13.5241,
+            "grad_norm": 1.3770707845687866,
+            "learning_rate": 0.0001086305258997523,
+            "epoch": 0.9691931047232758,
+            "step": 21625
+        },
+        {
+            "loss": 13.5229,
+            "grad_norm": 1.3707841634750366,
+            "learning_rate": 0.00010718294622630187,
+            "epoch": 0.9694171956145412,
+            "step": 21630
+        },
+        {
+            "loss": 13.4982,
+            "grad_norm": 1.2850505113601685,
+            "learning_rate": 0.00010574243997618415,
+            "epoch": 0.9696412865058067,
+            "step": 21635
+        },
+        {
+            "loss": 13.6666,
+            "grad_norm": 1.3031147718429565,
+            "learning_rate": 0.00010430907849459354,
+            "epoch": 0.9698653773970722,
+            "step": 21640
+        },
+        {
+            "loss": 13.599,
+            "grad_norm": 1.4308277368545532,
+            "learning_rate": 0.0001028829327728599,
+            "epoch": 0.9700894682883378,
+            "step": 21645
+        },
+        {
+            "loss": 13.5511,
+            "grad_norm": 1.3806493282318115,
+            "learning_rate": 0.00010146407344493186,
+            "epoch": 0.9703135591796033,
+            "step": 21650
+        },
+        {
+            "loss": 13.5986,
+            "grad_norm": 1.3425934314727783,
+            "learning_rate": 0.0001000525707838783,
+            "epoch": 0.9705376500708688,
+            "step": 21655
+        },
+        {
+            "loss": 13.6673,
+            "grad_norm": 1.3432285785675049,
+            "learning_rate": 9.864849469840822e-05,
+            "epoch": 0.9707617409621342,
+            "step": 21660
+        },
+        {
+            "loss": 13.6672,
+            "grad_norm": 1.3797509670257568,
+            "learning_rate": 9.725191472940837e-05,
+            "epoch": 0.9709858318533997,
+            "step": 21665
+        },
+        {
+            "loss": 13.6013,
+            "grad_norm": 1.3327072858810425,
+            "learning_rate": 9.586290004649867e-05,
+            "epoch": 0.9712099227446652,
+            "step": 21670
+        },
+        {
+            "loss": 13.6346,
+            "grad_norm": 1.3804453611373901,
+            "learning_rate": 9.448151944460656e-05,
+            "epoch": 0.9714340136359307,
+            "step": 21675
+        },
+        {
+            "loss": 13.5767,
+            "grad_norm": 1.3617254495620728,
+            "learning_rate": 9.31078413405601e-05,
+            "epoch": 0.9716581045271963,
+            "step": 21680
+        },
+        {
+            "loss": 13.5679,
+            "grad_norm": 1.3983831405639648,
+            "learning_rate": 9.174193376969866e-05,
+            "epoch": 0.9718821954184618,
+            "step": 21685
+        },
+        {
+            "loss": 13.5782,
+            "grad_norm": 1.3388102054595947,
+            "learning_rate": 9.038386438250415e-05,
+            "epoch": 0.9721062863097272,
+            "step": 21690
+        },
+        {
+            "loss": 13.5006,
+            "grad_norm": 1.3537145853042603,
+            "learning_rate": 8.903370044124967e-05,
+            "epoch": 0.9723303772009927,
+            "step": 21695
+        },
+        {
+            "loss": 13.5512,
+            "grad_norm": 1.3179186582565308,
+            "learning_rate": 8.769150881666851e-05,
+            "epoch": 0.9725544680922582,
+            "step": 21700
+        },
+        {
+            "loss": 13.5385,
+            "grad_norm": 1.3044551610946655,
+            "learning_rate": 8.635735598464243e-05,
+            "epoch": 0.9727785589835237,
+            "step": 21705
+        },
+        {
+            "loss": 13.5015,
+            "grad_norm": 1.3320732116699219,
+            "learning_rate": 8.503130802290862e-05,
+            "epoch": 0.9730026498747892,
+            "step": 21710
+        },
+        {
+            "loss": 13.457,
+            "grad_norm": 1.3160337209701538,
+            "learning_rate": 8.371343060778771e-05,
+            "epoch": 0.9732267407660548,
+            "step": 21715
+        },
+        {
+            "loss": 13.5789,
+            "grad_norm": 1.3159329891204834,
+            "learning_rate": 8.240378901093035e-05,
+            "epoch": 0.9734508316573202,
+            "step": 21720
+        },
+        {
+            "loss": 13.55,
+            "grad_norm": 1.3328814506530762,
+            "learning_rate": 8.110244809608495e-05,
+            "epoch": 0.9736749225485857,
+            "step": 21725
+        },
+        {
+            "loss": 13.47,
+            "grad_norm": 1.2326339483261108,
+            "learning_rate": 7.980947231588471e-05,
+            "epoch": 0.9738990134398512,
+            "step": 21730
+        },
+        {
+            "loss": 13.4876,
+            "grad_norm": 1.370176911354065,
+            "learning_rate": 7.852492570865557e-05,
+            "epoch": 0.9741231043311167,
+            "step": 21735
+        },
+        {
+            "loss": 13.4974,
+            "grad_norm": 1.4055640697479248,
+            "learning_rate": 7.724887189524485e-05,
+            "epoch": 0.9743471952223822,
+            "step": 21740
+        },
+        {
+            "loss": 13.4864,
+            "grad_norm": 1.3312071561813354,
+            "learning_rate": 7.598137407586958e-05,
+            "epoch": 0.9745712861136477,
+            "step": 21745
+        },
+        {
+            "loss": 13.5479,
+            "grad_norm": 1.2774869203567505,
+            "learning_rate": 7.472249502698686e-05,
+            "epoch": 0.9747953770049131,
+            "step": 21750
+        },
+        {
+            "loss": 13.5515,
+            "grad_norm": 1.2742916345596313,
+            "learning_rate": 7.347229709818453e-05,
+            "epoch": 0.9750194678961787,
+            "step": 21755
+        },
+        {
+            "loss": 13.6203,
+            "grad_norm": 1.309590220451355,
+            "learning_rate": 7.223084220909332e-05,
+            "epoch": 0.9752435587874442,
+            "step": 21760
+        },
+        {
+            "loss": 13.5797,
+            "grad_norm": 1.290747046470642,
+            "learning_rate": 7.099819184631928e-05,
+            "epoch": 0.9754676496787097,
+            "step": 21765
+        },
+        {
+            "loss": 13.5127,
+            "grad_norm": 1.37006676197052,
+            "learning_rate": 6.977440706039972e-05,
+            "epoch": 0.9756917405699752,
+            "step": 21770
+        },
+        {
+            "loss": 13.4896,
+            "grad_norm": 1.3133093118667603,
+            "learning_rate": 6.85595484627787e-05,
+            "epoch": 0.9759158314612407,
+            "step": 21775
+        },
+        {
+            "loss": 13.4632,
+            "grad_norm": 1.3016916513442993,
+            "learning_rate": 6.735367622280513e-05,
+            "epoch": 0.9761399223525061,
+            "step": 21780
+        },
+        {
+            "loss": 13.5153,
+            "grad_norm": 1.2847900390625,
+            "learning_rate": 6.615685006475284e-05,
+            "epoch": 0.9763640132437716,
+            "step": 21785
+        },
+        {
+            "loss": 13.5218,
+            "grad_norm": 1.3632951974868774,
+            "learning_rate": 6.496912926486279e-05,
+            "epoch": 0.9765881041350372,
+            "step": 21790
+        },
+        {
+            "loss": 13.5126,
+            "grad_norm": 1.2633512020111084,
+            "learning_rate": 6.379057264840679e-05,
+            "epoch": 0.9768121950263027,
+            "step": 21795
+        },
+        {
+            "loss": 13.5233,
+            "grad_norm": 1.3291096687316895,
+            "learning_rate": 6.262123858677426e-05,
+            "epoch": 0.9770362859175682,
+            "step": 21800
+        },
+        {
+            "loss": 13.5546,
+            "grad_norm": 1.2681336402893066,
+            "learning_rate": 6.146118499458131e-05,
+            "epoch": 0.9772603768088337,
+            "step": 21805
+        },
+        {
+            "loss": 13.5603,
+            "grad_norm": 1.3324332237243652,
+            "learning_rate": 6.0310469326802285e-05,
+            "epoch": 0.9774844677000991,
+            "step": 21810
+        },
+        {
+            "loss": 13.6537,
+            "grad_norm": 1.3532867431640625,
+            "learning_rate": 5.916914857592387e-05,
+            "epoch": 0.9777085585913646,
+            "step": 21815
+        },
+        {
+            "loss": 13.5717,
+            "grad_norm": 1.341113805770874,
+            "learning_rate": 5.803727926912269e-05,
+            "epoch": 0.9779326494826301,
+            "step": 21820
+        },
+        {
+            "loss": 13.5134,
+            "grad_norm": 1.4140442609786987,
+            "learning_rate": 5.691491746546573e-05,
+            "epoch": 0.9781567403738957,
+            "step": 21825
+        },
+        {
+            "loss": 13.5485,
+            "grad_norm": 1.2625435590744019,
+            "learning_rate": 5.580211875313346e-05,
+            "epoch": 0.9783808312651612,
+            "step": 21830
+        },
+        {
+            "loss": 13.4533,
+            "grad_norm": 1.3162455558776855,
+            "learning_rate": 5.469893824666686e-05,
+            "epoch": 0.9786049221564267,
+            "step": 21835
+        },
+        {
+            "loss": 13.4871,
+            "grad_norm": 1.3901221752166748,
+            "learning_rate": 5.3605430584238e-05,
+            "epoch": 0.9788290130476921,
+            "step": 21840
+        },
+        {
+            "loss": 13.5566,
+            "grad_norm": 1.3070961236953735,
+            "learning_rate": 5.252164992494338e-05,
+            "epoch": 0.9790531039389576,
+            "step": 21845
+        },
+        {
+            "loss": 13.5652,
+            "grad_norm": 1.3653221130371094,
+            "learning_rate": 5.1447649946122e-05,
+            "epoch": 0.9792771948302231,
+            "step": 21850
+        },
+        {
+            "loss": 13.4793,
+            "grad_norm": 1.3583656549453735,
+            "learning_rate": 5.038348384069663e-05,
+            "epoch": 0.9795012857214886,
+            "step": 21855
+        },
+        {
+            "loss": 13.5538,
+            "grad_norm": 1.2869338989257812,
+            "learning_rate": 4.9329204314539186e-05,
+            "epoch": 0.9797253766127542,
+            "step": 21860
+        },
+        {
+            "loss": 13.5843,
+            "grad_norm": 1.2691680192947388,
+            "learning_rate": 4.828486358386072e-05,
+            "epoch": 0.9799494675040197,
+            "step": 21865
+        },
+        {
+            "loss": 13.5242,
+            "grad_norm": 1.2550824880599976,
+            "learning_rate": 4.725051337262476e-05,
+            "epoch": 0.9801735583952851,
+            "step": 21870
+        },
+        {
+            "loss": 13.4907,
+            "grad_norm": 1.3409019708633423,
+            "learning_rate": 4.6226204909985777e-05,
+            "epoch": 0.9803976492865506,
+            "step": 21875
+        },
+        {
+            "loss": 13.5526,
+            "grad_norm": 1.2815920114517212,
+            "learning_rate": 4.521198892775202e-05,
+            "epoch": 0.9806217401778161,
+            "step": 21880
+        },
+        {
+            "loss": 13.4688,
+            "grad_norm": 1.2765311002731323,
+            "learning_rate": 4.420791565787288e-05,
+            "epoch": 0.9808458310690816,
+            "step": 21885
+        },
+        {
+            "loss": 13.4585,
+            "grad_norm": 1.2942756414413452,
+            "learning_rate": 4.3214034829950396e-05,
+            "epoch": 0.9810699219603471,
+            "step": 21890
+        },
+        {
+            "loss": 13.6144,
+            "grad_norm": 1.3017674684524536,
+            "learning_rate": 4.223039566877729e-05,
+            "epoch": 0.9812940128516127,
+            "step": 21895
+        },
+        {
+            "loss": 13.4723,
+            "grad_norm": 1.2312167882919312,
+            "learning_rate": 4.125704689189819e-05,
+            "epoch": 0.9815181037428781,
+            "step": 21900
+        },
+        {
+            "loss": 13.4998,
+            "grad_norm": 1.3000839948654175,
+            "learning_rate": 4.0294036707196945e-05,
+            "epoch": 0.9817421946341436,
+            "step": 21905
+        },
+        {
+            "loss": 13.5832,
+            "grad_norm": 1.3279622793197632,
+            "learning_rate": 3.93414128105091e-05,
+            "epoch": 0.9819662855254091,
+            "step": 21910
+        },
+        {
+            "loss": 13.4335,
+            "grad_norm": 1.3625671863555908,
+            "learning_rate": 3.83992223832596e-05,
+            "epoch": 0.9821903764166746,
+            "step": 21915
+        },
+        {
+            "loss": 13.4777,
+            "grad_norm": 1.2160567045211792,
+            "learning_rate": 3.7467512090126e-05,
+            "epoch": 0.9824144673079401,
+            "step": 21920
+        },
+        {
+            "loss": 13.5191,
+            "grad_norm": 1.3054990768432617,
+            "learning_rate": 3.654632807672695e-05,
+            "epoch": 0.9826385581992056,
+            "step": 21925
+        },
+        {
+            "loss": 13.5742,
+            "grad_norm": 1.334370732307434,
+            "learning_rate": 3.563571596733722e-05,
+            "epoch": 0.982862649090471,
+            "step": 21930
+        },
+        {
+            "loss": 13.6078,
+            "grad_norm": 1.3172982931137085,
+            "learning_rate": 3.473572086262783e-05,
+            "epoch": 0.9830867399817366,
+            "step": 21935
+        },
+        {
+            "loss": 13.4361,
+            "grad_norm": 1.299933671951294,
+            "learning_rate": 3.3846387337432034e-05,
+            "epoch": 0.9833108308730021,
+            "step": 21940
+        },
+        {
+            "loss": 13.547,
+            "grad_norm": 1.253339409828186,
+            "learning_rate": 3.2967759438537886e-05,
+            "epoch": 0.9835349217642676,
+            "step": 21945
+        },
+        {
+            "loss": 13.5014,
+            "grad_norm": 1.2532811164855957,
+            "learning_rate": 3.209988068250688e-05,
+            "epoch": 0.9837590126555331,
+            "step": 21950
+        },
+        {
+            "loss": 13.457,
+            "grad_norm": 1.2518850564956665,
+            "learning_rate": 3.1242794053518234e-05,
+            "epoch": 0.9839831035467986,
+            "step": 21955
+        },
+        {
+            "loss": 13.5006,
+            "grad_norm": 1.3577061891555786,
+            "learning_rate": 3.0396542001240145e-05,
+            "epoch": 0.984207194438064,
+            "step": 21960
+        },
+        {
+            "loss": 13.5796,
+            "grad_norm": 1.237846851348877,
+            "learning_rate": 2.9561166438727638e-05,
+            "epoch": 0.9844312853293296,
+            "step": 21965
+        },
+        {
+            "loss": 13.5357,
+            "grad_norm": 1.1801531314849854,
+            "learning_rate": 2.8736708740346146e-05,
+            "epoch": 0.9846553762205951,
+            "step": 21970
+        },
+        {
+            "loss": 13.5288,
+            "grad_norm": 1.3670539855957031,
+            "learning_rate": 2.7923209739722955e-05,
+            "epoch": 0.9848794671118606,
+            "step": 21975
+        },
+        {
+            "loss": 13.5071,
+            "grad_norm": 1.3142650127410889,
+            "learning_rate": 2.7120709727724207e-05,
+            "epoch": 0.9851035580031261,
+            "step": 21980
+        },
+        {
+            "loss": 13.5824,
+            "grad_norm": 1.2690343856811523,
+            "learning_rate": 2.632924845045975e-05,
+            "epoch": 0.9853276488943916,
+            "step": 21985
+        },
+        {
+            "loss": 13.4953,
+            "grad_norm": 1.2118849754333496,
+            "learning_rate": 2.5548865107314605e-05,
+            "epoch": 0.985551739785657,
+            "step": 21990
+        },
+        {
+            "loss": 13.458,
+            "grad_norm": 1.2431721687316895,
+            "learning_rate": 2.4779598349007227e-05,
+            "epoch": 0.9857758306769225,
+            "step": 21995
+        },
+        {
+            "loss": 13.5144,
+            "grad_norm": 1.2333933115005493,
+            "learning_rate": 2.402148627567555e-05,
+            "epoch": 0.985999921568188,
+            "step": 22000
+        },
+        {
+            "eval_loss": 1.6814430952072144,
+            "eval_runtime": 18.632,
+            "eval_samples_per_second": 879.348,
+            "eval_steps_per_second": 7.89,
+            "epoch": 0.985999921568188,
+            "step": 22000
+        },
+        {
+            "loss": 13.4373,
+            "grad_norm": 1.2443017959594727,
+            "learning_rate": 2.3274566434989626e-05,
+            "epoch": 0.9862240124594536,
+            "step": 22005
+        },
+        {
+            "loss": 13.4767,
+            "grad_norm": 1.2459810972213745,
+            "learning_rate": 2.2538875820292348e-05,
+            "epoch": 0.9864481033507191,
+            "step": 22010
+        },
+        {
+            "loss": 13.4842,
+            "grad_norm": 1.2842656373977661,
+            "learning_rate": 2.181445086876696e-05,
+            "epoch": 0.9866721942419846,
+            "step": 22015
+        },
+        {
+            "loss": 13.5223,
+            "grad_norm": 1.3632595539093018,
+            "learning_rate": 2.1101327459632445e-05,
+            "epoch": 0.98689628513325,
+            "step": 22020
+        },
+        {
+            "loss": 13.4355,
+            "grad_norm": 1.2427505254745483,
+            "learning_rate": 2.0399540912366675e-05,
+            "epoch": 0.9871203760245155,
+            "step": 22025
+        },
+        {
+            "loss": 13.6738,
+            "grad_norm": 1.2594168186187744,
+            "learning_rate": 1.970912598495689e-05,
+            "epoch": 0.987344466915781,
+            "step": 22030
+        },
+        {
+            "loss": 13.555,
+            "grad_norm": 1.211691975593567,
+            "learning_rate": 1.9030116872178316e-05,
+            "epoch": 0.9875685578070466,
+            "step": 22035
+        },
+        {
+            "loss": 13.5905,
+            "grad_norm": 1.255252480506897,
+            "learning_rate": 1.8362547203900625e-05,
+            "epoch": 0.9877926486983121,
+            "step": 22040
+        },
+        {
+            "loss": 13.4202,
+            "grad_norm": 1.2320276498794556,
+            "learning_rate": 1.7706450043422308e-05,
+            "epoch": 0.9880167395895776,
+            "step": 22045
+        },
+        {
+            "loss": 13.5955,
+            "grad_norm": 1.2293109893798828,
+            "learning_rate": 1.7061857885832893e-05,
+            "epoch": 0.988240830480843,
+            "step": 22050
+        },
+        {
+            "loss": 13.4793,
+            "grad_norm": 1.2270162105560303,
+            "learning_rate": 1.6428802656403842e-05,
+            "epoch": 0.9884649213721085,
+            "step": 22055
+        },
+        {
+            "loss": 13.4329,
+            "grad_norm": 1.2630739212036133,
+            "learning_rate": 1.580731570900723e-05,
+            "epoch": 0.988689012263374,
+            "step": 22060
+        },
+        {
+            "loss": 13.395,
+            "grad_norm": 1.3208829164505005,
+            "learning_rate": 1.519742782456282e-05,
+            "epoch": 0.9889131031546395,
+            "step": 22065
+        },
+        {
+            "loss": 13.4516,
+            "grad_norm": 1.225806474685669,
+            "learning_rate": 1.4599169209513568e-05,
+            "epoch": 0.989137194045905,
+            "step": 22070
+        },
+        {
+            "loss": 13.4567,
+            "grad_norm": 1.2742187976837158,
+            "learning_rate": 1.4012569494329664e-05,
+            "epoch": 0.9893612849371706,
+            "step": 22075
+        },
+        {
+            "loss": 13.5909,
+            "grad_norm": 1.2377104759216309,
+            "learning_rate": 1.3437657732040782e-05,
+            "epoch": 0.989585375828436,
+            "step": 22080
+        },
+        {
+            "loss": 13.5011,
+            "grad_norm": 1.279120683670044,
+            "learning_rate": 1.287446239679746e-05,
+            "epoch": 0.9898094667197015,
+            "step": 22085
+        },
+        {
+            "loss": 13.4,
+            "grad_norm": 1.2107067108154297,
+            "learning_rate": 1.232301138246042e-05,
+            "epoch": 0.990033557610967,
+            "step": 22090
+        },
+        {
+            "loss": 13.4454,
+            "grad_norm": 1.2431056499481201,
+            "learning_rate": 1.1783332001219533e-05,
+            "epoch": 0.9902576485022325,
+            "step": 22095
+        },
+        {
+            "loss": 13.4354,
+            "grad_norm": 1.2113475799560547,
+            "learning_rate": 1.1255450982240679e-05,
+            "epoch": 0.990481739393498,
+            "step": 22100
+        },
+        {
+            "loss": 13.4946,
+            "grad_norm": 1.2396379709243774,
+            "learning_rate": 1.0739394470342057e-05,
+            "epoch": 0.9907058302847636,
+            "step": 22105
+        },
+        {
+            "loss": 13.4944,
+            "grad_norm": 1.2629179954528809,
+            "learning_rate": 1.0235188024699471e-05,
+            "epoch": 0.990929921176029,
+            "step": 22110
+        },
+        {
+            "loss": 13.4847,
+            "grad_norm": 1.1953667402267456,
+            "learning_rate": 9.742856617580147e-06,
+            "epoch": 0.9911540120672945,
+            "step": 22115
+        },
+        {
+            "loss": 13.4528,
+            "grad_norm": 1.1882737874984741,
+            "learning_rate": 9.262424633106115e-06,
+            "epoch": 0.99137810295856,
+            "step": 22120
+        },
+        {
+            "loss": 13.5766,
+            "grad_norm": 1.2557233572006226,
+            "learning_rate": 8.793915866046358e-06,
+            "epoch": 0.9916021938498255,
+            "step": 22125
+        },
+        {
+            "loss": 13.5409,
+            "grad_norm": 1.2498853206634521,
+            "learning_rate": 8.337353520638468e-06,
+            "epoch": 0.991826284741091,
+            "step": 22130
+        },
+        {
+            "loss": 13.4702,
+            "grad_norm": 1.2380563020706177,
+            "learning_rate": 7.892760209439298e-06,
+            "epoch": 0.9920503756323564,
+            "step": 22135
+        },
+        {
+            "loss": 13.5216,
+            "grad_norm": 1.3622403144836426,
+            "learning_rate": 7.460157952205032e-06,
+            "epoch": 0.992274466523622,
+            "step": 22140
+        },
+        {
+            "loss": 13.4822,
+            "grad_norm": 1.223172903060913,
+            "learning_rate": 7.039568174800504e-06,
+            "epoch": 0.9924985574148875,
+            "step": 22145
+        },
+        {
+            "loss": 13.5725,
+            "grad_norm": 1.1964086294174194,
+            "learning_rate": 6.631011708138207e-06,
+            "epoch": 0.992722648306153,
+            "step": 22150
+        },
+        {
+            "loss": 13.3952,
+            "grad_norm": 1.2155392169952393,
+            "learning_rate": 6.234508787146543e-06,
+            "epoch": 0.9929467391974185,
+            "step": 22155
+        },
+        {
+            "loss": 13.5441,
+            "grad_norm": 1.1746368408203125,
+            "learning_rate": 5.850079049767309e-06,
+            "epoch": 0.993170830088684,
+            "step": 22160
+        },
+        {
+            "loss": 13.4357,
+            "grad_norm": 1.1912814378738403,
+            "learning_rate": 5.477741535983572e-06,
+            "epoch": 0.9933949209799494,
+            "step": 22165
+        },
+        {
+            "loss": 13.4152,
+            "grad_norm": 1.2137175798416138,
+            "learning_rate": 5.117514686876379e-06,
+            "epoch": 0.9936190118712149,
+            "step": 22170
+        },
+        {
+            "loss": 13.6085,
+            "grad_norm": 1.1917462348937988,
+            "learning_rate": 4.769416343711364e-06,
+            "epoch": 0.9938431027624804,
+            "step": 22175
+        },
+        {
+            "loss": 13.5033,
+            "grad_norm": 1.1859139204025269,
+            "learning_rate": 4.433463747055194e-06,
+            "epoch": 0.994067193653746,
+            "step": 22180
+        },
+        {
+            "loss": 13.4355,
+            "grad_norm": 1.1885240077972412,
+            "learning_rate": 4.10967353592176e-06,
+            "epoch": 0.9942912845450115,
+            "step": 22185
+        },
+        {
+            "loss": 13.4492,
+            "grad_norm": 1.2175772190093994,
+            "learning_rate": 3.798061746947995e-06,
+            "epoch": 0.994515375436277,
+            "step": 22190
+        },
+        {
+            "loss": 13.491,
+            "grad_norm": 1.2592718601226807,
+            "learning_rate": 3.498643813599517e-06,
+            "epoch": 0.9947394663275424,
+            "step": 22195
+        },
+        {
+            "loss": 13.4639,
+            "grad_norm": 1.1634012460708618,
+            "learning_rate": 3.211434565406457e-06,
+            "epoch": 0.9949635572188079,
+            "step": 22200
+        },
+        {
+            "loss": 13.5012,
+            "grad_norm": 1.233546257019043,
+            "learning_rate": 2.9364482272288273e-06,
+            "epoch": 0.9951876481100734,
+            "step": 22205
+        },
+        {
+            "loss": 13.4361,
+            "grad_norm": 1.2499737739562988,
+            "learning_rate": 2.6736984185520286e-06,
+            "epoch": 0.995411739001339,
+            "step": 22210
+        },
+        {
+            "loss": 13.4777,
+            "grad_norm": 1.1770788431167603,
+            "learning_rate": 2.423198152812306e-06,
+            "epoch": 0.9956358298926045,
+            "step": 22215
+        },
+        {
+            "loss": 13.4065,
+            "grad_norm": 1.223405361175537,
+            "learning_rate": 2.1849598367522926e-06,
+            "epoch": 0.99585992078387,
+            "step": 22220
+        },
+        {
+            "loss": 13.5333,
+            "grad_norm": 1.2335143089294434,
+            "learning_rate": 1.958995269806446e-06,
+            "epoch": 0.9960840116751354,
+            "step": 22225
+        },
+        {
+            "loss": 13.3936,
+            "grad_norm": 1.1790462732315063,
+            "learning_rate": 1.7453156435165983e-06,
+            "epoch": 0.9963081025664009,
+            "step": 22230
+        },
+        {
+            "loss": 13.4622,
+            "grad_norm": 1.1784507036209106,
+            "learning_rate": 1.5439315409778443e-06,
+            "epoch": 0.9965321934576664,
+            "step": 22235
+        },
+        {
+            "loss": 13.4254,
+            "grad_norm": 1.2225220203399658,
+            "learning_rate": 1.3548529363142104e-06,
+            "epoch": 0.9967562843489319,
+            "step": 22240
+        },
+        {
+            "loss": 13.588,
+            "grad_norm": 1.2711127996444702,
+            "learning_rate": 1.1780891941847448e-06,
+            "epoch": 0.9969803752401974,
+            "step": 22245
+        },
+        {
+            "loss": 13.5485,
+            "grad_norm": 1.226955771446228,
+            "learning_rate": 1.0136490693196666e-06,
+            "epoch": 0.997204466131463,
+            "step": 22250
+        },
+        {
+            "loss": 13.5763,
+            "grad_norm": 1.2618262767791748,
+            "learning_rate": 8.615407060867663e-07,
+            "epoch": 0.9974285570227284,
+            "step": 22255
+        },
+        {
+            "loss": 13.5191,
+            "grad_norm": 1.189131498336792,
+            "learning_rate": 7.217716380881478e-07,
+            "epoch": 0.9976526479139939,
+            "step": 22260
+        },
+        {
+            "loss": 13.3736,
+            "grad_norm": 1.207396149635315,
+            "learning_rate": 5.943487877868303e-07,
+            "epoch": 0.9978767388052594,
+            "step": 22265
+        },
+        {
+            "loss": 13.518,
+            "grad_norm": 1.1892287731170654,
+            "learning_rate": 4.792784661642458e-07,
+            "epoch": 0.9981008296965249,
+            "step": 22270
+        },
+        {
+            "loss": 13.5047,
+            "grad_norm": 1.2454142570495605,
+            "learning_rate": 3.7656637240732206e-07,
+            "epoch": 0.9983249205877904,
+            "step": 22275
+        },
+        {
+            "loss": 13.506,
+            "grad_norm": 1.1761161088943481,
+            "learning_rate": 2.862175936265421e-07,
+            "epoch": 0.998549011479056,
+            "step": 22280
+        },
+        {
+            "loss": 13.5602,
+            "grad_norm": 1.1850390434265137,
+            "learning_rate": 2.0823660460370098e-07,
+            "epoch": 0.9987731023703214,
+            "step": 22285
+        },
+        {
+            "loss": 13.3766,
+            "grad_norm": 1.171899676322937,
+            "learning_rate": 1.426272675704998e-07,
+            "epoch": 0.9989971932615869,
+            "step": 22290
+        },
+        {
+            "loss": 13.3907,
+            "grad_norm": 1.218133568763733,
+            "learning_rate": 8.939283201708782e-08,
+            "epoch": 0.9992212841528524,
+            "step": 22295
+        },
+        {
+            "loss": 13.3937,
+            "grad_norm": 1.1850651502609253,
+            "learning_rate": 4.8535934531274137e-08,
+            "epoch": 0.9994453750441179,
+            "step": 22300
+        },
+        {
+            "loss": 13.5193,
+            "grad_norm": 1.1924799680709839,
+            "learning_rate": 2.0058598667854756e-08,
+            "epoch": 0.9996694659353834,
+            "step": 22305
+        },
+        {
+            "loss": 13.5643,
+            "grad_norm": 1.1612070798873901,
+            "learning_rate": 3.962234848359225e-09,
+            "epoch": 0.9998935568266489,
+            "step": 22310
+        },
+        {
+            "train_runtime": 132800.8636,
+            "train_samples_per_second": 301.079,
+            "train_steps_per_second": 0.168,
+            "total_flos": 1.364299850107899e+19,
+            "train_loss": 14.025886625746066,
+            "epoch": 0.9999831931831551,
+            "step": 22312
+        },
+        {
+            "eval_loss": 1.681868314743042,
+            "eval_runtime": 18.5175,
+            "eval_samples_per_second": 884.782,
+            "eval_steps_per_second": 7.938,
+            "epoch": 0.9999831931831551,
+            "step": 22312
+        }
+    ],
+    "best_metric": null,
+    "best_model_checkpoint": null,
+    "is_local_process_zero": true,
+    "is_world_process_zero": true,
+    "is_hyper_param_search": false,
+    "trial_name": null,
+    "trial_params": null,
+    "stateful_callbacks": {
+        "TrainerControl": {
+            "args": {
+                "should_training_stop": true,
+                "should_epoch_stop": false,
+                "should_save": true,
+                "should_evaluate": false,
+                "should_log": false
+            },
+            "attributes": {}
+        }
+    }
+}
\ No newline at end of file