{
  "best_metric": 0.027184655889868736,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold7/checkpoint-1800",
  "epoch": 4.9961802902979375,
  "eval_steps": 50,
  "global_step": 3270,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015278838808250574,
      "grad_norm": 1.848912000656128,
      "learning_rate": 3.0581039755351682e-06,
      "loss": 1.7007,
      "step": 10
    },
    {
      "epoch": 0.030557677616501147,
      "grad_norm": 1.845645785331726,
      "learning_rate": 6.1162079510703365e-06,
      "loss": 1.6546,
      "step": 20
    },
    {
      "epoch": 0.04583651642475172,
      "grad_norm": 2.5136754512786865,
      "learning_rate": 9.174311926605506e-06,
      "loss": 1.5461,
      "step": 30
    },
    {
      "epoch": 0.061115355233002294,
      "grad_norm": 3.130951404571533,
      "learning_rate": 1.2232415902140673e-05,
      "loss": 1.3083,
      "step": 40
    },
    {
      "epoch": 0.07639419404125286,
      "grad_norm": 0.9243296384811401,
      "learning_rate": 1.5290519877675842e-05,
      "loss": 0.8412,
      "step": 50
    },
    {
      "epoch": 0.07639419404125286,
      "eval_loss": 0.6190279722213745,
      "eval_runtime": 219.3498,
      "eval_samples_per_second": 5.307,
      "eval_steps_per_second": 5.307,
      "step": 50
    },
    {
      "epoch": 0.09167303284950344,
      "grad_norm": 1.0339502096176147,
      "learning_rate": 1.834862385321101e-05,
      "loss": 0.6464,
      "step": 60
    },
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 1.244530200958252,
      "learning_rate": 2.140672782874618e-05,
      "loss": 0.4868,
      "step": 70
    },
    {
      "epoch": 0.12223071046600459,
      "grad_norm": 0.8109800815582275,
      "learning_rate": 2.4464831804281346e-05,
      "loss": 0.3051,
      "step": 80
    },
    {
      "epoch": 0.13750954927425516,
      "grad_norm": 0.551679253578186,
      "learning_rate": 2.7522935779816515e-05,
      "loss": 0.1913,
      "step": 90
    },
    {
      "epoch": 0.15278838808250572,
      "grad_norm": 0.790469765663147,
      "learning_rate": 3.0581039755351684e-05,
      "loss": 0.1455,
      "step": 100
    },
    {
      "epoch": 0.15278838808250572,
      "eval_loss": 0.10693324357271194,
      "eval_runtime": 219.1206,
      "eval_samples_per_second": 5.312,
      "eval_steps_per_second": 5.312,
      "step": 100
    },
    {
      "epoch": 0.16806722689075632,
      "grad_norm": 0.5916261672973633,
      "learning_rate": 3.363914373088685e-05,
      "loss": 0.1072,
      "step": 110
    },
    {
      "epoch": 0.18334606569900688,
      "grad_norm": 0.5760579705238342,
      "learning_rate": 3.669724770642202e-05,
      "loss": 0.096,
      "step": 120
    },
    {
      "epoch": 0.19862490450725745,
      "grad_norm": 0.6538326740264893,
      "learning_rate": 3.9755351681957185e-05,
      "loss": 0.0842,
      "step": 130
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 0.6428112983703613,
      "learning_rate": 4.281345565749236e-05,
      "loss": 0.0834,
      "step": 140
    },
    {
      "epoch": 0.22918258212375858,
      "grad_norm": 0.6178712248802185,
      "learning_rate": 4.587155963302753e-05,
      "loss": 0.0861,
      "step": 150
    },
    {
      "epoch": 0.22918258212375858,
      "eval_loss": 0.06472994387149811,
      "eval_runtime": 218.5605,
      "eval_samples_per_second": 5.326,
      "eval_steps_per_second": 5.326,
      "step": 150
    },
    {
      "epoch": 0.24446142093200918,
      "grad_norm": 0.7672821283340454,
      "learning_rate": 4.892966360856269e-05,
      "loss": 0.0823,
      "step": 160
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.593714714050293,
      "learning_rate": 5.1987767584097854e-05,
      "loss": 0.0676,
      "step": 170
    },
    {
      "epoch": 0.2750190985485103,
      "grad_norm": 0.5263367295265198,
      "learning_rate": 5.504587155963303e-05,
      "loss": 0.0731,
      "step": 180
    },
    {
      "epoch": 0.2902979373567609,
      "grad_norm": 0.8246402740478516,
      "learning_rate": 5.81039755351682e-05,
      "loss": 0.0799,
      "step": 190
    },
    {
      "epoch": 0.30557677616501144,
      "grad_norm": 0.7266839742660522,
      "learning_rate": 6.116207951070337e-05,
      "loss": 0.0575,
      "step": 200
    },
    {
      "epoch": 0.30557677616501144,
      "eval_loss": 0.05183817818760872,
      "eval_runtime": 217.88,
      "eval_samples_per_second": 5.342,
      "eval_steps_per_second": 5.342,
      "step": 200
    },
    {
      "epoch": 0.32085561497326204,
      "grad_norm": 0.5840321779251099,
      "learning_rate": 6.422018348623854e-05,
      "loss": 0.0501,
      "step": 210
    },
    {
      "epoch": 0.33613445378151263,
      "grad_norm": 0.6017265915870667,
      "learning_rate": 6.72782874617737e-05,
      "loss": 0.0637,
      "step": 220
    },
    {
      "epoch": 0.35141329258976317,
      "grad_norm": 0.6707488894462585,
      "learning_rate": 7.033639143730886e-05,
      "loss": 0.057,
      "step": 230
    },
    {
      "epoch": 0.36669213139801377,
      "grad_norm": 0.41309818625450134,
      "learning_rate": 7.339449541284404e-05,
      "loss": 0.0494,
      "step": 240
    },
    {
      "epoch": 0.3819709702062643,
      "grad_norm": 0.5365416407585144,
      "learning_rate": 7.645259938837921e-05,
      "loss": 0.0643,
      "step": 250
    },
    {
      "epoch": 0.3819709702062643,
      "eval_loss": 0.04688103497028351,
      "eval_runtime": 218.0282,
      "eval_samples_per_second": 5.339,
      "eval_steps_per_second": 5.339,
      "step": 250
    },
    {
      "epoch": 0.3972498090145149,
      "grad_norm": 0.2568855285644531,
      "learning_rate": 7.951070336391437e-05,
      "loss": 0.0436,
      "step": 260
    },
    {
      "epoch": 0.4125286478227655,
      "grad_norm": 0.808674693107605,
      "learning_rate": 8.256880733944955e-05,
      "loss": 0.0708,
      "step": 270
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 0.32300978899002075,
      "learning_rate": 8.562691131498472e-05,
      "loss": 0.0634,
      "step": 280
    },
    {
      "epoch": 0.4430863254392666,
      "grad_norm": 0.24263863265514374,
      "learning_rate": 8.868501529051988e-05,
      "loss": 0.0507,
      "step": 290
    },
    {
      "epoch": 0.45836516424751717,
      "grad_norm": 0.45118218660354614,
      "learning_rate": 9.174311926605506e-05,
      "loss": 0.0341,
      "step": 300
    },
    {
      "epoch": 0.45836516424751717,
      "eval_loss": 0.043478548526763916,
      "eval_runtime": 218.6129,
      "eval_samples_per_second": 5.324,
      "eval_steps_per_second": 5.324,
      "step": 300
    },
    {
      "epoch": 0.47364400305576776,
      "grad_norm": 0.44334620237350464,
      "learning_rate": 9.480122324159021e-05,
      "loss": 0.0361,
      "step": 310
    },
    {
      "epoch": 0.48892284186401835,
      "grad_norm": 0.27677738666534424,
      "learning_rate": 9.785932721712538e-05,
      "loss": 0.0466,
      "step": 320
    },
    {
      "epoch": 0.5042016806722689,
      "grad_norm": 0.4536680281162262,
      "learning_rate": 9.999974360983129e-05,
      "loss": 0.0556,
      "step": 330
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.4593944549560547,
      "learning_rate": 9.999518563553522e-05,
      "loss": 0.0652,
      "step": 340
    },
    {
      "epoch": 0.5347593582887701,
      "grad_norm": 0.290985107421875,
      "learning_rate": 9.998493069976636e-05,
      "loss": 0.0641,
      "step": 350
    },
    {
      "epoch": 0.5347593582887701,
      "eval_loss": 0.041326772421598434,
      "eval_runtime": 218.3999,
      "eval_samples_per_second": 5.33,
      "eval_steps_per_second": 5.33,
      "step": 350
    },
    {
      "epoch": 0.5500381970970206,
      "grad_norm": 0.4909442663192749,
      "learning_rate": 9.99689799710767e-05,
      "loss": 0.0574,
      "step": 360
    },
    {
      "epoch": 0.5653170359052712,
      "grad_norm": 0.5292761921882629,
      "learning_rate": 9.994733526705501e-05,
      "loss": 0.0535,
      "step": 370
    },
    {
      "epoch": 0.5805958747135218,
      "grad_norm": 0.3746335804462433,
      "learning_rate": 9.991999905411966e-05,
      "loss": 0.0432,
      "step": 380
    },
    {
      "epoch": 0.5958747135217723,
      "grad_norm": 0.18307775259017944,
      "learning_rate": 9.988697444723762e-05,
      "loss": 0.0345,
      "step": 390
    },
    {
      "epoch": 0.6111535523300229,
      "grad_norm": 0.22528858482837677,
      "learning_rate": 9.984826520956949e-05,
      "loss": 0.0405,
      "step": 400
    },
    {
      "epoch": 0.6111535523300229,
      "eval_loss": 0.04191767796874046,
      "eval_runtime": 218.4167,
      "eval_samples_per_second": 5.329,
      "eval_steps_per_second": 5.329,
      "step": 400
    },
    {
      "epoch": 0.6264323911382735,
      "grad_norm": 0.299173504114151,
      "learning_rate": 9.980387575204072e-05,
      "loss": 0.0363,
      "step": 410
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 0.5245795249938965,
      "learning_rate": 9.975381113283891e-05,
      "loss": 0.0346,
      "step": 420
    },
    {
      "epoch": 0.6569900687547746,
      "grad_norm": 0.3329148292541504,
      "learning_rate": 9.969807705683751e-05,
      "loss": 0.0441,
      "step": 430
    },
    {
      "epoch": 0.6722689075630253,
      "grad_norm": 0.23399081826210022,
      "learning_rate": 9.96366798749457e-05,
      "loss": 0.0301,
      "step": 440
    },
    {
      "epoch": 0.6875477463712758,
      "grad_norm": 0.1943167895078659,
      "learning_rate": 9.956962658338473e-05,
      "loss": 0.0531,
      "step": 450
    },
    {
      "epoch": 0.6875477463712758,
      "eval_loss": 0.03845739737153053,
      "eval_runtime": 217.3338,
      "eval_samples_per_second": 5.356,
      "eval_steps_per_second": 5.356,
      "step": 450
    },
    {
      "epoch": 0.7028265851795263,
      "grad_norm": 0.2697027027606964,
      "learning_rate": 9.94969248228907e-05,
      "loss": 0.0363,
      "step": 460
    },
    {
      "epoch": 0.7181054239877769,
      "grad_norm": 0.3219946026802063,
      "learning_rate": 9.941858287784383e-05,
      "loss": 0.0447,
      "step": 470
    },
    {
      "epoch": 0.7333842627960275,
      "grad_norm": 0.4683067798614502,
      "learning_rate": 9.933460967532453e-05,
      "loss": 0.043,
      "step": 480
    },
    {
      "epoch": 0.7486631016042781,
      "grad_norm": 0.3924793004989624,
      "learning_rate": 9.924501478409618e-05,
      "loss": 0.0468,
      "step": 490
    },
    {
      "epoch": 0.7639419404125286,
      "grad_norm": 0.25234106183052063,
      "learning_rate": 9.914980841351465e-05,
      "loss": 0.041,
      "step": 500
    },
    {
      "epoch": 0.7639419404125286,
      "eval_loss": 0.03724617511034012,
      "eval_runtime": 213.7645,
      "eval_samples_per_second": 5.445,
      "eval_steps_per_second": 5.445,
      "step": 500
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 0.21722671389579773,
      "learning_rate": 9.904900141236506e-05,
      "loss": 0.0254,
      "step": 510
    },
    {
      "epoch": 0.7944996180290298,
      "grad_norm": 0.3712560534477234,
      "learning_rate": 9.894260526762548e-05,
      "loss": 0.0495,
      "step": 520
    },
    {
      "epoch": 0.8097784568372803,
      "grad_norm": 0.32290956377983093,
      "learning_rate": 9.883063210315804e-05,
      "loss": 0.0556,
      "step": 530
    },
    {
      "epoch": 0.825057295645531,
      "grad_norm": 0.397980272769928,
      "learning_rate": 9.871309467832738e-05,
      "loss": 0.0381,
      "step": 540
    },
    {
      "epoch": 0.8403361344537815,
      "grad_norm": 0.24909217655658722,
      "learning_rate": 9.859000638654674e-05,
      "loss": 0.0283,
      "step": 550
    },
    {
      "epoch": 0.8403361344537815,
      "eval_loss": 0.03525731340050697,
      "eval_runtime": 209.1141,
      "eval_samples_per_second": 5.566,
      "eval_steps_per_second": 5.566,
      "step": 550
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 0.2012615203857422,
      "learning_rate": 9.846138125375175e-05,
      "loss": 0.0501,
      "step": 560
    },
    {
      "epoch": 0.8708938120702827,
      "grad_norm": 0.19848772883415222,
      "learning_rate": 9.83272339368022e-05,
      "loss": 0.0298,
      "step": 570
    },
    {
      "epoch": 0.8861726508785333,
      "grad_norm": 0.27441200613975525,
      "learning_rate": 9.818757972181191e-05,
      "loss": 0.043,
      "step": 580
    },
    {
      "epoch": 0.9014514896867838,
      "grad_norm": 0.3285536766052246,
      "learning_rate": 9.804243452240675e-05,
      "loss": 0.0335,
      "step": 590
    },
    {
      "epoch": 0.9167303284950343,
      "grad_norm": 0.48698437213897705,
      "learning_rate": 9.789181487791146e-05,
      "loss": 0.041,
      "step": 600
    },
    {
      "epoch": 0.9167303284950343,
      "eval_loss": 0.03299807012081146,
      "eval_runtime": 207.055,
      "eval_samples_per_second": 5.622,
      "eval_steps_per_second": 5.622,
      "step": 600
    },
    {
      "epoch": 0.932009167303285,
      "grad_norm": 0.37265974283218384,
      "learning_rate": 9.773573795146485e-05,
      "loss": 0.044,
      "step": 610
    },
    {
      "epoch": 0.9472880061115355,
      "grad_norm": 0.22519999742507935,
      "learning_rate": 9.757422152806415e-05,
      "loss": 0.0284,
      "step": 620
    },
    {
      "epoch": 0.9625668449197861,
      "grad_norm": 0.24463610351085663,
      "learning_rate": 9.74072840125383e-05,
      "loss": 0.0417,
      "step": 630
    },
    {
      "epoch": 0.9778456837280367,
      "grad_norm": 0.3929882347583771,
      "learning_rate": 9.723494442745085e-05,
      "loss": 0.0395,
      "step": 640
    },
    {
      "epoch": 0.9931245225362872,
      "grad_norm": 0.40377989411354065,
      "learning_rate": 9.705722241093223e-05,
      "loss": 0.0553,
      "step": 650
    },
    {
      "epoch": 0.9931245225362872,
      "eval_loss": 0.0362694077193737,
      "eval_runtime": 205.7921,
      "eval_samples_per_second": 5.656,
      "eval_steps_per_second": 5.656,
      "step": 650
    },
    {
      "epoch": 1.0084033613445378,
      "grad_norm": 0.14314551651477814,
      "learning_rate": 9.687413821444199e-05,
      "loss": 0.0376,
      "step": 660
    },
    {
      "epoch": 1.0236822001527883,
      "grad_norm": 0.2880457937717438,
      "learning_rate": 9.668571270046122e-05,
      "loss": 0.0351,
      "step": 670
    },
    {
      "epoch": 1.0389610389610389,
      "grad_norm": 0.19249747693538666,
      "learning_rate": 9.649196734011519e-05,
      "loss": 0.0288,
      "step": 680
    },
    {
      "epoch": 1.0542398777692896,
      "grad_norm": 0.46998459100723267,
      "learning_rate": 9.629292421072671e-05,
      "loss": 0.0295,
      "step": 690
    },
    {
      "epoch": 1.0695187165775402,
      "grad_norm": 0.1609886735677719,
      "learning_rate": 9.608860599330048e-05,
      "loss": 0.0314,
      "step": 700
    },
    {
      "epoch": 1.0695187165775402,
      "eval_loss": 0.03102213703095913,
      "eval_runtime": 204.7184,
      "eval_samples_per_second": 5.686,
      "eval_steps_per_second": 5.686,
      "step": 700
    },
    {
      "epoch": 1.0847975553857907,
      "grad_norm": 0.2983687222003937,
      "learning_rate": 9.587903596993854e-05,
      "loss": 0.0297,
      "step": 710
    },
    {
      "epoch": 1.1000763941940412,
      "grad_norm": 0.3572332561016083,
      "learning_rate": 9.566423802118724e-05,
      "loss": 0.0301,
      "step": 720
    },
    {
      "epoch": 1.1153552330022918,
      "grad_norm": 0.10540549457073212,
      "learning_rate": 9.544423662331612e-05,
      "loss": 0.0438,
      "step": 730
    },
    {
      "epoch": 1.1306340718105423,
      "grad_norm": 0.332472026348114,
      "learning_rate": 9.521905684552877e-05,
      "loss": 0.0354,
      "step": 740
    },
    {
      "epoch": 1.1459129106187929,
      "grad_norm": 0.19438593089580536,
      "learning_rate": 9.498872434710623e-05,
      "loss": 0.0211,
      "step": 750
    },
    {
      "epoch": 1.1459129106187929,
      "eval_loss": 0.031196942552924156,
      "eval_runtime": 204.1433,
      "eval_samples_per_second": 5.702,
      "eval_steps_per_second": 5.702,
      "step": 750
    },
    {
      "epoch": 1.1611917494270436,
      "grad_norm": 0.22115331888198853,
      "learning_rate": 9.475326537448307e-05,
      "loss": 0.0285,
      "step": 760
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 0.16889172792434692,
      "learning_rate": 9.451270675825665e-05,
      "loss": 0.0287,
      "step": 770
    },
    {
      "epoch": 1.1917494270435447,
      "grad_norm": 0.330474853515625,
      "learning_rate": 9.426707591012976e-05,
      "loss": 0.0261,
      "step": 780
    },
    {
      "epoch": 1.2070282658517952,
      "grad_norm": 0.4934603273868561,
      "learning_rate": 9.4016400819787e-05,
      "loss": 0.0404,
      "step": 790
    },
    {
      "epoch": 1.2223071046600458,
      "grad_norm": 0.39667800068855286,
      "learning_rate": 9.376071005170539e-05,
      "loss": 0.0314,
      "step": 800
    },
    {
      "epoch": 1.2223071046600458,
      "eval_loss": 0.03202859312295914,
      "eval_runtime": 203.9785,
      "eval_samples_per_second": 5.706,
      "eval_steps_per_second": 5.706,
      "step": 800
    },
    {
      "epoch": 1.2375859434682965,
      "grad_norm": 0.2728429138660431,
      "learning_rate": 9.350003274189949e-05,
      "loss": 0.0361,
      "step": 810
    },
    {
      "epoch": 1.2528647822765469,
      "grad_norm": 0.23270060122013092,
      "learning_rate": 9.323439859460122e-05,
      "loss": 0.0312,
      "step": 820
    },
    {
      "epoch": 1.2681436210847976,
      "grad_norm": 0.21153134107589722,
      "learning_rate": 9.296383787887519e-05,
      "loss": 0.0311,
      "step": 830
    },
    {
      "epoch": 1.2834224598930482,
      "grad_norm": 0.18999841809272766,
      "learning_rate": 9.268838142516943e-05,
      "loss": 0.0405,
      "step": 840
    },
    {
      "epoch": 1.2987012987012987,
      "grad_norm": 0.28957971930503845,
      "learning_rate": 9.240806062180234e-05,
      "loss": 0.0325,
      "step": 850
    },
    {
      "epoch": 1.2987012987012987,
      "eval_loss": 0.03146594762802124,
      "eval_runtime": 203.4351,
      "eval_samples_per_second": 5.722,
      "eval_steps_per_second": 5.722,
      "step": 850
    },
    {
      "epoch": 1.3139801375095492,
      "grad_norm": 0.12776818871498108,
      "learning_rate": 9.212290741138592e-05,
      "loss": 0.0172,
      "step": 860
    },
    {
      "epoch": 1.3292589763177998,
      "grad_norm": 0.1620066910982132,
      "learning_rate": 9.183295428718592e-05,
      "loss": 0.0285,
      "step": 870
    },
    {
      "epoch": 1.3445378151260505,
      "grad_norm": 0.2502717971801758,
      "learning_rate": 9.153823428941924e-05,
      "loss": 0.0395,
      "step": 880
    },
    {
      "epoch": 1.359816653934301,
      "grad_norm": 0.23712791502475739,
      "learning_rate": 9.1238781001489e-05,
      "loss": 0.0276,
      "step": 890
    },
    {
      "epoch": 1.3750954927425516,
      "grad_norm": 0.27235275506973267,
      "learning_rate": 9.093462854615766e-05,
      "loss": 0.0351,
      "step": 900
    },
    {
      "epoch": 1.3750954927425516,
      "eval_loss": 0.030502911657094955,
      "eval_runtime": 203.3489,
      "eval_samples_per_second": 5.724,
      "eval_steps_per_second": 5.724,
      "step": 900
    },
    {
      "epoch": 1.3903743315508021,
      "grad_norm": 0.2629202604293823,
      "learning_rate": 9.062581158165876e-05,
      "loss": 0.0323,
      "step": 910
    },
    {
      "epoch": 1.4056531703590527,
      "grad_norm": 0.2535449266433716,
      "learning_rate": 9.031236529774764e-05,
      "loss": 0.0222,
      "step": 920
    },
    {
      "epoch": 1.4209320091673032,
      "grad_norm": 0.19801892340183258,
      "learning_rate": 8.999432541169145e-05,
      "loss": 0.0273,
      "step": 930
    },
    {
      "epoch": 1.4362108479755538,
      "grad_norm": 0.19468438625335693,
      "learning_rate": 8.967172816419927e-05,
      "loss": 0.0274,
      "step": 940
    },
    {
      "epoch": 1.4514896867838045,
      "grad_norm": 0.268258273601532,
      "learning_rate": 8.934461031529242e-05,
      "loss": 0.0402,
      "step": 950
    },
    {
      "epoch": 1.4514896867838045,
      "eval_loss": 0.03139554709196091,
      "eval_runtime": 203.1165,
      "eval_samples_per_second": 5.731,
      "eval_steps_per_second": 5.731,
      "step": 950
    },
    {
      "epoch": 1.466768525592055,
      "grad_norm": 0.21832090616226196,
      "learning_rate": 8.901300914011569e-05,
      "loss": 0.0308,
      "step": 960
    },
    {
      "epoch": 1.4820473644003056,
      "grad_norm": 0.12288864701986313,
      "learning_rate": 8.867696242468976e-05,
      "loss": 0.0228,
      "step": 970
    },
    {
      "epoch": 1.4973262032085561,
      "grad_norm": 0.2422870695590973,
      "learning_rate": 8.833650846160555e-05,
      "loss": 0.0321,
      "step": 980
    },
    {
      "epoch": 1.5126050420168067,
      "grad_norm": 0.20795999467372894,
      "learning_rate": 8.79916860456607e-05,
      "loss": 0.0269,
      "step": 990
    },
    {
      "epoch": 1.5278838808250574,
      "grad_norm": 0.13397260010242462,
      "learning_rate": 8.7642534469439e-05,
      "loss": 0.0262,
      "step": 1000
    },
    {
      "epoch": 1.5278838808250574,
      "eval_loss": 0.029931625351309776,
      "eval_runtime": 202.6566,
      "eval_samples_per_second": 5.744,
      "eval_steps_per_second": 5.744,
      "step": 1000
    },
    {
      "epoch": 1.5431627196333078,
      "grad_norm": 0.5128399729728699,
      "learning_rate": 8.728909351883283e-05,
      "loss": 0.0295,
      "step": 1010
    },
    {
      "epoch": 1.5584415584415585,
      "grad_norm": 0.24673043191432953,
      "learning_rate": 8.693140346850975e-05,
      "loss": 0.0379,
      "step": 1020
    },
    {
      "epoch": 1.573720397249809,
      "grad_norm": 0.16341492533683777,
      "learning_rate": 8.656950507732303e-05,
      "loss": 0.0247,
      "step": 1030
    },
    {
      "epoch": 1.5889992360580596,
      "grad_norm": 0.11685626953840256,
      "learning_rate": 8.620343958366718e-05,
      "loss": 0.0284,
      "step": 1040
    },
    {
      "epoch": 1.6042780748663101,
      "grad_norm": 0.25839105248451233,
      "learning_rate": 8.5833248700779e-05,
      "loss": 0.026,
      "step": 1050
    },
    {
      "epoch": 1.6042780748663101,
      "eval_loss": 0.030233880504965782,
      "eval_runtime": 202.0854,
      "eval_samples_per_second": 5.76,
      "eval_steps_per_second": 5.76,
      "step": 1050
    },
    {
      "epoch": 1.6195569136745607,
      "grad_norm": 0.12446019053459167,
      "learning_rate": 8.545897461198413e-05,
      "loss": 0.0269,
      "step": 1060
    },
    {
      "epoch": 1.6348357524828114,
      "grad_norm": 0.13764119148254395,
      "learning_rate": 8.508065996589036e-05,
      "loss": 0.0295,
      "step": 1070
    },
    {
      "epoch": 1.6501145912910617,
      "grad_norm": 0.1417740285396576,
      "learning_rate": 8.469834787152783e-05,
      "loss": 0.0235,
      "step": 1080
    },
    {
      "epoch": 1.6653934300993125,
      "grad_norm": 0.1833825260400772,
      "learning_rate": 8.43120818934367e-05,
      "loss": 0.0267,
      "step": 1090
    },
    {
      "epoch": 1.680672268907563,
      "grad_norm": 0.259737104177475,
      "learning_rate": 8.392190604670293e-05,
      "loss": 0.024,
      "step": 1100
    },
    {
      "epoch": 1.680672268907563,
      "eval_loss": 0.031445007771253586,
      "eval_runtime": 201.7681,
      "eval_samples_per_second": 5.769,
      "eval_steps_per_second": 5.769,
      "step": 1100
    },
    {
      "epoch": 1.6959511077158136,
      "grad_norm": 0.3111629784107208,
      "learning_rate": 8.352786479194288e-05,
      "loss": 0.0351,
      "step": 1110
    },
    {
      "epoch": 1.7112299465240641,
      "grad_norm": 0.36554020643234253,
      "learning_rate": 8.313000303023688e-05,
      "loss": 0.0245,
      "step": 1120
    },
    {
      "epoch": 1.7265087853323147,
      "grad_norm": 0.18379883468151093,
      "learning_rate": 8.27283660980128e-05,
      "loss": 0.0371,
      "step": 1130
    },
    {
      "epoch": 1.7417876241405654,
      "grad_norm": 0.13640639185905457,
      "learning_rate": 8.232299976187999e-05,
      "loss": 0.0334,
      "step": 1140
    },
    {
      "epoch": 1.7570664629488157,
      "grad_norm": 0.205993190407753,
      "learning_rate": 8.191395021341408e-05,
      "loss": 0.0487,
      "step": 1150
    },
    {
      "epoch": 1.7570664629488157,
      "eval_loss": 0.03015304170548916,
      "eval_runtime": 201.7837,
      "eval_samples_per_second": 5.769,
      "eval_steps_per_second": 5.769,
      "step": 1150
    },
    {
      "epoch": 1.7723453017570665,
      "grad_norm": 0.30252721905708313,
      "learning_rate": 8.150126406389352e-05,
      "loss": 0.0299,
      "step": 1160
    },
    {
      "epoch": 1.787624140565317,
      "grad_norm": 0.21577638387680054,
      "learning_rate": 8.108498833898815e-05,
      "loss": 0.0315,
      "step": 1170
    },
    {
      "epoch": 1.8029029793735676,
      "grad_norm": 0.34054169058799744,
      "learning_rate": 8.066517047340066e-05,
      "loss": 0.0348,
      "step": 1180
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.06973261386156082,
      "learning_rate": 8.02418583054614e-05,
      "loss": 0.0213,
      "step": 1190
    },
    {
      "epoch": 1.8334606569900687,
      "grad_norm": 0.21064534783363342,
      "learning_rate": 7.981510007167719e-05,
      "loss": 0.0251,
      "step": 1200
    },
    {
      "epoch": 1.8334606569900687,
      "eval_loss": 0.030031004920601845,
      "eval_runtime": 201.7876,
      "eval_samples_per_second": 5.768,
      "eval_steps_per_second": 5.768,
      "step": 1200
    },
    {
      "epoch": 1.8487394957983194,
      "grad_norm": 0.3094477355480194,
      "learning_rate": 7.938494440123468e-05,
      "loss": 0.0295,
      "step": 1210
    },
    {
      "epoch": 1.8640183346065697,
      "grad_norm": 0.15358871221542358,
      "learning_rate": 7.895144031045918e-05,
      "loss": 0.0287,
      "step": 1220
    },
    {
      "epoch": 1.8792971734148205,
      "grad_norm": 0.14194321632385254,
      "learning_rate": 7.851463719722913e-05,
      "loss": 0.025,
      "step": 1230
    },
    {
      "epoch": 1.894576012223071,
      "grad_norm": 0.1547326147556305,
      "learning_rate": 7.80745848353473e-05,
      "loss": 0.0294,
      "step": 1240
    },
    {
      "epoch": 1.9098548510313216,
      "grad_norm": 0.42348071932792664,
      "learning_rate": 7.763133336886892e-05,
      "loss": 0.028,
      "step": 1250
    },
    {
      "epoch": 1.9098548510313216,
      "eval_loss": 0.031990259885787964,
      "eval_runtime": 201.9183,
      "eval_samples_per_second": 5.765,
      "eval_steps_per_second": 5.765,
      "step": 1250
    },
    {
      "epoch": 1.9251336898395723,
      "grad_norm": 0.10496515035629272,
      "learning_rate": 7.718493330638789e-05,
      "loss": 0.0205,
      "step": 1260
    },
    {
      "epoch": 1.9404125286478227,
      "grad_norm": 0.07516542822122574,
      "learning_rate": 7.673543551528122e-05,
      "loss": 0.0346,
      "step": 1270
    },
    {
      "epoch": 1.9556913674560734,
      "grad_norm": 0.5417054891586304,
      "learning_rate": 7.628289121591277e-05,
      "loss": 0.0314,
      "step": 1280
    },
    {
      "epoch": 1.970970206264324,
      "grad_norm": 0.17793487012386322,
      "learning_rate": 7.582735197579656e-05,
      "loss": 0.0282,
      "step": 1290
    },
    {
      "epoch": 1.9862490450725745,
      "grad_norm": 0.1875259131193161,
      "learning_rate": 7.536886970372078e-05,
      "loss": 0.0244,
      "step": 1300
    },
    {
      "epoch": 1.9862490450725745,
      "eval_loss": 0.029929550364613533,
      "eval_runtime": 201.5395,
      "eval_samples_per_second": 5.776,
      "eval_steps_per_second": 5.776,
      "step": 1300
    },
    {
      "epoch": 2.0015278838808253,
      "grad_norm": 0.14435645937919617,
      "learning_rate": 7.490749664383271e-05,
      "loss": 0.0372,
      "step": 1310
    },
    {
      "epoch": 2.0168067226890756,
      "grad_norm": 0.09202131628990173,
      "learning_rate": 7.444328536968538e-05,
      "loss": 0.0258,
      "step": 1320
    },
    {
      "epoch": 2.0320855614973263,
      "grad_norm": 0.43767938017845154,
      "learning_rate": 7.397628877824701e-05,
      "loss": 0.024,
      "step": 1330
    },
    {
      "epoch": 2.0473644003055766,
      "grad_norm": 0.13443796336650848,
      "learning_rate": 7.350656008387327e-05,
      "loss": 0.0192,
      "step": 1340
    },
    {
      "epoch": 2.0626432391138274,
      "grad_norm": 0.21498188376426697,
      "learning_rate": 7.303415281224346e-05,
      "loss": 0.0211,
      "step": 1350
    },
    {
      "epoch": 2.0626432391138274,
      "eval_loss": 0.0282374769449234,
      "eval_runtime": 202.0409,
      "eval_samples_per_second": 5.761,
      "eval_steps_per_second": 5.761,
      "step": 1350
    },
    {
      "epoch": 2.0779220779220777,
      "grad_norm": 0.1617782562971115,
      "learning_rate": 7.255912079426136e-05,
      "loss": 0.0155,
      "step": 1360
    },
    {
      "epoch": 2.0932009167303285,
      "grad_norm": 0.07547878473997116,
      "learning_rate": 7.208151815992107e-05,
      "loss": 0.0172,
      "step": 1370
    },
    {
      "epoch": 2.1084797555385792,
      "grad_norm": 0.34869226813316345,
      "learning_rate": 7.160139933213898e-05,
      "loss": 0.0254,
      "step": 1380
    },
    {
      "epoch": 2.1237585943468296,
      "grad_norm": 0.2867926061153412,
      "learning_rate": 7.111881902055223e-05,
      "loss": 0.0226,
      "step": 1390
    },
    {
      "epoch": 2.1390374331550803,
      "grad_norm": 0.26209303736686707,
      "learning_rate": 7.06338322152845e-05,
      "loss": 0.019,
      "step": 1400
    },
    {
      "epoch": 2.1390374331550803,
      "eval_loss": 0.02854863740503788,
      "eval_runtime": 201.0589,
      "eval_samples_per_second": 5.789,
      "eval_steps_per_second": 5.789,
      "step": 1400
    },
    {
      "epoch": 2.1543162719633306,
      "grad_norm": 0.17289575934410095,
      "learning_rate": 7.014649418067994e-05,
      "loss": 0.0176,
      "step": 1410
    },
    {
      "epoch": 2.1695951107715814,
      "grad_norm": 0.22034431993961334,
      "learning_rate": 6.965686044900577e-05,
      "loss": 0.0163,
      "step": 1420
    },
    {
      "epoch": 2.184873949579832,
      "grad_norm": 0.2867283821105957,
      "learning_rate": 6.91649868141243e-05,
      "loss": 0.0154,
      "step": 1430
    },
    {
      "epoch": 2.2001527883880825,
      "grad_norm": 0.17423681914806366,
      "learning_rate": 6.86709293251353e-05,
      "loss": 0.0142,
      "step": 1440
    },
    {
      "epoch": 2.2154316271963332,
      "grad_norm": 0.10424409806728363,
      "learning_rate": 6.817474427998916e-05,
      "loss": 0.012,
      "step": 1450
    },
    {
      "epoch": 2.2154316271963332,
      "eval_loss": 0.030223889276385307,
      "eval_runtime": 200.9988,
      "eval_samples_per_second": 5.791,
      "eval_steps_per_second": 5.791,
      "step": 1450
    },
    {
      "epoch": 2.2307104660045836,
      "grad_norm": 0.325139582157135,
      "learning_rate": 6.767648821907172e-05,
      "loss": 0.0192,
      "step": 1460
    },
    {
      "epoch": 2.2459893048128343,
      "grad_norm": 0.11585790663957596,
      "learning_rate": 6.717621791876147e-05,
      "loss": 0.0302,
      "step": 1470
    },
    {
      "epoch": 2.2612681436210846,
      "grad_norm": 0.16039828956127167,
      "learning_rate": 6.667399038495986e-05,
      "loss": 0.0207,
      "step": 1480
    },
    {
      "epoch": 2.2765469824293354,
      "grad_norm": 0.18742448091506958,
      "learning_rate": 6.616986284659556e-05,
      "loss": 0.0247,
      "step": 1490
    },
    {
      "epoch": 2.2918258212375857,
      "grad_norm": 0.07378096133470535,
      "learning_rate": 6.566389274910309e-05,
      "loss": 0.0181,
      "step": 1500
    },
    {
      "epoch": 2.2918258212375857,
      "eval_loss": 0.02827996388077736,
      "eval_runtime": 201.0916,
      "eval_samples_per_second": 5.788,
      "eval_steps_per_second": 5.788,
      "step": 1500
    },
    {
      "epoch": 2.3071046600458365,
      "grad_norm": 0.10507321357727051,
      "learning_rate": 6.515613774787697e-05,
      "loss": 0.019,
      "step": 1510
    },
    {
      "epoch": 2.3223834988540872,
      "grad_norm": 0.10831980407238007,
      "learning_rate": 6.464665570170186e-05,
      "loss": 0.0215,
      "step": 1520
    },
    {
      "epoch": 2.3376623376623376,
      "grad_norm": 0.10839013010263443,
      "learning_rate": 6.413550466615952e-05,
      "loss": 0.017,
      "step": 1530
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 0.12335723638534546,
      "learning_rate": 6.362274288701342e-05,
      "loss": 0.0174,
      "step": 1540
    },
    {
      "epoch": 2.3682200152788386,
      "grad_norm": 0.16996541619300842,
      "learning_rate": 6.310842879357158e-05,
      "loss": 0.0176,
      "step": 1550
    },
    {
      "epoch": 2.3682200152788386,
      "eval_loss": 0.028811601921916008,
      "eval_runtime": 201.2139,
      "eval_samples_per_second": 5.785,
      "eval_steps_per_second": 5.785,
      "step": 1550
    },
    {
      "epoch": 2.3834988540870894,
      "grad_norm": 0.22573326528072357,
      "learning_rate": 6.25926209920285e-05,
      "loss": 0.0117,
      "step": 1560
    },
    {
      "epoch": 2.39877769289534,
      "grad_norm": 0.24418701231479645,
      "learning_rate": 6.207537825878708e-05,
      "loss": 0.0165,
      "step": 1570
    },
    {
      "epoch": 2.4140565317035905,
      "grad_norm": 0.15098631381988525,
      "learning_rate": 6.155675953376095e-05,
      "loss": 0.027,
      "step": 1580
    },
    {
      "epoch": 2.4293353705118412,
      "grad_norm": 0.1565188765525818,
      "learning_rate": 6.103682391365828e-05,
      "loss": 0.0185,
      "step": 1590
    },
    {
      "epoch": 2.4446142093200915,
      "grad_norm": 0.0722580999135971,
      "learning_rate": 6.05156306452477e-05,
      "loss": 0.0136,
      "step": 1600
    },
    {
      "epoch": 2.4446142093200915,
      "eval_loss": 0.027694666758179665,
      "eval_runtime": 201.4813,
      "eval_samples_per_second": 5.777,
      "eval_steps_per_second": 5.777,
      "step": 1600
    },
    {
      "epoch": 2.4598930481283423,
      "grad_norm": 0.2664923667907715,
      "learning_rate": 5.9993239118607124e-05,
      "loss": 0.0251,
      "step": 1610
    },
    {
      "epoch": 2.475171886936593,
      "grad_norm": 0.15546078979969025,
      "learning_rate": 5.9469708860356246e-05,
      "loss": 0.0207,
      "step": 1620
    },
    {
      "epoch": 2.4904507257448434,
      "grad_norm": 0.14643099904060364,
      "learning_rate": 5.89450995268734e-05,
      "loss": 0.0195,
      "step": 1630
    },
    {
      "epoch": 2.5057295645530937,
      "grad_norm": 0.1590198427438736,
      "learning_rate": 5.841947089749783e-05,
      "loss": 0.0251,
      "step": 1640
    },
    {
      "epoch": 2.5210084033613445,
      "grad_norm": 0.19299504160881042,
      "learning_rate": 5.78928828677177e-05,
      "loss": 0.0217,
      "step": 1650
    },
    {
      "epoch": 2.5210084033613445,
      "eval_loss": 0.028578663244843483,
      "eval_runtime": 201.0706,
      "eval_samples_per_second": 5.789,
      "eval_steps_per_second": 5.789,
      "step": 1650
    },
    {
      "epoch": 2.5362872421695952,
      "grad_norm": 0.18513061106204987,
      "learning_rate": 5.7365395442345085e-05,
      "loss": 0.0143,
      "step": 1660
    },
    {
      "epoch": 2.5515660809778455,
      "grad_norm": 0.1994353085756302,
      "learning_rate": 5.683706872867833e-05,
      "loss": 0.0149,
      "step": 1670
    },
    {
      "epoch": 2.5668449197860963,
      "grad_norm": 0.20116868615150452,
      "learning_rate": 5.630796292965288e-05,
      "loss": 0.0255,
      "step": 1680
    },
    {
      "epoch": 2.5821237585943466,
      "grad_norm": 0.1676681637763977,
      "learning_rate": 5.57781383369811e-05,
      "loss": 0.02,
      "step": 1690
    },
    {
      "epoch": 2.5974025974025974,
      "grad_norm": 0.1904657781124115,
      "learning_rate": 5.524765532428203e-05,
      "loss": 0.0156,
      "step": 1700
    },
    {
      "epoch": 2.5974025974025974,
      "eval_loss": 0.029410608112812042,
      "eval_runtime": 200.9522,
      "eval_samples_per_second": 5.792,
      "eval_steps_per_second": 5.792,
      "step": 1700
    },
    {
      "epoch": 2.612681436210848,
      "grad_norm": 0.15733115375041962,
      "learning_rate": 5.471657434020182e-05,
      "loss": 0.0187,
      "step": 1710
    },
    {
      "epoch": 2.6279602750190985,
      "grad_norm": 0.20242205262184143,
      "learning_rate": 5.418495590152557e-05,
      "loss": 0.0231,
      "step": 1720
    },
    {
      "epoch": 2.643239113827349,
      "grad_norm": 0.1593477725982666,
      "learning_rate": 5.365286058628145e-05,
      "loss": 0.0212,
      "step": 1730
    },
    {
      "epoch": 2.6585179526355995,
      "grad_norm": 0.3499566912651062,
      "learning_rate": 5.312034902683779e-05,
      "loss": 0.0268,
      "step": 1740
    },
    {
      "epoch": 2.6737967914438503,
      "grad_norm": 0.1606576293706894,
      "learning_rate": 5.258748190299404e-05,
      "loss": 0.0191,
      "step": 1750
    },
    {
      "epoch": 2.6737967914438503,
      "eval_loss": 0.028579337522387505,
      "eval_runtime": 200.7556,
      "eval_samples_per_second": 5.798,
      "eval_steps_per_second": 5.798,
      "step": 1750
    },
    {
      "epoch": 2.689075630252101,
      "grad_norm": 0.26182451844215393,
      "learning_rate": 5.20543199350663e-05,
      "loss": 0.0217,
      "step": 1760
    },
    {
      "epoch": 2.7043544690603514,
      "grad_norm": 0.17532721161842346,
      "learning_rate": 5.152092387696821e-05,
      "loss": 0.0215,
      "step": 1770
    },
    {
      "epoch": 2.719633307868602,
      "grad_norm": 0.23950158059597015,
      "learning_rate": 5.0987354509287985e-05,
      "loss": 0.0224,
      "step": 1780
    },
    {
      "epoch": 2.7349121466768525,
      "grad_norm": 0.3977627754211426,
      "learning_rate": 5.045367263236257e-05,
      "loss": 0.017,
      "step": 1790
    },
    {
      "epoch": 2.750190985485103,
      "grad_norm": 0.10810764133930206,
      "learning_rate": 4.991993905934931e-05,
      "loss": 0.0249,
      "step": 1800
    },
    {
      "epoch": 2.750190985485103,
      "eval_loss": 0.027184655889868736,
      "eval_runtime": 200.6455,
      "eval_samples_per_second": 5.801,
      "eval_steps_per_second": 5.801,
      "step": 1800
    },
    {
      "epoch": 2.765469824293354,
      "grad_norm": 0.23987551033496857,
      "learning_rate": 4.938621460929639e-05,
      "loss": 0.0182,
      "step": 1810
    },
    {
      "epoch": 2.7807486631016043,
      "grad_norm": 0.2798612415790558,
      "learning_rate": 4.885256010021233e-05,
      "loss": 0.0193,
      "step": 1820
    },
    {
      "epoch": 2.7960275019098546,
      "grad_norm": 0.16607950627803802,
      "learning_rate": 4.831903634213599e-05,
      "loss": 0.0199,
      "step": 1830
    },
    {
      "epoch": 2.8113063407181054,
      "grad_norm": 0.04422255605459213,
      "learning_rate": 4.778570413020702e-05,
      "loss": 0.0211,
      "step": 1840
    },
    {
      "epoch": 2.826585179526356,
      "grad_norm": 0.24929310381412506,
      "learning_rate": 4.725262423773838e-05,
      "loss": 0.0237,
      "step": 1850
    },
    {
      "epoch": 2.826585179526356,
      "eval_loss": 0.02900101989507675,
      "eval_runtime": 201.0849,
      "eval_samples_per_second": 5.789,
      "eval_steps_per_second": 5.789,
      "step": 1850
    },
    {
      "epoch": 2.8418640183346064,
      "grad_norm": 0.2282838672399521,
      "learning_rate": 4.671985740929123e-05,
      "loss": 0.0165,
      "step": 1860
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.06340071558952332,
      "learning_rate": 4.618746435375295e-05,
      "loss": 0.023,
      "step": 1870
    },
    {
      "epoch": 2.8724216959511075,
      "grad_norm": 0.04884711280465126,
      "learning_rate": 4.565550573741942e-05,
      "loss": 0.0159,
      "step": 1880
    },
    {
      "epoch": 2.8877005347593583,
      "grad_norm": 0.24351294338703156,
      "learning_rate": 4.512404217708217e-05,
      "loss": 0.0227,
      "step": 1890
    },
    {
      "epoch": 2.902979373567609,
      "grad_norm": 0.12852653861045837,
      "learning_rate": 4.45931342331209e-05,
      "loss": 0.021,
      "step": 1900
    },
    {
      "epoch": 2.902979373567609,
      "eval_loss": 0.02777683548629284,
      "eval_runtime": 200.6223,
      "eval_samples_per_second": 5.802,
      "eval_steps_per_second": 5.802,
      "step": 1900
    },
    {
      "epoch": 2.9182582123758594,
      "grad_norm": 0.1962100714445114,
      "learning_rate": 4.406284240260278e-05,
      "loss": 0.023,
      "step": 1910
    },
    {
      "epoch": 2.93353705118411,
      "grad_norm": 0.16242529451847076,
      "learning_rate": 4.3533227112388694e-05,
      "loss": 0.0165,
      "step": 1920
    },
    {
      "epoch": 2.9488158899923604,
      "grad_norm": 0.15565882623195648,
      "learning_rate": 4.300434871224763e-05,
      "loss": 0.0234,
      "step": 1930
    },
    {
      "epoch": 2.964094728800611,
      "grad_norm": 0.1874658316373825,
      "learning_rate": 4.247626746797983e-05,
      "loss": 0.0207,
      "step": 1940
    },
    {
      "epoch": 2.979373567608862,
      "grad_norm": 0.1462504267692566,
      "learning_rate": 4.1949043554549406e-05,
      "loss": 0.0174,
      "step": 1950
    },
    {
      "epoch": 2.979373567608862,
      "eval_loss": 0.028289297595620155,
      "eval_runtime": 200.8115,
      "eval_samples_per_second": 5.796,
      "eval_steps_per_second": 5.796,
      "step": 1950
    },
    {
      "epoch": 2.9946524064171123,
      "grad_norm": 0.3352892994880676,
      "learning_rate": 4.14227370492275e-05,
      "loss": 0.0261,
      "step": 1960
    },
    {
      "epoch": 3.009931245225363,
      "grad_norm": 0.15284225344657898,
      "learning_rate": 4.08974079247464e-05,
      "loss": 0.0172,
      "step": 1970
    },
    {
      "epoch": 3.0252100840336134,
      "grad_norm": 0.06005563586950302,
      "learning_rate": 4.037311604246565e-05,
      "loss": 0.0115,
      "step": 1980
    },
    {
      "epoch": 3.040488922841864,
      "grad_norm": 0.21838952600955963,
      "learning_rate": 3.9849921145550805e-05,
      "loss": 0.0129,
      "step": 1990
    },
    {
      "epoch": 3.0557677616501144,
      "grad_norm": 0.11845885962247849,
      "learning_rate": 3.9327882852165795e-05,
      "loss": 0.0122,
      "step": 2000
    },
    {
      "epoch": 3.0557677616501144,
      "eval_loss": 0.028973255306482315,
      "eval_runtime": 200.9411,
      "eval_samples_per_second": 5.793,
      "eval_steps_per_second": 5.793,
      "step": 2000
    },
    {
      "epoch": 3.071046600458365,
      "grad_norm": 0.09495562314987183,
      "learning_rate": 3.880706064867926e-05,
      "loss": 0.0125,
      "step": 2010
    },
    {
      "epoch": 3.0863254392666155,
      "grad_norm": 0.16284717619419098,
      "learning_rate": 3.8287513882886196e-05,
      "loss": 0.009,
      "step": 2020
    },
    {
      "epoch": 3.1016042780748663,
      "grad_norm": 0.1330808848142624,
      "learning_rate": 3.776930175724521e-05,
      "loss": 0.011,
      "step": 2030
    },
    {
      "epoch": 3.116883116883117,
      "grad_norm": 0.1752845197916031,
      "learning_rate": 3.7252483322132386e-05,
      "loss": 0.0128,
      "step": 2040
    },
    {
      "epoch": 3.1321619556913673,
      "grad_norm": 0.09726885706186295,
      "learning_rate": 3.673711746911252e-05,
      "loss": 0.0137,
      "step": 2050
    },
    {
      "epoch": 3.1321619556913673,
      "eval_loss": 0.030082957819104195,
      "eval_runtime": 200.9521,
      "eval_samples_per_second": 5.792,
      "eval_steps_per_second": 5.792,
      "step": 2050
    },
    {
      "epoch": 3.147440794499618,
      "grad_norm": 0.09027769416570663,
      "learning_rate": 3.6223262924228344e-05,
      "loss": 0.0078,
      "step": 2060
    },
    {
      "epoch": 3.1627196333078684,
      "grad_norm": 0.08661149442195892,
      "learning_rate": 3.5710978241308733e-05,
      "loss": 0.013,
      "step": 2070
    },
    {
      "epoch": 3.177998472116119,
      "grad_norm": 0.19561336934566498,
      "learning_rate": 3.520032179529652e-05,
      "loss": 0.0123,
      "step": 2080
    },
    {
      "epoch": 3.19327731092437,
      "grad_norm": 0.28915536403656006,
      "learning_rate": 3.4691351775596564e-05,
      "loss": 0.0135,
      "step": 2090
    },
    {
      "epoch": 3.2085561497326203,
      "grad_norm": 0.12461555749177933,
      "learning_rate": 3.41841261794451e-05,
      "loss": 0.0086,
      "step": 2100
    },
    {
      "epoch": 3.2085561497326203,
      "eval_loss": 0.030945168808102608,
      "eval_runtime": 200.348,
      "eval_samples_per_second": 5.81,
      "eval_steps_per_second": 5.81,
      "step": 2100
    },
    {
      "epoch": 3.223834988540871,
      "grad_norm": 0.17976081371307373,
      "learning_rate": 3.367870280530101e-05,
      "loss": 0.0176,
      "step": 2110
    },
    {
      "epoch": 3.2391138273491213,
      "grad_norm": 0.17633405327796936,
      "learning_rate": 3.3175139246259536e-05,
      "loss": 0.0109,
      "step": 2120
    },
    {
      "epoch": 3.254392666157372,
      "grad_norm": 0.23417995870113373,
      "learning_rate": 3.2673492883489696e-05,
      "loss": 0.0105,
      "step": 2130
    },
    {
      "epoch": 3.2696715049656224,
      "grad_norm": 0.385770320892334,
      "learning_rate": 3.2173820879695535e-05,
      "loss": 0.0166,
      "step": 2140
    },
    {
      "epoch": 3.284950343773873,
      "grad_norm": 0.1962185502052307,
      "learning_rate": 3.1676180172602525e-05,
      "loss": 0.0136,
      "step": 2150
    },
    {
      "epoch": 3.284950343773873,
      "eval_loss": 0.030590800568461418,
      "eval_runtime": 199.5546,
      "eval_samples_per_second": 5.833,
      "eval_steps_per_second": 5.833,
      "step": 2150
    },
    {
      "epoch": 3.300229182582124,
      "grad_norm": 0.07931424677371979,
      "learning_rate": 3.11806274684695e-05,
      "loss": 0.0115,
      "step": 2160
    },
    {
      "epoch": 3.3155080213903743,
      "grad_norm": 0.12415704876184464,
      "learning_rate": 3.068721923562688e-05,
      "loss": 0.0076,
      "step": 2170
    },
    {
      "epoch": 3.330786860198625,
      "grad_norm": 0.1858009248971939,
      "learning_rate": 3.019601169804216e-05,
      "loss": 0.0109,
      "step": 2180
    },
    {
      "epoch": 3.3460656990068753,
      "grad_norm": 0.15695886313915253,
      "learning_rate": 2.9707060828913225e-05,
      "loss": 0.0086,
      "step": 2190
    },
    {
      "epoch": 3.361344537815126,
      "grad_norm": 0.5040724277496338,
      "learning_rate": 2.9220422344290056e-05,
      "loss": 0.0111,
      "step": 2200
    },
    {
      "epoch": 3.361344537815126,
      "eval_loss": 0.03102887235581875,
      "eval_runtime": 200.5885,
      "eval_samples_per_second": 5.803,
      "eval_steps_per_second": 5.803,
      "step": 2200
    },
    {
      "epoch": 3.3766233766233764,
      "grad_norm": 0.2534728944301605,
      "learning_rate": 2.873615169672601e-05,
      "loss": 0.0099,
      "step": 2210
    },
    {
      "epoch": 3.391902215431627,
      "grad_norm": 0.10433147847652435,
      "learning_rate": 2.8254304068958927e-05,
      "loss": 0.011,
      "step": 2220
    },
    {
      "epoch": 3.407181054239878,
      "grad_norm": 0.09524795413017273,
      "learning_rate": 2.7774934367622996e-05,
      "loss": 0.0159,
      "step": 2230
    },
    {
      "epoch": 3.4224598930481283,
      "grad_norm": 0.1627880036830902,
      "learning_rate": 2.7298097216992284e-05,
      "loss": 0.0095,
      "step": 2240
    },
    {
      "epoch": 3.437738731856379,
      "grad_norm": 0.24664515256881714,
      "learning_rate": 2.6823846952756125e-05,
      "loss": 0.0142,
      "step": 2250
    },
    {
      "epoch": 3.437738731856379,
      "eval_loss": 0.03270566090941429,
      "eval_runtime": 200.5417,
      "eval_samples_per_second": 5.804,
      "eval_steps_per_second": 5.804,
      "step": 2250
    },
    {
      "epoch": 3.4530175706646293,
      "grad_norm": 0.2718111276626587,
      "learning_rate": 2.6352237615827636e-05,
      "loss": 0.0077,
      "step": 2260
    },
    {
      "epoch": 3.46829640947288,
      "grad_norm": 0.2555138170719147,
      "learning_rate": 2.5883322946185777e-05,
      "loss": 0.0129,
      "step": 2270
    },
    {
      "epoch": 3.483575248281131,
      "grad_norm": 0.23297752439975739,
      "learning_rate": 2.5417156376751562e-05,
      "loss": 0.0124,
      "step": 2280
    },
    {
      "epoch": 3.498854087089381,
      "grad_norm": 0.24713806807994843,
      "learning_rate": 2.4953791027299506e-05,
      "loss": 0.008,
      "step": 2290
    },
    {
      "epoch": 3.514132925897632,
      "grad_norm": 0.15914295613765717,
      "learning_rate": 2.4493279698404493e-05,
      "loss": 0.0114,
      "step": 2300
    },
    {
      "epoch": 3.514132925897632,
      "eval_loss": 0.031200075522065163,
      "eval_runtime": 200.6578,
      "eval_samples_per_second": 5.801,
      "eval_steps_per_second": 5.801,
      "step": 2300
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 0.14958317577838898,
      "learning_rate": 2.403567486542518e-05,
      "loss": 0.0107,
      "step": 2310
    },
    {
      "epoch": 3.544690603514133,
      "grad_norm": 0.10404475033283234,
      "learning_rate": 2.3581028672524485e-05,
      "loss": 0.0111,
      "step": 2320
    },
    {
      "epoch": 3.5599694423223838,
      "grad_norm": 0.11570907384157181,
      "learning_rate": 2.312939292672765e-05,
      "loss": 0.0088,
      "step": 2330
    },
    {
      "epoch": 3.575248281130634,
      "grad_norm": 0.1450202912092209,
      "learning_rate": 2.268081909201885e-05,
      "loss": 0.0087,
      "step": 2340
    },
    {
      "epoch": 3.5905271199388844,
      "grad_norm": 0.1596309393644333,
      "learning_rate": 2.2235358283476936e-05,
      "loss": 0.015,
      "step": 2350
    },
    {
      "epoch": 3.5905271199388844,
      "eval_loss": 0.031868599355220795,
      "eval_runtime": 200.4001,
      "eval_samples_per_second": 5.808,
      "eval_steps_per_second": 5.808,
      "step": 2350
    },
    {
      "epoch": 3.605805958747135,
      "grad_norm": 0.13130627572536469,
      "learning_rate": 2.179306126145075e-05,
      "loss": 0.0135,
      "step": 2360
    },
    {
      "epoch": 3.621084797555386,
      "grad_norm": 0.2653070390224457,
      "learning_rate": 2.1353978425775008e-05,
      "loss": 0.0129,
      "step": 2370
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.20412729680538177,
      "learning_rate": 2.091815981002731e-05,
      "loss": 0.0101,
      "step": 2380
    },
    {
      "epoch": 3.651642475171887,
      "grad_norm": 0.18179409205913544,
      "learning_rate": 2.0485655075826667e-05,
      "loss": 0.0141,
      "step": 2390
    },
    {
      "epoch": 3.6669213139801373,
      "grad_norm": 0.3348677158355713,
      "learning_rate": 2.0056513507174685e-05,
      "loss": 0.0088,
      "step": 2400
    },
    {
      "epoch": 3.6669213139801373,
      "eval_loss": 0.030020447447896004,
      "eval_runtime": 200.0438,
      "eval_samples_per_second": 5.819,
      "eval_steps_per_second": 5.819,
      "step": 2400
    },
    {
      "epoch": 3.682200152788388,
      "grad_norm": 0.25341227650642395,
      "learning_rate": 1.963078400483953e-05,
      "loss": 0.0111,
      "step": 2410
    },
    {
      "epoch": 3.697478991596639,
      "grad_norm": 0.2232055962085724,
      "learning_rate": 1.9208515080783723e-05,
      "loss": 0.0087,
      "step": 2420
    },
    {
      "epoch": 3.712757830404889,
      "grad_norm": 0.29884758591651917,
      "learning_rate": 1.8789754852636245e-05,
      "loss": 0.0084,
      "step": 2430
    },
    {
      "epoch": 3.72803666921314,
      "grad_norm": 0.2963765859603882,
      "learning_rate": 1.837455103820942e-05,
      "loss": 0.0173,
      "step": 2440
    },
    {
      "epoch": 3.7433155080213902,
      "grad_norm": 0.06125696748495102,
      "learning_rate": 1.7962950950061502e-05,
      "loss": 0.0068,
      "step": 2450
    },
    {
      "epoch": 3.7433155080213902,
      "eval_loss": 0.03097730502486229,
      "eval_runtime": 199.2868,
      "eval_samples_per_second": 5.841,
      "eval_steps_per_second": 5.841,
      "step": 2450
    },
    {
      "epoch": 3.758594346829641,
      "grad_norm": 0.330082505941391,
      "learning_rate": 1.7555001490105488e-05,
      "loss": 0.0103,
      "step": 2460
    },
    {
      "epoch": 3.7738731856378918,
      "grad_norm": 0.3047909438610077,
      "learning_rate": 1.7150749144264462e-05,
      "loss": 0.0123,
      "step": 2470
    },
    {
      "epoch": 3.789152024446142,
      "grad_norm": 0.09779535233974457,
      "learning_rate": 1.6750239977174682e-05,
      "loss": 0.0097,
      "step": 2480
    },
    {
      "epoch": 3.8044308632543924,
      "grad_norm": 0.3834091126918793,
      "learning_rate": 1.6353519626936397e-05,
      "loss": 0.0145,
      "step": 2490
    },
    {
      "epoch": 3.819709702062643,
      "grad_norm": 0.12687507271766663,
      "learning_rate": 1.596063329991341e-05,
      "loss": 0.0098,
      "step": 2500
    },
    {
      "epoch": 3.819709702062643,
      "eval_loss": 0.02996741607785225,
      "eval_runtime": 199.4136,
      "eval_samples_per_second": 5.837,
      "eval_steps_per_second": 5.837,
      "step": 2500
    },
    {
      "epoch": 3.834988540870894,
      "grad_norm": 0.10428875684738159,
      "learning_rate": 1.5571625765581832e-05,
      "loss": 0.0081,
      "step": 2510
    },
    {
      "epoch": 3.8502673796791442,
      "grad_norm": 0.2717987895011902,
      "learning_rate": 1.5186541351428545e-05,
      "loss": 0.012,
      "step": 2520
    },
    {
      "epoch": 3.865546218487395,
      "grad_norm": 0.6217739582061768,
      "learning_rate": 1.4805423937900087e-05,
      "loss": 0.0138,
      "step": 2530
    },
    {
      "epoch": 3.8808250572956453,
      "grad_norm": 0.14223617315292358,
      "learning_rate": 1.4428316953402526e-05,
      "loss": 0.0095,
      "step": 2540
    },
    {
      "epoch": 3.896103896103896,
      "grad_norm": 0.42163944244384766,
      "learning_rate": 1.4055263369352672e-05,
      "loss": 0.0088,
      "step": 2550
    },
    {
      "epoch": 3.896103896103896,
      "eval_loss": 0.02977598085999489,
      "eval_runtime": 199.4878,
      "eval_samples_per_second": 5.835,
      "eval_steps_per_second": 5.835,
      "step": 2550
    },
    {
      "epoch": 3.911382734912147,
      "grad_norm": 0.27528712153434753,
      "learning_rate": 1.3686305695281559e-05,
      "loss": 0.0111,
      "step": 2560
    },
    {
      "epoch": 3.926661573720397,
      "grad_norm": 0.07628080993890762,
      "learning_rate": 1.3321485973990494e-05,
      "loss": 0.0083,
      "step": 2570
    },
    {
      "epoch": 3.941940412528648,
      "grad_norm": 0.3497057855129242,
      "learning_rate": 1.2960845776760156e-05,
      "loss": 0.0164,
      "step": 2580
    },
    {
      "epoch": 3.9572192513368982,
      "grad_norm": 0.1905433088541031,
      "learning_rate": 1.2604426198613688e-05,
      "loss": 0.0089,
      "step": 2590
    },
    {
      "epoch": 3.972498090145149,
      "grad_norm": 0.29687440395355225,
      "learning_rate": 1.2252267853633798e-05,
      "loss": 0.0081,
      "step": 2600
    },
    {
      "epoch": 3.972498090145149,
      "eval_loss": 0.030624018982052803,
      "eval_runtime": 199.6483,
      "eval_samples_per_second": 5.83,
      "eval_steps_per_second": 5.83,
      "step": 2600
    },
    {
      "epoch": 3.9877769289533997,
      "grad_norm": 0.16313646733760834,
      "learning_rate": 1.1904410870334803e-05,
      "loss": 0.0132,
      "step": 2610
    },
    {
      "epoch": 4.0030557677616505,
      "grad_norm": 0.20121262967586517,
      "learning_rate": 1.1560894887090052e-05,
      "loss": 0.011,
      "step": 2620
    },
    {
      "epoch": 4.0183346065699,
      "grad_norm": 0.3056090772151947,
      "learning_rate": 1.1221759047615004e-05,
      "loss": 0.0084,
      "step": 2630
    },
    {
      "epoch": 4.033613445378151,
      "grad_norm": 0.07052377611398697,
      "learning_rate": 1.0887041996506859e-05,
      "loss": 0.0041,
      "step": 2640
    },
    {
      "epoch": 4.048892284186402,
      "grad_norm": 0.07718514651060104,
      "learning_rate": 1.0556781874841027e-05,
      "loss": 0.0052,
      "step": 2650
    },
    {
      "epoch": 4.048892284186402,
      "eval_loss": 0.031373970210552216,
      "eval_runtime": 199.7835,
      "eval_samples_per_second": 5.826,
      "eval_steps_per_second": 5.826,
      "step": 2650
    },
    {
      "epoch": 4.064171122994653,
      "grad_norm": 0.09308119118213654,
      "learning_rate": 1.0231016315824875e-05,
      "loss": 0.0042,
      "step": 2660
    },
    {
      "epoch": 4.0794499618029025,
      "grad_norm": 0.04948989301919937,
      "learning_rate": 9.909782440509491e-06,
      "loss": 0.0063,
      "step": 2670
    },
    {
      "epoch": 4.094728800611153,
      "grad_norm": 0.328868567943573,
      "learning_rate": 9.593116853559648e-06,
      "loss": 0.0076,
      "step": 2680
    },
    {
      "epoch": 4.110007639419404,
      "grad_norm": 0.03805866092443466,
      "learning_rate": 9.281055639082747e-06,
      "loss": 0.009,
      "step": 2690
    },
    {
      "epoch": 4.125286478227655,
      "grad_norm": 0.16321790218353271,
      "learning_rate": 8.973634356517063e-06,
      "loss": 0.0076,
      "step": 2700
    },
    {
      "epoch": 4.125286478227655,
      "eval_loss": 0.03259371221065521,
      "eval_runtime": 199.7271,
      "eval_samples_per_second": 5.828,
      "eval_steps_per_second": 5.828,
      "step": 2700
    },
    {
      "epoch": 4.140565317035906,
      "grad_norm": 0.2903512120246887,
      "learning_rate": 8.670888036579639e-06,
      "loss": 0.0056,
      "step": 2710
    },
    {
      "epoch": 4.1558441558441555,
      "grad_norm": 0.06461212784051895,
      "learning_rate": 8.372851177274604e-06,
      "loss": 0.0064,
      "step": 2720
    },
    {
      "epoch": 4.171122994652406,
      "grad_norm": 0.3396087884902954,
      "learning_rate": 8.079557739962128e-06,
      "loss": 0.0063,
      "step": 2730
    },
    {
      "epoch": 4.186401833460657,
      "grad_norm": 0.14997313916683197,
      "learning_rate": 7.791041145488454e-06,
      "loss": 0.0065,
      "step": 2740
    },
    {
      "epoch": 4.201680672268908,
      "grad_norm": 0.1173219084739685,
      "learning_rate": 7.507334270377619e-06,
      "loss": 0.0091,
      "step": 2750
    },
    {
      "epoch": 4.201680672268908,
      "eval_loss": 0.03307883441448212,
      "eval_runtime": 199.4548,
      "eval_samples_per_second": 5.836,
      "eval_steps_per_second": 5.836,
      "step": 2750
    },
    {
      "epoch": 4.2169595110771585,
      "grad_norm": 0.14040367305278778,
      "learning_rate": 7.228469443085206e-06,
      "loss": 0.0072,
      "step": 2760
    },
    {
      "epoch": 4.232238349885408,
      "grad_norm": 0.11947222054004669,
      "learning_rate": 6.954478440314427e-06,
      "loss": 0.0079,
      "step": 2770
    },
    {
      "epoch": 4.247517188693659,
      "grad_norm": 0.08814125508069992,
      "learning_rate": 6.685392483395259e-06,
      "loss": 0.0029,
      "step": 2780
    },
    {
      "epoch": 4.26279602750191,
      "grad_norm": 0.13022877275943756,
      "learning_rate": 6.421242234726682e-06,
      "loss": 0.0055,
      "step": 2790
    },
    {
      "epoch": 4.278074866310161,
      "grad_norm": 0.22816994786262512,
      "learning_rate": 6.1620577942827166e-06,
      "loss": 0.0045,
      "step": 2800
    },
    {
      "epoch": 4.278074866310161,
      "eval_loss": 0.0341590940952301,
      "eval_runtime": 199.141,
      "eval_samples_per_second": 5.845,
      "eval_steps_per_second": 5.845,
      "step": 2800
    },
    {
      "epoch": 4.293353705118411,
      "grad_norm": 0.7083064317703247,
      "learning_rate": 5.907868696182584e-06,
      "loss": 0.0058,
      "step": 2810
    },
    {
      "epoch": 4.308632543926661,
      "grad_norm": 0.029391461983323097,
      "learning_rate": 5.658703905325186e-06,
      "loss": 0.0078,
      "step": 2820
    },
    {
      "epoch": 4.323911382734912,
      "grad_norm": 0.17884346842765808,
      "learning_rate": 5.414591814088627e-06,
      "loss": 0.0077,
      "step": 2830
    },
    {
      "epoch": 4.339190221543163,
      "grad_norm": 0.07933899760246277,
      "learning_rate": 5.17556023909489e-06,
      "loss": 0.0059,
      "step": 2840
    },
    {
      "epoch": 4.354469060351414,
      "grad_norm": 0.22485603392124176,
      "learning_rate": 4.941636418040058e-06,
      "loss": 0.0047,
      "step": 2850
    },
    {
      "epoch": 4.354469060351414,
      "eval_loss": 0.034740038216114044,
      "eval_runtime": 199.1325,
      "eval_samples_per_second": 5.845,
      "eval_steps_per_second": 5.845,
      "step": 2850
    },
    {
      "epoch": 4.369747899159664,
      "grad_norm": 0.30258139967918396,
      "learning_rate": 4.7128470065906925e-06,
      "loss": 0.0041,
      "step": 2860
    },
    {
      "epoch": 4.385026737967914,
      "grad_norm": 0.20538049936294556,
      "learning_rate": 4.4892180753462744e-06,
      "loss": 0.0039,
      "step": 2870
    },
    {
      "epoch": 4.400305576776165,
      "grad_norm": 0.1709292232990265,
      "learning_rate": 4.270775106868586e-06,
      "loss": 0.004,
      "step": 2880
    },
    {
      "epoch": 4.415584415584416,
      "grad_norm": 0.12943388521671295,
      "learning_rate": 4.057542992777868e-06,
      "loss": 0.0057,
      "step": 2890
    },
    {
      "epoch": 4.4308632543926665,
      "grad_norm": 0.07701294124126434,
      "learning_rate": 3.849546030916473e-06,
      "loss": 0.0047,
      "step": 2900
    },
    {
      "epoch": 4.4308632543926665,
      "eval_loss": 0.03577398136258125,
      "eval_runtime": 199.2545,
      "eval_samples_per_second": 5.842,
      "eval_steps_per_second": 5.842,
      "step": 2900
    },
    {
      "epoch": 4.446142093200916,
      "grad_norm": 0.42437708377838135,
      "learning_rate": 3.646807922580098e-06,
      "loss": 0.004,
      "step": 2910
    },
    {
      "epoch": 4.461420932009167,
      "grad_norm": 0.21150662004947662,
      "learning_rate": 3.4493517698170164e-06,
      "loss": 0.0082,
      "step": 2920
    },
    {
      "epoch": 4.476699770817418,
      "grad_norm": 0.15437455475330353,
      "learning_rate": 3.2572000727956186e-06,
      "loss": 0.0058,
      "step": 2930
    },
    {
      "epoch": 4.491978609625669,
      "grad_norm": 0.3453219532966614,
      "learning_rate": 3.070374727240466e-06,
      "loss": 0.0043,
      "step": 2940
    },
    {
      "epoch": 4.507257448433919,
      "grad_norm": 0.1276451051235199,
      "learning_rate": 2.8888970219373314e-06,
      "loss": 0.005,
      "step": 2950
    },
    {
      "epoch": 4.507257448433919,
      "eval_loss": 0.03592822700738907,
      "eval_runtime": 199.3782,
      "eval_samples_per_second": 5.838,
      "eval_steps_per_second": 5.838,
      "step": 2950
    },
    {
      "epoch": 4.522536287242169,
      "grad_norm": 0.09260442107915878,
      "learning_rate": 2.7127876363072736e-06,
      "loss": 0.0031,
      "step": 2960
    },
    {
      "epoch": 4.53781512605042,
      "grad_norm": 0.21995659172534943,
      "learning_rate": 2.54206663805025e-06,
      "loss": 0.0057,
      "step": 2970
    },
    {
      "epoch": 4.553093964858671,
      "grad_norm": 0.0280003622174263,
      "learning_rate": 2.3767534808584125e-06,
      "loss": 0.0026,
      "step": 2980
    },
    {
      "epoch": 4.5683728036669216,
      "grad_norm": 0.11262509971857071,
      "learning_rate": 2.2168670021993075e-06,
      "loss": 0.0078,
      "step": 2990
    },
    {
      "epoch": 4.583651642475171,
      "grad_norm": 0.32838335633277893,
      "learning_rate": 2.0624254211693894e-06,
      "loss": 0.0049,
      "step": 3000
    },
    {
      "epoch": 4.583651642475171,
      "eval_loss": 0.0363006666302681,
      "eval_runtime": 199.2158,
      "eval_samples_per_second": 5.843,
      "eval_steps_per_second": 5.843,
      "step": 3000
    },
    {
      "epoch": 4.598930481283422,
      "grad_norm": 0.17691144347190857,
      "learning_rate": 1.9134463364179177e-06,
      "loss": 0.007,
      "step": 3010
    },
    {
      "epoch": 4.614209320091673,
      "grad_norm": 0.4434875547885895,
      "learning_rate": 1.7699467241416024e-06,
      "loss": 0.0073,
      "step": 3020
    },
    {
      "epoch": 4.629488158899924,
      "grad_norm": 0.17472083866596222,
      "learning_rate": 1.6319429361501714e-06,
      "loss": 0.0045,
      "step": 3030
    },
    {
      "epoch": 4.6447669977081745,
      "grad_norm": 0.05675290897488594,
      "learning_rate": 1.4994506980030577e-06,
      "loss": 0.0019,
      "step": 3040
    },
    {
      "epoch": 4.660045836516424,
      "grad_norm": 0.3393559157848358,
      "learning_rate": 1.3724851072174917e-06,
      "loss": 0.0039,
      "step": 3050
    },
    {
      "epoch": 4.660045836516424,
      "eval_loss": 0.036286067217588425,
      "eval_runtime": 198.942,
      "eval_samples_per_second": 5.851,
      "eval_steps_per_second": 5.851,
      "step": 3050
    },
    {
      "epoch": 4.675324675324675,
      "grad_norm": 0.2996921241283417,
      "learning_rate": 1.251060631548112e-06,
      "loss": 0.0107,
      "step": 3060
    },
    {
      "epoch": 4.690603514132926,
      "grad_norm": 0.1846129596233368,
      "learning_rate": 1.135191107338368e-06,
      "loss": 0.0029,
      "step": 3070
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 0.33350446820259094,
      "learning_rate": 1.0248897379438904e-06,
      "loss": 0.0054,
      "step": 3080
    },
    {
      "epoch": 4.721161191749427,
      "grad_norm": 0.18361401557922363,
      "learning_rate": 9.201690922279405e-07,
      "loss": 0.0052,
      "step": 3090
    },
    {
      "epoch": 4.736440030557677,
      "grad_norm": 0.1535758525133133,
      "learning_rate": 8.210411031291776e-07,
      "loss": 0.0062,
      "step": 3100
    },
    {
      "epoch": 4.736440030557677,
      "eval_loss": 0.03661744296550751,
      "eval_runtime": 199.1326,
      "eval_samples_per_second": 5.845,
      "eval_steps_per_second": 5.845,
      "step": 3100
    },
    {
      "epoch": 4.751718869365928,
      "grad_norm": 0.1135258898139,
      "learning_rate": 7.275170663019415e-07,
      "loss": 0.0061,
      "step": 3110
    },
    {
      "epoch": 4.766997708174179,
      "grad_norm": 0.3599923253059387,
      "learning_rate": 6.396076388290484e-07,
      "loss": 0.0043,
      "step": 3120
    },
    {
      "epoch": 4.7822765469824295,
      "grad_norm": 0.06559276580810547,
      "learning_rate": 5.573228380074736e-07,
      "loss": 0.0054,
      "step": 3130
    },
    {
      "epoch": 4.79755538579068,
      "grad_norm": 0.02168980799615383,
      "learning_rate": 4.806720402068477e-07,
      "loss": 0.0031,
      "step": 3140
    },
    {
      "epoch": 4.81283422459893,
      "grad_norm": 0.08103174716234207,
      "learning_rate": 4.0966397980100604e-07,
      "loss": 0.0054,
      "step": 3150
    },
    {
      "epoch": 4.81283422459893,
      "eval_loss": 0.03663839399814606,
      "eval_runtime": 198.8473,
      "eval_samples_per_second": 5.854,
      "eval_steps_per_second": 5.854,
      "step": 3150
    },
    {
      "epoch": 4.828113063407181,
      "grad_norm": 0.16072829067707062,
      "learning_rate": 3.4430674817274575e-07,
      "loss": 0.0043,
      "step": 3160
    },
    {
      "epoch": 4.843391902215432,
      "grad_norm": 0.05386335402727127,
      "learning_rate": 2.8460779279176896e-07,
      "loss": 0.0071,
      "step": 3170
    },
    {
      "epoch": 4.8586707410236825,
      "grad_norm": 0.18377594649791718,
      "learning_rate": 2.3057391636606696e-07,
      "loss": 0.0035,
      "step": 3180
    },
    {
      "epoch": 4.873949579831933,
      "grad_norm": 0.5214228630065918,
      "learning_rate": 1.8221127606674605e-07,
      "loss": 0.0078,
      "step": 3190
    },
    {
      "epoch": 4.889228418640183,
      "grad_norm": 0.09072000533342361,
      "learning_rate": 1.3952538282639982e-07,
      "loss": 0.0041,
      "step": 3200
    },
    {
      "epoch": 4.889228418640183,
      "eval_loss": 0.03660115599632263,
      "eval_runtime": 198.4598,
      "eval_samples_per_second": 5.865,
      "eval_steps_per_second": 5.865,
      "step": 3200
    },
    {
      "epoch": 4.904507257448434,
      "grad_norm": 0.4262472093105316,
      "learning_rate": 1.025211007111615e-07,
      "loss": 0.0048,
      "step": 3210
    },
    {
      "epoch": 4.919786096256685,
      "grad_norm": 0.13413435220718384,
      "learning_rate": 7.120264636643615e-08,
      "loss": 0.0023,
      "step": 3220
    },
    {
      "epoch": 4.935064935064935,
      "grad_norm": 0.03055586665868759,
      "learning_rate": 4.5573588536407254e-08,
      "loss": 0.006,
      "step": 3230
    },
    {
      "epoch": 4.950343773873186,
      "grad_norm": 0.36961618065834045,
      "learning_rate": 2.5636847657367623e-08,
      "loss": 0.0094,
      "step": 3240
    },
    {
      "epoch": 4.965622612681436,
      "grad_norm": 0.05908333510160446,
      "learning_rate": 1.1394695524963306e-08,
      "loss": 0.0047,
      "step": 3250
    },
    {
      "epoch": 4.965622612681436,
      "eval_loss": 0.03658132627606392,
      "eval_runtime": 198.452,
      "eval_samples_per_second": 5.865,
      "eval_steps_per_second": 5.865,
      "step": 3250
    },
    {
      "epoch": 4.980901451489687,
      "grad_norm": 0.10072652250528336,
      "learning_rate": 2.8487550352951363e-09,
      "loss": 0.0064,
      "step": 3260
    },
    {
      "epoch": 4.9961802902979375,
      "grad_norm": 0.12432490289211273,
      "learning_rate": 0.0,
      "loss": 0.0078,
      "step": 3270
    },
    {
      "epoch": 4.9961802902979375,
      "step": 3270,
      "total_flos": 8.84620635703935e+17,
      "train_loss": 0.04931148957939596,
      "train_runtime": 42656.8883,
      "train_samples_per_second": 1.227,
      "train_steps_per_second": 0.077
    }
  ],
  "logging_steps": 10,
  "max_steps": 3270,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.84620635703935e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}