{
  "model": "unsloth/Phi-4-unsloth-bnb-4bit",
  "dataset": "Mathieu-Thomas-JOSSET/michael_abab_conversations_infini_instruct.jsonl",
  "examples_total": 2872,
  "examples_train": 1436,
  "examples_eval": 1436,
  "world_size": 1,
  "effective_batch_size": 8,
  "steps_per_epoch_approx": 179.5,
  "max_steps": 2000,
  "eval_steps": 50,
  "save_steps": 50,
  "learning_rate": 9.95267419777795e-06,
  "warmup_steps": 10,
  "lr_scheduler_type": "linear",
  "weight_decay": 0.009206070410847844,
  "lora_r": 32,
  "lora_alpha": 64,
  "lora_dropout": 0.0,
  "best_checkpoint": "outputs/continue_r1_from_350_20260112_073729/checkpoint-100",
  "LR_AUTO_ENABLED": true,
  "LR_AUTO_USE_N": "train",
  "LR_AUTO_N_REF": 1436,
  "LR_AUTO_BASE": 1e-05,
  "LR_AUTO_MULT": 0.5,
  "LR_AUTO_FINAL": 5e-06,
  "best_step": 100,
  "best_eval_loss": 2.2380564212799072,
  "best_blended": 1.3520409573791146,
  "best_blended_step": 600
}

{
  "MODEL_NAME": "unsloth/Phi-4-unsloth-bnb-4bit",
  "CHAT_TEMPLATE": "phi-4",
  "MAX_SEQ_LENGTH": 2048,
  "LOAD_IN_4BIT": true,
  "DATASET_NAME": "Mathieu-Thomas-JOSSET/michael_abab_conversations_infini_instruct.jsonl",
  "DATASET_SPLIT": "train",
  "PER_DEVICE_TRAIN_BATCH_SIZE": 2,
  "GRADIENT_ACCUMULATION_STEPS": 4,
  "WARMUP_STEPS": 10,
  "MAX_STEPS": 2000,
  "LEARNING_RATE": 9.95267419777795e-06,
  "WEIGHT_DECAY": 0.009206070410847844,
  "LR_SCHEDULER_TYPE": "linear",
  "SEED": 3407,
  "PLOTLY_DARK_MODE": true,
  "PLOTLY_BASE_COLOR": "#00CC96",
  "PLOTLY_EMA_SPAN": 25,
  "LR_AUTO_ENABLED": true,
  "LR_AUTO_USE_N": "train",
  "LR_AUTO_N_REF": 1436,
  "LR_AUTO_BASE": 1e-05,
  "LR_AUTO_MULT": 0.5,
  "LR_AUTO_FINAL": 5e-06
}

{
  "model_name": "unsloth/Phi-4-unsloth-bnb-4bit",
  "dataset": {
    "name": "Mathieu-Thomas-JOSSET/michael_abab_conversations_infini_instruct.jsonl",
    "split": "train"
  },
  "training": {
    "max_steps": 2000,
    "learning_rate": 9.95267419777795e-06,
    "per_device_train_batch_size": 2,
    "gradient_accumulation_steps": 4,
    "max_seq_length": 2048,
    "seed": 3407,
    "optimizer": "adamw_8bit",
    "lr_scheduler_type": "linear"
  },
  "auto_lr": {
    "enabled": true,
    "use_n": "train",
    "n_ref": 1436,
    "base": 1e-05,
    "mult": 0.5,
    "final": 5e-06
  },
  "best": {
    "checkpoint": "/content/outputs/continue_r1_from_350_20260112_073729/checkpoint-100",
    "metric": 2.2380564212799072,
    "metric_name": "eval_loss"
  },
  "plotly": {
    "html": "training_loss_step.html"
  }
}

[
  {
    "loss": 2.2041,
    "grad_norm": 4.026190757751465,
    "learning_rate": 0.0,
    "epoch": 0.005571030640668524,
    "step": 1
  },
  {
    "loss": 2.6901,
    "grad_norm": 1.616629719734192,
    "learning_rate": 1.4636285584967574e-07,
    "epoch": 0.011142061281337047,
    "step": 2
  },
  {
    "loss": 2.6774,
    "grad_norm": 13.836981773376465,
    "learning_rate": 2.927257116993515e-07,
    "epoch": 0.016713091922005572,
    "step": 3
  },
  {
    "loss": 2.4478,
    "grad_norm": 1.857710361480713,
    "learning_rate": 4.3908856754902726e-07,
    "epoch": 0.022284122562674095,
    "step": 4
  },
  {
    "loss": 1.9988,
    "grad_norm": 1.4818029403686523,
    "learning_rate": 5.85451423398703e-07,
    "epoch": 0.027855153203342618,
    "step": 5
  },
  {
    "loss": 2.0358,
    "grad_norm": 1.726440191268921,
    "learning_rate": 7.318142792483787e-07,
    "epoch": 0.033426183844011144,
    "step": 6
  },
  {
    "loss": 2.5824,
    "grad_norm": 2.0604233741760254,
    "learning_rate": 8.781771350980545e-07,
    "epoch": 0.03899721448467967,
    "step": 7
  },
  {
    "loss": 2.3479,
    "grad_norm": 1.7288694381713867,
    "learning_rate": 1.0245399909477302e-06,
    "epoch": 0.04456824512534819,
    "step": 8
  },
  {
    "loss": 2.6387,
    "grad_norm": 1.9069620370864868,
    "learning_rate": 1.170902846797406e-06,
    "epoch": 0.05013927576601671,
    "step": 9
  },
  {
    "loss": 2.6083,
    "grad_norm": 1.4719465970993042,
    "learning_rate": 1.3172657026470817e-06,
    "epoch": 0.055710306406685235,
    "step": 10
  },
  {
    "loss": 2.3015,
    "grad_norm": 1.6306267976760864,
    "learning_rate": 1.4636285584967574e-06,
    "epoch": 0.06128133704735376,
    "step": 11
  },
  {
    "loss": 2.7042,
    "grad_norm": 1.4724116325378418,
    "learning_rate": 1.6099914143464333e-06,
    "epoch": 0.06685236768802229,
    "step": 12
  },
  {
    "loss": 2.5386,
    "grad_norm": 1.5470020771026611,
    "learning_rate": 1.756354270196109e-06,
    "epoch": 0.07242339832869081,
    "step": 13
  },
  {
    "loss": 2.5886,
    "grad_norm": 2.022662401199341,
    "learning_rate": 1.9027171260457846e-06,
    "epoch": 0.07799442896935933,
    "step": 14
  },
  {
    "loss": 2.7168,
    "grad_norm": 1.8387386798858643,
    "learning_rate": 2.0490799818954605e-06,
    "epoch": 0.08356545961002786,
    "step": 15
  },
  {
    "loss": 2.1886,
    "grad_norm": 1.9359395503997803,
    "learning_rate": 2.195442837745136e-06,
    "epoch": 0.08913649025069638,
    "step": 16
  },
  {
    "loss": 2.4191,
    "grad_norm": 1.5662318468093872,
    "learning_rate": 2.341805693594812e-06,
    "epoch": 0.0947075208913649,
    "step": 17
  },
  {
    "loss": 2.2736,
    "grad_norm": 1.7207640409469604,
    "learning_rate": 2.4881685494444876e-06,
    "epoch": 0.10027855153203342,
    "step": 18
  },
  {
    "loss": 2.7336,
    "grad_norm": 1.6225577592849731,
    "learning_rate": 2.6345314052941634e-06,
    "epoch": 0.10584958217270195,
    "step": 19
  },
  {
    "loss": 2.5288,
    "grad_norm": 1.6348892450332642,
    "learning_rate": 2.780894261143839e-06,
    "epoch": 0.11142061281337047,
    "step": 20
  },
  {
    "loss": 2.0214,
    "grad_norm": 1.5059679746627808,
    "learning_rate": 2.927257116993515e-06,
    "epoch": 0.116991643454039,
    "step": 21
  },
  {
    "loss": 2.3708,
    "grad_norm": 1.3699105978012085,
    "learning_rate": 3.073619972843191e-06,
    "epoch": 0.12256267409470752,
    "step": 22
  },
  {
    "loss": 2.5341,
    "grad_norm": 2.241403341293335,
    "learning_rate": 3.2199828286928667e-06,
    "epoch": 0.12813370473537605,
    "step": 23
  },
  {
    "loss": 2.4852,
    "grad_norm": 1.7692517042160034,
    "learning_rate": 3.3663456845425424e-06,
    "epoch": 0.13370473537604458,
    "step": 24
  },
  {
    "loss": 2.4959,
    "grad_norm": 1.9559876918792725,
    "learning_rate": 3.512708540392218e-06,
    "epoch": 0.1392757660167131,
    "step": 25
  },
  {
    "loss": 2.243,
    "grad_norm": 1.7536145448684692,
    "learning_rate": 3.659071396241894e-06,
    "epoch": 0.14484679665738162,
    "step": 26
  },
  {
    "loss": 2.1622,
    "grad_norm": 1.8103671073913574,
    "learning_rate": 3.805434252091569e-06,
    "epoch": 0.15041782729805014,
    "step": 27
  },
  {
    "loss": 2.1827,
    "grad_norm": 1.5473895072937012,
    "learning_rate": 3.951797107941245e-06,
    "epoch": 0.15598885793871867,
    "step": 28
  },
  {
    "loss": 2.3346,
    "grad_norm": 1.7137048244476318,
    "learning_rate": 4.098159963790921e-06,
    "epoch": 0.1615598885793872,
    "step": 29
  },
  {
    "loss": 1.9959,
    "grad_norm": 1.7803088426589966,
    "learning_rate": 4.244522819640596e-06,
    "epoch": 0.1671309192200557,
    "step": 30
  },
  {
    "loss": 2.5963,
    "grad_norm": 1.3565785884857178,
    "learning_rate": 4.390885675490272e-06,
    "epoch": 0.17270194986072424,
    "step": 31
  },
  {
    "loss": 2.3312,
    "grad_norm": 1.4962913990020752,
    "learning_rate": 4.537248531339948e-06,
    "epoch": 0.17827298050139276,
    "step": 32
  },
  {
    "loss": 2.2653,
    "grad_norm": 1.4864503145217896,
    "learning_rate": 4.683611387189624e-06,
    "epoch": 0.18384401114206128,
    "step": 33
  },
  {
    "loss": 2.6019,
    "grad_norm": 2.2222468852996826,
    "learning_rate": 4.829974243039299e-06,
    "epoch": 0.1894150417827298,
    "step": 34
  },
  {
    "loss": 2.2858,
    "grad_norm": 1.6111881732940674,
    "learning_rate": 4.976337098888975e-06,
    "epoch": 0.19498607242339833,
    "step": 35
  },
  {
    "loss": 2.1479,
    "grad_norm": 2.307185173034668,
    "learning_rate": 5.1226999547386506e-06,
    "epoch": 0.20055710306406685,
    "step": 36
  },
  {
    "loss": 2.3149,
    "grad_norm": 1.9714685678482056,
    "learning_rate": 5.269062810588327e-06,
    "epoch": 0.20612813370473537,
    "step": 37
  },
  {
    "loss": 2.4299,
    "grad_norm": 1.6915550231933594,
    "learning_rate": 5.415425666438002e-06,
    "epoch": 0.2116991643454039,
    "step": 38
  },
  {
    "loss": 2.1583,
    "grad_norm": 1.9084649085998535,
    "learning_rate": 5.561788522287678e-06,
    "epoch": 0.21727019498607242,
    "step": 39
  },
  {
    "loss": 2.1631,
    "grad_norm": 1.882629632949829,
    "learning_rate": 5.7081513781373534e-06,
    "epoch": 0.22284122562674094,
    "step": 40
  },
  {
    "loss": 2.0198,
    "grad_norm": 1.335666537284851,
    "learning_rate": 5.85451423398703e-06,
    "epoch": 0.22841225626740946,
    "step": 41
  },
  {
    "loss": 2.3086,
    "grad_norm": 2.620265007019043,
    "learning_rate": 6.000877089836705e-06,
    "epoch": 0.233983286908078,
    "step": 42
  },
  {
    "loss": 2.1388,
    "grad_norm": 2.0138227939605713,
    "learning_rate": 6.147239945686382e-06,
    "epoch": 0.2395543175487465,
    "step": 43
  },
  {
    "loss": 1.8651,
    "grad_norm": 1.6108520030975342,
    "learning_rate": 6.293602801536056e-06,
    "epoch": 0.24512534818941503,
    "step": 44
  },
  {
    "loss": 1.8407,
    "grad_norm": 1.5935970544815063,
    "learning_rate": 6.439965657385733e-06,
    "epoch": 0.25069637883008355,
    "step": 45
  },
  {
    "loss": 1.9392,
    "grad_norm": 1.3794289827346802,
    "learning_rate": 6.586328513235409e-06,
    "epoch": 0.2562674094707521,
    "step": 46
  },
  {
    "loss": 2.4167,
    "grad_norm": 1.23729407787323,
    "learning_rate": 6.732691369085085e-06,
    "epoch": 0.2618384401114206,
    "step": 47
  },
  {
    "loss": 2.2915,
    "grad_norm": 1.4265947341918945,
    "learning_rate": 6.87905422493476e-06,
    "epoch": 0.26740947075208915,
    "step": 48
  },
  {
    "loss": 1.7683,
    "grad_norm": 1.696736216545105,
    "learning_rate": 7.025417080784436e-06,
    "epoch": 0.27298050139275765,
    "step": 49
  },
  {
    "loss": 2.0858,
    "grad_norm": 1.5071961879730225,
    "learning_rate": 7.1717799366341115e-06,
    "epoch": 0.2785515320334262,
    "step": 50
  },
  {
    "eval_loss": 2.315699815750122,
    "eval_runtime": 35.9132,
    "eval_samples_per_second": 39.985,
    "eval_steps_per_second": 2.005,
    "epoch": 0.2785515320334262,
    "step": 50
  },
  {
    "loss": 2.1761,
    "grad_norm": 1.5999361276626587,
    "learning_rate": 7.318142792483788e-06,
    "epoch": 0.2841225626740947,
    "step": 51
  },
  {
    "loss": 2.145,
    "grad_norm": 2.0915520191192627,
    "learning_rate": 7.464505648333463e-06,
    "epoch": 0.28969359331476324,
    "step": 52
  },
  {
    "loss": 1.829,
    "grad_norm": 4.090714931488037,
    "learning_rate": 7.610868504183138e-06,
    "epoch": 0.29526462395543174,
    "step": 53
  },
  {
    "loss": 2.345,
    "grad_norm": 1.6347575187683105,
    "learning_rate": 7.757231360032815e-06,
    "epoch": 0.3008356545961003,
    "step": 54
  },
  {
    "loss": 1.9989,
    "grad_norm": 1.5609041452407837,
    "learning_rate": 7.90359421588249e-06,
    "epoch": 0.3064066852367688,
    "step": 55
  },
  {
    "loss": 2.3336,
    "grad_norm": 1.6561325788497925,
    "learning_rate": 8.049957071732166e-06,
    "epoch": 0.31197771587743733,
    "step": 56
  },
  {
    "loss": 2.0979,
    "grad_norm": 1.6579258441925049,
    "learning_rate": 8.196319927581842e-06,
    "epoch": 0.31754874651810583,
    "step": 57
  },
  {
    "loss": 2.4041,
    "grad_norm": 1.8354761600494385,
    "learning_rate": 8.342682783431518e-06,
    "epoch": 0.3231197771587744,
    "step": 58
  },
  {
    "loss": 2.1076,
    "grad_norm": 1.7043126821517944,
    "learning_rate": 8.489045639281193e-06,
    "epoch": 0.3286908077994429,
    "step": 59
  },
  {
    "loss": 2.2672,
    "grad_norm": 1.5663846731185913,
    "learning_rate": 8.635408495130869e-06,
    "epoch": 0.3342618384401114,
    "step": 60
  },
  {
    "loss": 2.1081,
    "grad_norm": 1.8134770393371582,
    "learning_rate": 8.781771350980545e-06,
    "epoch": 0.3398328690807799,
    "step": 61
  },
  {
    "loss": 2.034,
    "grad_norm": 1.3617796897888184,
    "learning_rate": 8.928134206830221e-06,
    "epoch": 0.34540389972144847,
    "step": 62
  },
  {
    "loss": 2.1147,
    "grad_norm": 1.7525910139083862,
    "learning_rate": 9.074497062679895e-06,
    "epoch": 0.35097493036211697,
    "step": 63
  },
  {
    "loss": 1.9639,
    "grad_norm": 1.7186285257339478,
    "learning_rate": 9.220859918529572e-06,
    "epoch": 0.3565459610027855,
    "step": 64
  },
  {
    "loss": 1.7557,
    "grad_norm": 1.9141530990600586,
    "learning_rate": 9.367222774379248e-06,
    "epoch": 0.362116991643454,
    "step": 65
  },
  {
    "loss": 1.8861,
    "grad_norm": 1.696165680885315,
    "learning_rate": 9.513585630228924e-06,
    "epoch": 0.36768802228412256,
    "step": 66
  },
  {
    "loss": 2.272,
    "grad_norm": 1.24228036403656,
    "learning_rate": 9.659948486078598e-06,
    "epoch": 0.3732590529247911,
    "step": 67
  },
  {
    "loss": 1.7173,
    "grad_norm": 1.9760662317276,
    "learning_rate": 9.806311341928276e-06,
    "epoch": 0.3788300835654596,
    "step": 68
  },
  {
    "loss": 1.8545,
    "grad_norm": 1.3207972049713135,
    "learning_rate": 9.95267419777795e-06,
    "epoch": 0.38440111420612816,
    "step": 69
  },
  {
    "loss": 2.0492,
    "grad_norm": 1.5849637985229492,
    "learning_rate": 1.0099037053627625e-05,
    "epoch": 0.38997214484679665,
    "step": 70
  },
  {
    "loss": 1.9203,
    "grad_norm": 2.7006468772888184,
    "learning_rate": 1.0245399909477301e-05,
    "epoch": 0.3955431754874652,
    "step": 71
  },
  {
    "loss": 2.1317,
    "grad_norm": 1.9178322553634644,
    "learning_rate": 1.0391762765326979e-05,
    "epoch": 0.4011142061281337,
    "step": 72
  },
  {
    "loss": 2.282,
    "grad_norm": 1.5044149160385132,
    "learning_rate": 1.0538125621176653e-05,
    "epoch": 0.40668523676880225,
    "step": 73
  },
  {
    "loss": 2.2023,
    "grad_norm": 1.9386659860610962,
    "learning_rate": 1.068448847702633e-05,
    "epoch": 0.41225626740947074,
    "step": 74
  },
  {
    "loss": 2.3545,
    "grad_norm": 1.3408238887786865,
    "learning_rate": 1.0830851332876004e-05,
    "epoch": 0.4178272980501393,
    "step": 75
  },
  {
    "loss": 2.1984,
    "grad_norm": 2.221109390258789,
    "learning_rate": 1.0977214188725682e-05,
    "epoch": 0.4233983286908078,
    "step": 76
  },
  {
    "loss": 1.9843,
    "grad_norm": 1.7843296527862549,
    "learning_rate": 1.1123577044575356e-05,
    "epoch": 0.42896935933147634,
    "step": 77
  },
  {
    "loss": 2.2899,
    "grad_norm": 1.6259101629257202,
    "learning_rate": 1.1269939900425032e-05,
    "epoch": 0.43454038997214484,
    "step": 78
  },
  {
    "loss": 1.8448,
    "grad_norm": 1.718583345413208,
    "learning_rate": 1.1416302756274707e-05,
    "epoch": 0.4401114206128134,
    "step": 79
  },
  {
    "loss": 1.857,
    "grad_norm": 1.8396937847137451,
    "learning_rate": 1.1562665612124385e-05,
    "epoch": 0.4456824512534819,
    "step": 80
  },
  {
    "loss": 2.1387,
    "grad_norm": 1.808605670928955,
    "learning_rate": 1.170902846797406e-05,
    "epoch": 0.45125348189415043,
    "step": 81
  },
  {
    "loss": 1.9587,
    "grad_norm": 2.590714931488037,
    "learning_rate": 1.1855391323823735e-05,
    "epoch": 0.4568245125348189,
    "step": 82
  },
  {
    "loss": 2.5776,
    "grad_norm": 1.550307273864746,
    "learning_rate": 1.200175417967341e-05,
    "epoch": 0.4623955431754875,
    "step": 83
  },
  {
    "loss": 2.1138,
    "grad_norm": 1.7622662782669067,
    "learning_rate": 1.2148117035523088e-05,
    "epoch": 0.467966573816156,
    "step": 84
  },
  {
    "loss": 2.0632,
    "grad_norm": 2.1933865547180176,
    "learning_rate": 1.2294479891372764e-05,
    "epoch": 0.4735376044568245,
    "step": 85
  },
  {
    "loss": 2.1885,
    "grad_norm": 1.6188870668411255,
    "learning_rate": 1.2440842747222438e-05,
    "epoch": 0.479108635097493,
    "step": 86
  },
  {
    "loss": 2.2515,
    "grad_norm": 1.6533507108688354,
    "learning_rate": 1.2587205603072113e-05,
    "epoch": 0.48467966573816157,
    "step": 87
  },
  {
    "loss": 2.3173,
    "grad_norm": 1.295457363128662,
    "learning_rate": 1.2733568458921789e-05,
    "epoch": 0.49025069637883006,
    "step": 88
  },
  {
    "loss": 1.9536,
    "grad_norm": 1.5764713287353516,
    "learning_rate": 1.2879931314771467e-05,
    "epoch": 0.4958217270194986,
    "step": 89
  },
  {
    "loss": 1.9134,
    "grad_norm": 1.8399816751480103,
    "learning_rate": 1.3026294170621141e-05,
    "epoch": 0.5013927576601671,
    "step": 90
  },
  {
    "loss": 2.2808,
    "grad_norm": 1.7519652843475342,
    "learning_rate": 1.3172657026470817e-05,
    "epoch": 0.5069637883008357,
    "step": 91
  },
  {
    "loss": 2.2055,
    "grad_norm": 1.4549530744552612,
    "learning_rate": 1.3319019882320492e-05,
    "epoch": 0.5125348189415042,
    "step": 92
  },
  {
    "loss": 1.9373,
    "grad_norm": 2.0461559295654297,
    "learning_rate": 1.346538273817017e-05,
    "epoch": 0.5181058495821727,
    "step": 93
  },
  {
    "loss": 1.7517,
    "grad_norm": 1.5427114963531494,
    "learning_rate": 1.3611745594019844e-05,
    "epoch": 0.5236768802228412,
    "step": 94
  },
  {
    "loss": 1.9707,
    "grad_norm": 1.5442962646484375,
    "learning_rate": 1.375810844986952e-05,
    "epoch": 0.5292479108635098,
    "step": 95
  },
  {
    "loss": 1.9629,
    "grad_norm": 1.939523458480835,
    "learning_rate": 1.3904471305719195e-05,
    "epoch": 0.5348189415041783,
    "step": 96
  },
  {
    "loss": 1.8158,
    "grad_norm": 1.9389022588729858,
    "learning_rate": 1.4050834161568872e-05,
    "epoch": 0.5403899721448467,
    "step": 97
  },
  {
    "loss": 2.5133,
    "grad_norm": 1.9970468282699585,
    "learning_rate": 1.4197197017418547e-05,
    "epoch": 0.5459610027855153,
    "step": 98
  },
  {
    "loss": 2.263,
    "grad_norm": 1.5786551237106323,
    "learning_rate": 1.4343559873268223e-05,
    "epoch": 0.5515320334261838,
    "step": 99
  },
  {
    "loss": 2.1035,
    "grad_norm": 2.2139763832092285,
    "learning_rate": 1.4489922729117897e-05,
    "epoch": 0.5571030640668524,
    "step": 100
  },
  {
    "eval_loss": 2.2380564212799072,
    "eval_runtime": 35.9418,
    "eval_samples_per_second": 39.954,
    "eval_steps_per_second": 2.003,
    "epoch": 0.5571030640668524,
    "step": 100
  },
  {
    "loss": 1.8114,
    "grad_norm": 2.2652857303619385,
    "learning_rate": 1.4636285584967575e-05,
    "epoch": 0.5626740947075209,
    "step": 101
  },
  {
    "loss": 2.3733,
    "grad_norm": 1.688621997833252,
    "learning_rate": 1.4782648440817251e-05,
    "epoch": 0.5682451253481894,
    "step": 102
  },
  {
    "loss": 1.929,
    "grad_norm": 2.500704765319824,
    "learning_rate": 1.4929011296666926e-05,
    "epoch": 0.5738161559888579,
    "step": 103
  },
  {
    "loss": 1.9718,
    "grad_norm": 1.492704153060913,
    "learning_rate": 1.50753741525166e-05,
    "epoch": 0.5793871866295265,
    "step": 104
  },
  {
    "loss": 2.1026,
    "grad_norm": 1.6980139017105103,
    "learning_rate": 1.5221737008366276e-05,
    "epoch": 0.584958217270195,
    "step": 105
  },
  {
    "loss": 1.7999,
    "grad_norm": 1.7127199172973633,
    "learning_rate": 1.5368099864215953e-05,
    "epoch": 0.5905292479108635,
    "step": 106
  },
  {
    "loss": 2.0349,
    "grad_norm": 1.8260376453399658,
    "learning_rate": 1.551446272006563e-05,
    "epoch": 0.596100278551532,
    "step": 107
  },
  {
    "loss": 2.2571,
    "grad_norm": 1.8122572898864746,
    "learning_rate": 1.5660825575915305e-05,
    "epoch": 0.6016713091922006,
    "step": 108
  },
  {
    "loss": 2.0794,
    "grad_norm": 2.299410343170166,
    "learning_rate": 1.580718843176498e-05,
    "epoch": 0.6072423398328691,
    "step": 109
  },
  {
    "loss": 2.1671,
    "grad_norm": 1.4942196607589722,
    "learning_rate": 1.5953551287614657e-05,
    "epoch": 0.6128133704735376,
    "step": 110
  },
  {
    "loss": 2.1292,
    "grad_norm": 1.6794716119766235,
    "learning_rate": 1.609991414346433e-05,
    "epoch": 0.6183844011142061,
    "step": 111
  },
  {
    "loss": 2.2534,
    "grad_norm": 1.8196300268173218,
    "learning_rate": 1.6246276999314006e-05,
    "epoch": 0.6239554317548747,
    "step": 112
  },
  {
    "loss": 2.0333,
    "grad_norm": 1.5703504085540771,
    "learning_rate": 1.6392639855163684e-05,
    "epoch": 0.6295264623955432,
    "step": 113
  },
  {
    "loss": 2.1495,
    "grad_norm": 1.766376256942749,
    "learning_rate": 1.6539002711013358e-05,
    "epoch": 0.6350974930362117,
    "step": 114
  },
  {
    "loss": 1.5522,
    "grad_norm": 2.6598968505859375,
    "learning_rate": 1.6685365566863036e-05,
    "epoch": 0.6406685236768802,
    "step": 115
  },
  {
    "loss": 2.2221,
    "grad_norm": 10.731008529663086,
    "learning_rate": 1.683172842271271e-05,
    "epoch": 0.6462395543175488,
    "step": 116
  },
  {
    "loss": 2.4227,
    "grad_norm": 2.2150168418884277,
    "learning_rate": 1.6978091278562385e-05,
    "epoch": 0.6518105849582173,
    "step": 117
  },
  {
    "loss": 2.3886,
    "grad_norm": 2.283031940460205,
    "learning_rate": 1.7124454134412063e-05,
    "epoch": 0.6573816155988857,
    "step": 118
  },
  {
    "loss": 2.0651,
    "grad_norm": 2.6018834114074707,
    "learning_rate": 1.7270816990261737e-05,
    "epoch": 0.6629526462395543,
    "step": 119
  },
  {
    "loss": 1.9465,
    "grad_norm": 1.8486937284469604,
    "learning_rate": 1.7417179846111412e-05,
    "epoch": 0.6685236768802229,
    "step": 120
  },
  {
    "loss": 2.588,
    "grad_norm": 2.0970637798309326,
    "learning_rate": 1.756354270196109e-05,
    "epoch": 0.6740947075208914,
    "step": 121
  },
  {
    "loss": 1.7602,
    "grad_norm": 1.5886075496673584,
    "learning_rate": 1.7709905557810764e-05,
    "epoch": 0.6796657381615598,
    "step": 122
  },
  {
    "loss": 2.2738,
    "grad_norm": 2.4414422512054443,
    "learning_rate": 1.7856268413660442e-05,
    "epoch": 0.6852367688022284,
    "step": 123
  },
  {
    "loss": 1.8145,
    "grad_norm": 1.7890093326568604,
    "learning_rate": 1.8002631269510116e-05,
    "epoch": 0.6908077994428969,
    "step": 124
  },
  {
    "loss": 1.7572,
    "grad_norm": 1.7805349826812744,
    "learning_rate": 1.814899412535979e-05,
    "epoch": 0.6963788300835655,
    "step": 125
  },
  {
    "loss": 2.0206,
    "grad_norm": 1.9520258903503418,
    "learning_rate": 1.829535698120947e-05,
    "epoch": 0.7019498607242339,
    "step": 126
  },
  {
    "loss": 2.1292,
    "grad_norm": 1.6244016885757446,
    "learning_rate": 1.8441719837059143e-05,
    "epoch": 0.7075208913649025,
    "step": 127
  },
  {
    "loss": 2.1207,
    "grad_norm": 1.6681342124938965,
    "learning_rate": 1.858808269290882e-05,
    "epoch": 0.713091922005571,
    "step": 128
  },
  {
    "loss": 2.1515,
    "grad_norm": 2.1032838821411133,
    "learning_rate": 1.8734445548758495e-05,
    "epoch": 0.7186629526462396,
    "step": 129
  },
  {
    "loss": 1.7264,
    "grad_norm": 2.093341588973999,
    "learning_rate": 1.888080840460817e-05,
    "epoch": 0.724233983286908,
    "step": 130
  },
  {
    "loss": 1.6036,
    "grad_norm": 1.9431419372558594,
    "learning_rate": 1.9027171260457848e-05,
    "epoch": 0.7298050139275766,
    "step": 131
  },
  {
    "loss": 2.132,
    "grad_norm": 3.0380795001983643,
    "learning_rate": 1.9173534116307522e-05,
    "epoch": 0.7353760445682451,
    "step": 132
  },
  {
    "loss": 2.061,
    "grad_norm": 3.623516321182251,
    "learning_rate": 1.9319896972157197e-05,
    "epoch": 0.7409470752089137,
    "step": 133
  },
  {
    "loss": 1.8765,
    "grad_norm": 2.320667266845703,
    "learning_rate": 1.9466259828006874e-05,
    "epoch": 0.7465181058495822,
    "step": 134
  },
  {
    "loss": 1.8337,
    "grad_norm": 1.9040995836257935,
    "learning_rate": 1.9612622683856552e-05,
    "epoch": 0.7520891364902507,
    "step": 135
  },
  {
    "loss": 2.0664,
    "grad_norm": 1.8677185773849487,
    "learning_rate": 1.9758985539706227e-05,
    "epoch": 0.7576601671309192,
    "step": 136
  },
  {
    "loss": 1.9651,
    "grad_norm": 2.414144992828369,
    "learning_rate": 1.99053483955559e-05,
    "epoch": 0.7632311977715878,
    "step": 137
  },
  {
    "loss": 1.9243,
    "grad_norm": 2.5697357654571533,
    "learning_rate": 1.989220087349807e-05,
    "epoch": 0.7688022284122563,
    "step": 138
  },
  {
    "loss": 2.0289,
    "grad_norm": 2.384612560272217,
    "learning_rate": 1.987905335144024e-05,
    "epoch": 0.7743732590529248,
    "step": 139
  },
  {
    "loss": 1.7206,
    "grad_norm": 2.284605026245117,
    "learning_rate": 1.986590582938241e-05,
    "epoch": 0.7799442896935933,
    "step": 140
  },
  {
    "loss": 1.7212,
    "grad_norm": 2.3488142490386963,
    "learning_rate": 1.985275830732458e-05,
    "epoch": 0.7855153203342619,
    "step": 141
  },
  {
    "loss": 1.559,
    "grad_norm": 1.849543809890747,
    "learning_rate": 1.983961078526675e-05,
    "epoch": 0.7910863509749304,
    "step": 142
  },
  {
    "loss": 1.9442,
    "grad_norm": 2.343719720840454,
    "learning_rate": 1.9826463263208915e-05,
    "epoch": 0.7966573816155988,
    "step": 143
  },
  {
    "loss": 1.8913,
    "grad_norm": 2.6115176677703857,
    "learning_rate": 1.9813315741151084e-05,
    "epoch": 0.8022284122562674,
    "step": 144
  },
  {
    "loss": 2.1248,
    "grad_norm": 2.703418731689453,
    "learning_rate": 1.9800168219093254e-05,
    "epoch": 0.807799442896936,
    "step": 145
  },
  {
    "loss": 2.1745,
    "grad_norm": 2.379194736480713,
    "learning_rate": 1.9787020697035423e-05,
    "epoch": 0.8133704735376045,
    "step": 146
  },
  {
    "loss": 2.0572,
    "grad_norm": 2.4916770458221436,
    "learning_rate": 1.9773873174977593e-05,
    "epoch": 0.8189415041782729,
    "step": 147
  },
  {
    "loss": 1.8028,
    "grad_norm": 3.7550413608551025,
    "learning_rate": 1.9760725652919762e-05,
    "epoch": 0.8245125348189415,
    "step": 148
  },
  {
    "loss": 2.2202,
    "grad_norm": 1.704113483428955,
    "learning_rate": 1.974757813086193e-05,
    "epoch": 0.83008356545961,
    "step": 149
  },
  {
    "loss": 1.8407,
    "grad_norm": 2.14805269241333,
    "learning_rate": 1.9734430608804098e-05,
    "epoch": 0.8356545961002786,
    "step": 150
  },
  {
    "eval_loss": 2.264693021774292,
    "eval_runtime": 35.9452,
    "eval_samples_per_second": 39.95,
    "eval_steps_per_second": 2.003,
    "epoch": 0.8356545961002786,
    "step": 150
  },
  {
    "loss": 2.3419,
    "grad_norm": 2.3600826263427734,
    "learning_rate": 1.972128308674627e-05,
    "epoch": 0.841225626740947,
    "step": 151
  },
  {
    "loss": 2.0321,
    "grad_norm": 2.7362117767333984,
    "learning_rate": 1.970813556468844e-05,
    "epoch": 0.8467966573816156,
    "step": 152
  },
  {
    "loss": 1.7054,
    "grad_norm": 2.982322931289673,
    "learning_rate": 1.9694988042630607e-05,
    "epoch": 0.8523676880222841,
    "step": 153
  },
  {
    "loss": 1.9304,
    "grad_norm": 2.8210840225219727,
    "learning_rate": 1.9681840520572776e-05,
    "epoch": 0.8579387186629527,
    "step": 154
  },
  {
    "loss": 2.1154,
    "grad_norm": 2.412022113800049,
    "learning_rate": 1.9668692998514946e-05,
    "epoch": 0.8635097493036211,
    "step": 155
  },
  {
    "loss": 2.0367,
    "grad_norm": 2.439105987548828,
    "learning_rate": 1.9655545476457115e-05,
    "epoch": 0.8690807799442897,
    "step": 156
  },
  {
    "loss": 1.5693,
    "grad_norm": 2.276296854019165,
    "learning_rate": 1.9642397954399285e-05,
    "epoch": 0.8746518105849582,
    "step": 157
  },
  {
    "loss": 1.726,
    "grad_norm": 2.12568998336792,
    "learning_rate": 1.9629250432341454e-05,
    "epoch": 0.8802228412256268,
    "step": 158
  },
  {
    "loss": 1.7872,
    "grad_norm": 2.1106767654418945,
    "learning_rate": 1.9616102910283624e-05,
    "epoch": 0.8857938718662952,
    "step": 159
  },
  {
    "loss": 1.9831,
    "grad_norm": 1.9893423318862915,
    "learning_rate": 1.960295538822579e-05,
    "epoch": 0.8913649025069638,
    "step": 160
  },
  {
    "loss": 2.0571,
    "grad_norm": 2.222038984298706,
    "learning_rate": 1.958980786616796e-05,
    "epoch": 0.8969359331476323,
    "step": 161
  },
  {
    "loss": 1.8674,
    "grad_norm": 2.5205395221710205,
    "learning_rate": 1.957666034411013e-05,
    "epoch": 0.9025069637883009,
    "step": 162
  },
  {
    "loss": 1.7982,
    "grad_norm": 2.212405204772949,
    "learning_rate": 1.95635128220523e-05,
    "epoch": 0.9080779944289693,
    "step": 163
  },
  {
    "loss": 1.7992,
    "grad_norm": 2.304945468902588,
    "learning_rate": 1.9550365299994468e-05,
    "epoch": 0.9136490250696379,
    "step": 164
  },
  {
    "loss": 2.2013,
    "grad_norm": 2.8349928855895996,
    "learning_rate": 1.9537217777936638e-05,
    "epoch": 0.9192200557103064,
    "step": 165
  },
  {
    "loss": 1.7908,
    "grad_norm": 2.2040209770202637,
    "learning_rate": 1.9524070255878807e-05,
    "epoch": 0.924791086350975,
    "step": 166
  },
  {
    "loss": 2.1308,
    "grad_norm": 2.550541400909424,
    "learning_rate": 1.9510922733820977e-05,
    "epoch": 0.9303621169916435,
    "step": 167
  },
  {
    "loss": 1.735,
    "grad_norm": 2.9808292388916016,
    "learning_rate": 1.9497775211763146e-05,
    "epoch": 0.935933147632312,
    "step": 168
  },
  {
    "loss": 2.1536,
    "grad_norm": 2.4572677612304688,
    "learning_rate": 1.9484627689705316e-05,
    "epoch": 0.9415041782729805,
    "step": 169
  },
  {
    "loss": 1.8616,
    "grad_norm": 2.414435863494873,
    "learning_rate": 1.9471480167647482e-05,
    "epoch": 0.947075208913649,
    "step": 170
  },
  {
    "loss": 1.9501,
    "grad_norm": 2.490251064300537,
    "learning_rate": 1.945833264558965e-05,
    "epoch": 0.9526462395543176,
    "step": 171
  },
  {
    "loss": 1.7965,
    "grad_norm": 3.2512645721435547,
    "learning_rate": 1.944518512353182e-05,
    "epoch": 0.958217270194986,
    "step": 172
  },
  {
    "loss": 1.8903,
    "grad_norm": 2.0697317123413086,
    "learning_rate": 1.943203760147399e-05,
    "epoch": 0.9637883008356546,
    "step": 173
  },
  {
    "loss": 1.9153,
    "grad_norm": 2.869088888168335,
    "learning_rate": 1.941889007941616e-05,
    "epoch": 0.9693593314763231,
    "step": 174
  },
  {
    "loss": 2.0043,
    "grad_norm": 2.5188841819763184,
    "learning_rate": 1.940574255735833e-05,
    "epoch": 0.9749303621169917,
    "step": 175
  },
  {
    "loss": 2.0106,
    "grad_norm": 2.2558531761169434,
    "learning_rate": 1.93925950353005e-05,
    "epoch": 0.9805013927576601,
    "step": 176
  },
  {
    "loss": 2.0356,
    "grad_norm": 2.78887677192688,
    "learning_rate": 1.9379447513242665e-05,
    "epoch": 0.9860724233983287,
    "step": 177
  },
  {
    "loss": 1.4849,
    "grad_norm": 2.9200024604797363,
    "learning_rate": 1.9366299991184838e-05,
    "epoch": 0.9916434540389972,
    "step": 178
  },
  {
    "loss": 1.6636,
    "grad_norm": 2.443997621536255,
    "learning_rate": 1.9353152469127007e-05,
    "epoch": 0.9972144846796658,
    "step": 179
  },
  {
    "loss": 1.4097,
    "grad_norm": 3.399275779724121,
    "learning_rate": 1.9340004947069174e-05,
    "epoch": 1.0,
    "step": 180
  },
  {
    "loss": 1.7611,
    "grad_norm": 2.312861442565918,
    "learning_rate": 1.9326857425011343e-05,
    "epoch": 1.0055710306406684,
    "step": 181
  },
  {
    "loss": 2.0967,
    "grad_norm": 2.799191951751709,
    "learning_rate": 1.9313709902953513e-05,
    "epoch": 1.011142061281337,
    "step": 182
  },
  {
    "loss": 2.3108,
    "grad_norm": 2.4845213890075684,
    "learning_rate": 1.9300562380895682e-05,
    "epoch": 1.0167130919220055,
    "step": 183
  },
  {
    "loss": 1.6263,
    "grad_norm": 2.72027325630188,
    "learning_rate": 1.928741485883785e-05,
    "epoch": 1.0222841225626742,
    "step": 184
  },
  {
    "loss": 1.6228,
    "grad_norm": 3.2783589363098145,
    "learning_rate": 1.927426733678002e-05,
    "epoch": 1.0278551532033426,
    "step": 185
  },
  {
    "loss": 1.9384,
    "grad_norm": 2.455291986465454,
    "learning_rate": 1.926111981472219e-05,
    "epoch": 1.033426183844011,
    "step": 186
  },
  {
    "loss": 1.5646,
    "grad_norm": 2.2230939865112305,
    "learning_rate": 1.9247972292664357e-05,
    "epoch": 1.0389972144846797,
    "step": 187
  },
  {
    "loss": 1.8545,
    "grad_norm": 2.596928119659424,
    "learning_rate": 1.9234824770606526e-05,
    "epoch": 1.0445682451253482,
    "step": 188
  },
  {
    "loss": 1.8436,
    "grad_norm": 2.5703697204589844,
    "learning_rate": 1.9221677248548696e-05,
    "epoch": 1.0501392757660166,
    "step": 189
  },
  {
    "loss": 2.2574,
    "grad_norm": 3.021871566772461,
    "learning_rate": 1.920852972649087e-05,
    "epoch": 1.0557103064066853,
    "step": 190
  },
  {
    "loss": 1.8046,
    "grad_norm": 2.35603404045105,
    "learning_rate": 1.9195382204433035e-05,
    "epoch": 1.0612813370473537,
    "step": 191
  },
  {
    "loss": 1.6635,
    "grad_norm": 2.453967809677124,
    "learning_rate": 1.9182234682375204e-05,
    "epoch": 1.0668523676880224,
    "step": 192
  },
  {
    "loss": 1.9118,
    "grad_norm": 3.2305331230163574,
    "learning_rate": 1.9169087160317374e-05,
    "epoch": 1.0724233983286908,
    "step": 193
  },
  {
    "loss": 1.5529,
    "grad_norm": 2.248871326446533,
    "learning_rate": 1.9155939638259543e-05,
    "epoch": 1.0779944289693593,
    "step": 194
  },
  {
    "loss": 1.7932,
    "grad_norm": 3.0331363677978516,
    "learning_rate": 1.9142792116201713e-05,
    "epoch": 1.083565459610028,
    "step": 195
  },
  {
    "loss": 1.7632,
    "grad_norm": 3.543948173522949,
    "learning_rate": 1.9129644594143882e-05,
    "epoch": 1.0891364902506964,
    "step": 196
  },
  {
    "loss": 1.8788,
    "grad_norm": 3.4173591136932373,
    "learning_rate": 1.911649707208605e-05,
    "epoch": 1.0947075208913648,
    "step": 197
  },
  {
    "loss": 2.0881,
    "grad_norm": 3.4639406204223633,
    "learning_rate": 1.9103349550028218e-05,
    "epoch": 1.1002785515320335,
    "step": 198
  },
  {
    "loss": 1.9197,
    "grad_norm": 3.6082725524902344,
    "learning_rate": 1.9090202027970388e-05,
    "epoch": 1.105849582172702,
    "step": 199
  },
  {
    "loss": 1.4541,
    "grad_norm": 2.834181070327759,
    "learning_rate": 1.9077054505912557e-05,
    "epoch": 1.1114206128133706,
    "step": 200
  },
  {
    "eval_loss": 2.3124067783355713,
    "eval_runtime": 35.9422,
    "eval_samples_per_second": 39.953,
    "eval_steps_per_second": 2.003,
    "epoch": 1.1114206128133706,
    "step": 200
  },
  {
    "loss": 1.8766,
    "grad_norm": 2.44728422164917,
    "learning_rate": 1.9063906983854727e-05,
    "epoch": 1.116991643454039,
    "step": 201
  },
  {
    "loss": 1.8877,
    "grad_norm": 3.1577866077423096,
    "learning_rate": 1.9050759461796896e-05,
    "epoch": 1.1225626740947074,
    "step": 202
  },
  {
    "loss": 1.6045,
    "grad_norm": 3.5458521842956543,
    "learning_rate": 1.9037611939739066e-05,
    "epoch": 1.128133704735376,
    "step": 203
  },
  {
    "loss": 1.705,
    "grad_norm": 2.496349811553955,
    "learning_rate": 1.9024464417681232e-05,
    "epoch": 1.1337047353760445,
    "step": 204
  },
  {
    "loss": 1.6478,
    "grad_norm": 3.2897088527679443,
    "learning_rate": 1.9011316895623405e-05,
    "epoch": 1.1392757660167132,
    "step": 205
  },
  {
    "loss": 1.6703,
    "grad_norm": 3.1694509983062744,
    "learning_rate": 1.8998169373565574e-05,
    "epoch": 1.1448467966573816,
    "step": 206
  },
  {
    "loss": 2.0232,
    "grad_norm": 2.8644907474517822,
    "learning_rate": 1.8985021851507744e-05,
    "epoch": 1.15041782729805,
    "step": 207
  },
  {
    "loss": 1.581,
    "grad_norm": 2.930053472518921,
    "learning_rate": 1.897187432944991e-05,
    "epoch": 1.1559888579387188,
    "step": 208
  },
  {
    "loss": 1.6617,
    "grad_norm": 2.9067940711975098,
    "learning_rate": 1.895872680739208e-05,
    "epoch": 1.1615598885793872,
    "step": 209
  },
  {
    "loss": 2.173,
    "grad_norm": 3.746903419494629,
    "learning_rate": 1.894557928533425e-05,
    "epoch": 1.1671309192200556,
    "step": 210
  },
  {
    "loss": 1.5917,
    "grad_norm": 4.83465576171875,
    "learning_rate": 1.893243176327642e-05,
    "epoch": 1.1727019498607243,
    "step": 211
  },
  {
    "loss": 1.531,
    "grad_norm": 3.0352439880371094,
    "learning_rate": 1.8919284241218588e-05,
    "epoch": 1.1782729805013927,
    "step": 212
  },
  {
    "loss": 2.193,
    "grad_norm": 2.738152027130127,
    "learning_rate": 1.8906136719160758e-05,
    "epoch": 1.1838440111420612,
    "step": 213
  },
  {
    "loss": 1.9884,
    "grad_norm": 3.005979061126709,
    "learning_rate": 1.8892989197102927e-05,
    "epoch": 1.1894150417827298,
    "step": 214
  },
  {
    "loss": 1.8659,
    "grad_norm": 3.930433750152588,
    "learning_rate": 1.8879841675045093e-05,
    "epoch": 1.1949860724233983,
    "step": 215
  },
  {
    "loss": 2.191,
    "grad_norm": 3.3943190574645996,
    "learning_rate": 1.8866694152987263e-05,
    "epoch": 1.200557103064067,
    "step": 216
  },
  {
    "loss": 2.1538,
    "grad_norm": 3.4692654609680176,
    "learning_rate": 1.8853546630929436e-05,
    "epoch": 1.2061281337047354,
    "step": 217
  },
  {
    "loss": 1.9939,
    "grad_norm": 2.889341354370117,
    "learning_rate": 1.8840399108871602e-05,
    "epoch": 1.2116991643454038,
    "step": 218
  },
  {
    "loss": 2.1667,
    "grad_norm": 3.123650550842285,
    "learning_rate": 1.882725158681377e-05,
    "epoch": 1.2172701949860725,
    "step": 219
  },
  {
    "loss": 1.9743,
    "grad_norm": 2.6485071182250977,
    "learning_rate": 1.881410406475594e-05,
    "epoch": 1.222841225626741,
    "step": 220
  },
  {
    "loss": 1.5679,
    "grad_norm": 3.791811227798462,
    "learning_rate": 1.880095654269811e-05,
    "epoch": 1.2284122562674096,
    "step": 221
  },
  {
    "loss": 2.1841,
    "grad_norm": 3.286864757537842,
    "learning_rate": 1.878780902064028e-05,
    "epoch": 1.233983286908078,
    "step": 222
  },
  {
    "loss": 2.1165,
    "grad_norm": 2.930072784423828,
    "learning_rate": 1.877466149858245e-05,
    "epoch": 1.2395543175487465,
    "step": 223
  },
  {
    "loss": 1.7816,
    "grad_norm": 2.936857223510742,
    "learning_rate": 1.876151397652462e-05,
    "epoch": 1.2451253481894151,
    "step": 224
  },
  {
    "loss": 1.702,
    "grad_norm": 2.3516695499420166,
    "learning_rate": 1.8748366454466785e-05,
    "epoch": 1.2506963788300836,
    "step": 225
  },
  {
    "loss": 1.7222,
    "grad_norm": 3.2817559242248535,
    "learning_rate": 1.8735218932408955e-05,
    "epoch": 1.2562674094707522,
    "step": 226
  },
  {
    "loss": 1.7151,
    "grad_norm": 2.987518548965454,
    "learning_rate": 1.8722071410351124e-05,
    "epoch": 1.2618384401114207,
    "step": 227
  },
  {
    "loss": 2.0545,
    "grad_norm": 3.132258415222168,
    "learning_rate": 1.8708923888293294e-05,
    "epoch": 1.267409470752089,
    "step": 228
  },
  {
    "loss": 1.4732,
    "grad_norm": 3.2233877182006836,
    "learning_rate": 1.8695776366235463e-05,
    "epoch": 1.2729805013927575,
    "step": 229
  },
  {
    "loss": 1.7168,
    "grad_norm": 3.2920405864715576,
    "learning_rate": 1.8682628844177633e-05,
    "epoch": 1.2785515320334262,
    "step": 230
  },
  {
    "loss": 1.497,
    "grad_norm": 2.536219596862793,
    "learning_rate": 1.8669481322119802e-05,
    "epoch": 1.2841225626740946,
    "step": 231
  },
  {
    "loss": 1.8177,
    "grad_norm": 4.246109485626221,
    "learning_rate": 1.865633380006197e-05,
    "epoch": 1.2896935933147633,
    "step": 232
  },
  {
    "loss": 1.9449,
    "grad_norm": 2.6518428325653076,
    "learning_rate": 1.864318627800414e-05,
    "epoch": 1.2952646239554317,
    "step": 233
  },
  {
    "loss": 1.4673,
    "grad_norm": 3.7276058197021484,
    "learning_rate": 1.863003875594631e-05,
    "epoch": 1.3008356545961002,
    "step": 234
  },
  {
    "loss": 1.865,
    "grad_norm": 3.2901997566223145,
    "learning_rate": 1.8616891233888477e-05,
    "epoch": 1.3064066852367688,
    "step": 235
  },
  {
    "loss": 1.7449,
    "grad_norm": 2.6417624950408936,
    "learning_rate": 1.8603743711830646e-05,
    "epoch": 1.3119777158774373,
    "step": 236
  },
  {
    "loss": 2.0348,
    "grad_norm": 3.81978702545166,
    "learning_rate": 1.8590596189772816e-05,
    "epoch": 1.317548746518106,
    "step": 237
  },
  {
    "loss": 1.8765,
    "grad_norm": 2.615661382675171,
    "learning_rate": 1.8577448667714985e-05,
    "epoch": 1.3231197771587744,
    "step": 238
  },
  {
    "loss": 1.7559,
    "grad_norm": 3.2889416217803955,
    "learning_rate": 1.8564301145657155e-05,
    "epoch": 1.3286908077994428,
    "step": 239
  },
  {
    "loss": 1.9031,
    "grad_norm": 4.006824970245361,
    "learning_rate": 1.8551153623599324e-05,
    "epoch": 1.3342618384401115,
    "step": 240
  },
  {
    "loss": 1.6102,
    "grad_norm": 3.3491599559783936,
    "learning_rate": 1.8538006101541494e-05,
    "epoch": 1.33983286908078,
    "step": 241
  },
  {
    "loss": 2.0716,
    "grad_norm": 3.2669260501861572,
    "learning_rate": 1.852485857948366e-05,
    "epoch": 1.3454038997214486,
    "step": 242
  },
  {
    "loss": 2.0298,
    "grad_norm": 4.218564510345459,
    "learning_rate": 1.851171105742583e-05,
    "epoch": 1.350974930362117,
    "step": 243
  },
  {
    "loss": 1.9911,
    "grad_norm": 3.5515315532684326,
    "learning_rate": 1.8498563535368003e-05,
    "epoch": 1.3565459610027855,
    "step": 244
  },
  {
    "loss": 1.3477,
    "grad_norm": 4.0060343742370605,
    "learning_rate": 1.848541601331017e-05,
    "epoch": 1.362116991643454,
    "step": 245
  },
  {
    "loss": 1.4686,
    "grad_norm": 3.574927568435669,
    "learning_rate": 1.8472268491252338e-05,
    "epoch": 1.3676880222841226,
    "step": 246
  },
  {
    "loss": 1.617,
    "grad_norm": 3.4316840171813965,
    "learning_rate": 1.8459120969194508e-05,
    "epoch": 1.3732590529247912,
    "step": 247
  },
  {
    "loss": 1.6593,
    "grad_norm": 3.2629754543304443,
    "learning_rate": 1.8445973447136677e-05,
    "epoch": 1.3788300835654597,
    "step": 248
  },
  {
    "loss": 1.2608,
    "grad_norm": 3.133815050125122,
    "learning_rate": 1.8432825925078847e-05,
    "epoch": 1.384401114206128,
    "step": 249
  },
  {
    "loss": 1.8523,
    "grad_norm": 3.742141008377075,
    "learning_rate": 1.8419678403021016e-05,
    "epoch": 1.3899721448467965,
    "step": 250
  },
  {
    "eval_loss": 2.335228443145752,
    "eval_runtime": 35.9668,
    "eval_samples_per_second": 39.926,
    "eval_steps_per_second": 2.002,
    "epoch": 1.3899721448467965,
    "step": 250
  },
  {
    "loss": 1.7768,
    "grad_norm": 3.9163429737091064,
    "learning_rate": 1.8406530880963186e-05,
    "epoch": 1.3955431754874652,
    "step": 251
  },
  {
    "loss": 1.7455,
    "grad_norm": 3.3456947803497314,
    "learning_rate": 1.8393383358905352e-05,
    "epoch": 1.4011142061281336,
    "step": 252
  },
  {
    "loss": 1.7103,
    "grad_norm": 4.220420837402344,
    "learning_rate": 1.838023583684752e-05,
    "epoch": 1.4066852367688023,
    "step": 253
  },
  {
    "loss": 2.0054,
    "grad_norm": 4.233839511871338,
    "learning_rate": 1.836708831478969e-05,
    "epoch": 1.4122562674094707,
    "step": 254
  },
  {
    "loss": 1.7175,
    "grad_norm": 3.703934669494629,
    "learning_rate": 1.8353940792731864e-05,
    "epoch": 1.4178272980501392,
    "step": 255
  },
  {
    "loss": 1.7225,
    "grad_norm": 4.210822105407715,
    "learning_rate": 1.834079327067403e-05,
    "epoch": 1.4233983286908078,
    "step": 256
  },
  {
    "loss": 1.6882,
    "grad_norm": 3.8861896991729736,
    "learning_rate": 1.83276457486162e-05,
    "epoch": 1.4289693593314763,
    "step": 257
  },
  {
    "loss": 2.0721,
    "grad_norm": 4.4140424728393555,
    "learning_rate": 1.831449822655837e-05,
    "epoch": 1.434540389972145,
    "step": 258
  },
  {
    "loss": 1.6198,
    "grad_norm": 3.1098673343658447,
    "learning_rate": 1.830135070450054e-05,
    "epoch": 1.4401114206128134,
    "step": 259
  },
  {
    "loss": 1.9632,
    "grad_norm": 2.9485561847686768,
    "learning_rate": 1.8288203182442708e-05,
    "epoch": 1.4456824512534818,
    "step": 260
  },
  {
    "loss": 1.9262,
    "grad_norm": 3.842655658721924,
    "learning_rate": 1.8275055660384878e-05,
    "epoch": 1.4512534818941505,
    "step": 261
  },
  {
    "loss": 2.0807,
    "grad_norm": 4.122529983520508,
    "learning_rate": 1.8261908138327047e-05,
    "epoch": 1.456824512534819,
    "step": 262
  },
  {
    "loss": 2.0099,
    "grad_norm": 3.6181795597076416,
    "learning_rate": 1.8248760616269213e-05,
    "epoch": 1.4623955431754876,
    "step": 263
  },
  {
    "loss": 1.7435,
    "grad_norm": 3.9433975219726562,
    "learning_rate": 1.8235613094211383e-05,
    "epoch": 1.467966573816156,
    "step": 264
  },
  {
    "loss": 1.4648,
    "grad_norm": 5.496665000915527,
    "learning_rate": 1.8222465572153552e-05,
    "epoch": 1.4735376044568245,
    "step": 265
  },
  {
    "loss": 2.106,
    "grad_norm": 3.3920114040374756,
    "learning_rate": 1.8209318050095722e-05,
    "epoch": 1.479108635097493,
    "step": 266
  },
  {
    "loss": 1.4486,
    "grad_norm": 4.195888519287109,
    "learning_rate": 1.819617052803789e-05,
    "epoch": 1.4846796657381616,
    "step": 267
  },
  {
    "loss": 1.4996,
    "grad_norm": 3.5301265716552734,
    "learning_rate": 1.818302300598006e-05,
    "epoch": 1.49025069637883,
    "step": 268
  },
  {
    "loss": 1.8247,
    "grad_norm": 3.3157520294189453,
    "learning_rate": 1.8169875483922227e-05,
    "epoch": 1.4958217270194987,
    "step": 269
  },
  {
    "loss": 1.6092,
    "grad_norm": 4.3797383308410645,
    "learning_rate": 1.8156727961864397e-05,
    "epoch": 1.501392757660167,
    "step": 270
  },
  {
    "loss": 1.6071,
    "grad_norm": 3.3917229175567627,
    "learning_rate": 1.814358043980657e-05,
    "epoch": 1.5069637883008355,
    "step": 271
  },
  {
    "loss": 1.9553,
    "grad_norm": 3.171808958053589,
    "learning_rate": 1.813043291774874e-05,
    "epoch": 1.5125348189415042,
    "step": 272
  },
  {
    "loss": 1.8105,
    "grad_norm": 3.1904940605163574,
    "learning_rate": 1.8117285395690905e-05,
    "epoch": 1.5181058495821727,
    "step": 273
  },
  {
    "loss": 1.5718,
    "grad_norm": 3.7544777393341064,
    "learning_rate": 1.8104137873633075e-05,
    "epoch": 1.5236768802228413,
    "step": 274
  },
  {
    "loss": 1.9999,
    "grad_norm": 4.143693923950195,
    "learning_rate": 1.8090990351575244e-05,
    "epoch": 1.5292479108635098,
    "step": 275
  },
  {
    "loss": 2.0393,
    "grad_norm": 3.505359411239624,
    "learning_rate": 1.8077842829517414e-05,
    "epoch": 1.5348189415041782,
    "step": 276
  },
  {
    "loss": 1.6101,
    "grad_norm": 4.118677139282227,
    "learning_rate": 1.8064695307459583e-05,
    "epoch": 1.5403899721448466,
    "step": 277
  },
  {
    "loss": 1.6718,
    "grad_norm": 4.947996139526367,
    "learning_rate": 1.8051547785401753e-05,
    "epoch": 1.5459610027855153,
    "step": 278
  },
  {
    "loss": 2.2007,
    "grad_norm": 4.226828575134277,
    "learning_rate": 1.8038400263343922e-05,
    "epoch": 1.551532033426184,
    "step": 279
  },
  {
    "loss": 1.7025,
    "grad_norm": 4.085235118865967,
    "learning_rate": 1.802525274128609e-05,
    "epoch": 1.5571030640668524,
    "step": 280
  },
  {
    "loss": 1.7632,
    "grad_norm": 3.5451292991638184,
    "learning_rate": 1.8012105219228258e-05,
    "epoch": 1.5626740947075208,
    "step": 281
  },
  {
    "loss": 1.4975,
    "grad_norm": 5.2698540687561035,
    "learning_rate": 1.799895769717043e-05,
    "epoch": 1.5682451253481893,
    "step": 282
  },
  {
    "loss": 1.2189,
    "grad_norm": 3.662693738937378,
    "learning_rate": 1.7985810175112597e-05,
    "epoch": 1.573816155988858,
    "step": 283
  },
  {
    "loss": 2.1889,
    "grad_norm": 3.9369843006134033,
    "learning_rate": 1.7972662653054766e-05,
    "epoch": 1.5793871866295266,
    "step": 284
  },
  {
    "loss": 1.782,
    "grad_norm": 5.153691291809082,
    "learning_rate": 1.7959515130996936e-05,
    "epoch": 1.584958217270195,
    "step": 285
  },
  {
    "loss": 1.7055,
    "grad_norm": 3.5153331756591797,
    "learning_rate": 1.7946367608939105e-05,
    "epoch": 1.5905292479108635,
    "step": 286
  },
  {
    "loss": 2.0713,
    "grad_norm": 3.8740577697753906,
    "learning_rate": 1.7933220086881275e-05,
    "epoch": 1.596100278551532,
    "step": 287
  },
  {
    "loss": 1.6159,
    "grad_norm": 2.977501153945923,
    "learning_rate": 1.7920072564823445e-05,
    "epoch": 1.6016713091922006,
    "step": 288
  },
  {
    "loss": 2.0388,
    "grad_norm": 4.873539447784424,
    "learning_rate": 1.7906925042765614e-05,
    "epoch": 1.6072423398328692,
    "step": 289
  },
  {
    "loss": 1.7656,
    "grad_norm": 3.6297993659973145,
    "learning_rate": 1.789377752070778e-05,
    "epoch": 1.6128133704735377,
    "step": 290
  },
  {
    "loss": 1.9818,
    "grad_norm": 2.868178367614746,
    "learning_rate": 1.788062999864995e-05,
    "epoch": 1.6183844011142061,
    "step": 291
  },
  {
    "loss": 1.6421,
    "grad_norm": 4.532885551452637,
    "learning_rate": 1.786748247659212e-05,
    "epoch": 1.6239554317548746,
    "step": 292
  },
  {
    "loss": 1.653,
    "grad_norm": 5.63344669342041,
    "learning_rate": 1.785433495453429e-05,
    "epoch": 1.6295264623955432,
    "step": 293
  },
  {
    "loss": 1.8727,
    "grad_norm": 4.235146999359131,
    "learning_rate": 1.7841187432476458e-05,
    "epoch": 1.6350974930362117,
    "step": 294
  },
  {
    "loss": 1.3509,
    "grad_norm": 4.512764930725098,
    "learning_rate": 1.7828039910418628e-05,
    "epoch": 1.6406685236768803,
    "step": 295
  },
  {
    "loss": 1.7836,
    "grad_norm": 3.72898268699646,
    "learning_rate": 1.7814892388360797e-05,
    "epoch": 1.6462395543175488,
    "step": 296
  },
  {
    "loss": 1.6315,
    "grad_norm": 3.1936659812927246,
    "learning_rate": 1.7801744866302963e-05,
    "epoch": 1.6518105849582172,
    "step": 297
  },
  {
    "loss": 1.9805,
    "grad_norm": 3.1188321113586426,
    "learning_rate": 1.7788597344245136e-05,
    "epoch": 1.6573816155988856,
    "step": 298
  },
  {
    "loss": 1.8716,
    "grad_norm": 4.88875150680542,
    "learning_rate": 1.7775449822187306e-05,
    "epoch": 1.6629526462395543,
    "step": 299
  },
  {
    "loss": 1.4669,
    "grad_norm": 4.494915962219238,
    "learning_rate": 1.7762302300129472e-05,
    "epoch": 1.668523676880223,
    "step": 300
  },
  {
    "eval_loss": 2.3116097450256348,
    "eval_runtime": 35.9294,
    "eval_samples_per_second": 39.967,
    "eval_steps_per_second": 2.004,
    "epoch": 1.668523676880223,
    "step": 300
  },
  {
    "loss": 1.3418,
    "grad_norm": 4.365106582641602,
    "learning_rate": 1.774915477807164e-05,
    "epoch": 1.6740947075208914,
    "step": 301
  },
  {
    "loss": 1.4561,
    "grad_norm": 4.683363914489746,
    "learning_rate": 1.773600725601381e-05,
    "epoch": 1.6796657381615598,
    "step": 302
  },
  {
    "loss": 1.8321,
    "grad_norm": 4.195693492889404,
    "learning_rate": 1.772285973395598e-05,
    "epoch": 1.6852367688022283,
    "step": 303
  },
  {
    "loss": 1.8932,
    "grad_norm": 4.681265830993652,
    "learning_rate": 1.770971221189815e-05,
    "epoch": 1.690807799442897,
    "step": 304
  },
  {
    "loss": 2.0071,
    "grad_norm": 5.034351348876953,
    "learning_rate": 1.769656468984032e-05,
    "epoch": 1.6963788300835656,
    "step": 305
  },
  {
    "loss": 1.9824,
    "grad_norm": 3.9581334590911865,
    "learning_rate": 1.768341716778249e-05,
    "epoch": 1.701949860724234,
    "step": 306
  },
  {
    "loss": 2.2225,
    "grad_norm": 3.9467825889587402,
    "learning_rate": 1.7670269645724655e-05,
    "epoch": 1.7075208913649025,
    "step": 307
  },
  {
    "loss": 1.671,
    "grad_norm": 3.7253997325897217,
    "learning_rate": 1.7657122123666825e-05,
    "epoch": 1.713091922005571,
    "step": 308
  },
  {
    "loss": 1.7876,
    "grad_norm": 4.8212480545043945,
    "learning_rate": 1.7643974601608998e-05,
    "epoch": 1.7186629526462396,
    "step": 309
  },
  {
    "loss": 1.1102,
    "grad_norm": 4.235992431640625,
    "learning_rate": 1.7630827079551164e-05,
    "epoch": 1.724233983286908,
    "step": 310
  },
  {
    "loss": 1.7577,
    "grad_norm": 3.5870513916015625,
    "learning_rate": 1.7617679557493333e-05,
    "epoch": 1.7298050139275767,
    "step": 311
  },
  {
    "loss": 1.3948,
    "grad_norm": 4.27365779876709,
    "learning_rate": 1.7604532035435503e-05,
    "epoch": 1.7353760445682451,
    "step": 312
  },
  {
    "loss": 1.5507,
    "grad_norm": 4.927708625793457,
    "learning_rate": 1.7591384513377672e-05,
    "epoch": 1.7409470752089136,
    "step": 313
  },
  {
    "loss": 1.5299,
    "grad_norm": 4.702437877655029,
    "learning_rate": 1.7578236991319842e-05,
    "epoch": 1.7465181058495822,
    "step": 314
  },
  {
    "loss": 1.6187,
    "grad_norm": 4.205385684967041,
    "learning_rate": 1.756508946926201e-05,
    "epoch": 1.7520891364902507,
    "step": 315
  },
  {
    "loss": 1.6467,
    "grad_norm": 3.724274158477783,
    "learning_rate": 1.755194194720418e-05,
    "epoch": 1.7576601671309193,
    "step": 316
  },
  {
    "loss": 1.48,
    "grad_norm": 5.0788187980651855,
    "learning_rate": 1.7538794425146347e-05,
    "epoch": 1.7632311977715878,
    "step": 317
  },
  {
    "loss": 1.2413,
    "grad_norm": 4.211026191711426,
    "learning_rate": 1.7525646903088517e-05,
    "epoch": 1.7688022284122562,
    "step": 318
  },
  {
    "loss": 1.2792,
    "grad_norm": 4.383068561553955,
    "learning_rate": 1.7512499381030686e-05,
    "epoch": 1.7743732590529246,
    "step": 319
  },
  {
    "loss": 2.0635,
    "grad_norm": 5.2455668449401855,
    "learning_rate": 1.7499351858972856e-05,
    "epoch": 1.7799442896935933,
    "step": 320
  },
  {
    "loss": 1.9011,
    "grad_norm": 4.73854398727417,
    "learning_rate": 1.7486204336915025e-05,
    "epoch": 1.785515320334262,
    "step": 321
  },
  {
    "loss": 1.9017,
    "grad_norm": 5.136256217956543,
    "learning_rate": 1.7473056814857195e-05,
    "epoch": 1.7910863509749304,
    "step": 322
  },
  {
    "loss": 1.7304,
    "grad_norm": 5.707761764526367,
    "learning_rate": 1.7459909292799364e-05,
    "epoch": 1.7966573816155988,
    "step": 323
  },
  {
    "loss": 1.9703,
    "grad_norm": 4.81571102142334,
    "learning_rate": 1.744676177074153e-05,
    "epoch": 1.8022284122562673,
    "step": 324
  },
  {
    "loss": 1.6825,
    "grad_norm": 6.157602310180664,
    "learning_rate": 1.7433614248683703e-05,
    "epoch": 1.807799442896936,
    "step": 325
  },
  {
    "loss": 1.7945,
    "grad_norm": 5.200462818145752,
    "learning_rate": 1.7420466726625873e-05,
    "epoch": 1.8133704735376046,
    "step": 326
  },
  {
    "loss": 1.7701,
    "grad_norm": 5.342528820037842,
    "learning_rate": 1.7407319204568042e-05,
    "epoch": 1.818941504178273,
    "step": 327
  },
  {
    "loss": 1.8,
    "grad_norm": 4.419646739959717,
    "learning_rate": 1.739417168251021e-05,
    "epoch": 1.8245125348189415,
    "step": 328
  },
  {
    "loss": 1.3064,
    "grad_norm": 5.106484889984131,
    "learning_rate": 1.7381024160452378e-05,
    "epoch": 1.83008356545961,
    "step": 329
  },
  {
    "loss": 2.0357,
    "grad_norm": 4.221576690673828,
    "learning_rate": 1.7367876638394547e-05,
    "epoch": 1.8356545961002786,
    "step": 330
  },
  {
    "loss": 1.6015,
    "grad_norm": 6.323553562164307,
    "learning_rate": 1.7354729116336717e-05,
    "epoch": 1.841225626740947,
    "step": 331
  },
  {
    "loss": 1.5858,
    "grad_norm": 4.978970527648926,
    "learning_rate": 1.7341581594278887e-05,
    "epoch": 1.8467966573816157,
    "step": 332
  },
  {
    "loss": 1.9489,
    "grad_norm": 3.1882030963897705,
    "learning_rate": 1.7328434072221056e-05,
    "epoch": 1.8523676880222841,
    "step": 333
  },
  {
    "loss": 2.1722,
    "grad_norm": 4.047868251800537,
    "learning_rate": 1.7315286550163222e-05,
    "epoch": 1.8579387186629526,
    "step": 334
  },
  {
    "loss": 1.5027,
    "grad_norm": 4.2307448387146,
    "learning_rate": 1.730213902810539e-05,
    "epoch": 1.863509749303621,
    "step": 335
  },
  {
    "loss": 1.481,
    "grad_norm": 6.048774242401123,
    "learning_rate": 1.7288991506047565e-05,
    "epoch": 1.8690807799442897,
    "step": 336
  },
  {
    "loss": 1.8746,
    "grad_norm": 5.389241695404053,
    "learning_rate": 1.7275843983989734e-05,
    "epoch": 1.8746518105849583,
    "step": 337
  },
  {
    "loss": 2.0807,
    "grad_norm": 4.036198139190674,
    "learning_rate": 1.72626964619319e-05,
    "epoch": 1.8802228412256268,
    "step": 338
  },
  {
    "loss": 1.7448,
    "grad_norm": 5.005743503570557,
    "learning_rate": 1.724954893987407e-05,
    "epoch": 1.8857938718662952,
    "step": 339
  },
  {
    "loss": 1.9092,
    "grad_norm": 4.462837219238281,
    "learning_rate": 1.723640141781624e-05,
    "epoch": 1.8913649025069637,
    "step": 340
  },
  {
    "loss": 1.7032,
    "grad_norm": 4.945067405700684,
    "learning_rate": 1.722325389575841e-05,
    "epoch": 1.8969359331476323,
    "step": 341
  },
  {
    "loss": 1.9141,
    "grad_norm": 3.7232062816619873,
    "learning_rate": 1.721010637370058e-05,
    "epoch": 1.902506963788301,
    "step": 342
  },
  {
    "loss": 1.8258,
    "grad_norm": 3.8830628395080566,
    "learning_rate": 1.7196958851642748e-05,
    "epoch": 1.9080779944289694,
    "step": 343
  },
  {
    "loss": 1.7998,
    "grad_norm": 4.693456649780273,
    "learning_rate": 1.7183811329584917e-05,
    "epoch": 1.9136490250696379,
    "step": 344
  },
  {
    "loss": 2.0583,
    "grad_norm": 4.737421989440918,
    "learning_rate": 1.7170663807527083e-05,
    "epoch": 1.9192200557103063,
    "step": 345
  },
  {
    "loss": 1.494,
    "grad_norm": 2.78582501411438,
    "learning_rate": 1.7157516285469253e-05,
    "epoch": 1.924791086350975,
    "step": 346
  },
  {
    "loss": 1.7167,
    "grad_norm": 4.305075168609619,
    "learning_rate": 1.7144368763411423e-05,
    "epoch": 1.9303621169916436,
    "step": 347
  },
  {
    "loss": 1.7753,
    "grad_norm": 3.9957072734832764,
    "learning_rate": 1.7131221241353592e-05,
    "epoch": 1.935933147632312,
    "step": 348
  },
  {
    "loss": 1.8852,
    "grad_norm": 4.9537434577941895,
    "learning_rate": 1.711807371929576e-05,
    "epoch": 1.9415041782729805,
    "step": 349
  },
  {
    "loss": 1.8729,
    "grad_norm": 3.9404208660125732,
    "learning_rate": 1.710492619723793e-05,
    "epoch": 1.947075208913649,
    "step": 350
  },
  {
    "eval_loss": 2.3213632106781006,
    "eval_runtime": 35.9387,
    "eval_samples_per_second": 39.957,
    "eval_steps_per_second": 2.003,
    "epoch": 1.947075208913649,
    "step": 350
  },
  {
    "loss": 2.1419,
    "grad_norm": 3.202141046524048,
    "learning_rate": 1.70917786751801e-05,
    "epoch": 1.9526462395543176,
    "step": 351
  },
  {
    "loss": 1.83,
    "grad_norm": 4.432948112487793,
    "learning_rate": 1.707863115312227e-05,
    "epoch": 1.958217270194986,
    "step": 352
  },
  {
    "loss": 2.3556,
    "grad_norm": 5.213648796081543,
    "learning_rate": 1.706548363106444e-05,
    "epoch": 1.9637883008356547,
    "step": 353
  },
  {
    "loss": 2.1396,
    "grad_norm": 4.155479431152344,
    "learning_rate": 1.705233610900661e-05,
    "epoch": 1.9693593314763231,
    "step": 354
  },
  {
    "loss": 1.4222,
    "grad_norm": 5.146358013153076,
    "learning_rate": 1.7039188586948775e-05,
    "epoch": 1.9749303621169916,
    "step": 355
  },
  {
    "loss": 1.9362,
    "grad_norm": 3.264761447906494,
    "learning_rate": 1.7026041064890945e-05,
    "epoch": 1.98050139275766,
    "step": 356
  },
  {
    "loss": 1.9471,
    "grad_norm": 3.308243989944458,
    "learning_rate": 1.7012893542833114e-05,
    "epoch": 1.9860724233983287,
    "step": 357
  },
  {
    "loss": 2.0193,
    "grad_norm": 4.1630859375,
    "learning_rate": 1.6999746020775284e-05,
    "epoch": 1.9916434540389973,
    "step": 358
  },
  {
    "loss": 2.1048,
    "grad_norm": 4.196152210235596,
    "learning_rate": 1.6986598498717453e-05,
    "epoch": 1.9972144846796658,
    "step": 359
  },
  {
    "loss": 2.0755,
    "grad_norm": 4.194087028503418,
    "learning_rate": 1.6973450976659623e-05,
    "epoch": 2.0,
    "step": 360
  },
  {
    "loss": 1.4388,
    "grad_norm": 4.208454132080078,
    "learning_rate": 1.6960303454601792e-05,
    "epoch": 2.0055710306406684,
    "step": 361
  },
  {
    "loss": 1.7819,
    "grad_norm": 3.549447774887085,
    "learning_rate": 1.694715593254396e-05,
    "epoch": 2.011142061281337,
    "step": 362
  },
  {
    "loss": 1.5135,
    "grad_norm": 3.6767420768737793,
    "learning_rate": 1.693400841048613e-05,
    "epoch": 2.0167130919220058,
    "step": 363
  },
  {
    "loss": 1.7713,
    "grad_norm": 3.816209554672241,
    "learning_rate": 1.69208608884283e-05,
    "epoch": 2.022284122562674,
    "step": 364
  },
  {
    "loss": 1.6624,
    "grad_norm": 3.2220561504364014,
    "learning_rate": 1.6907713366370467e-05,
    "epoch": 2.0278551532033426,
    "step": 365
  },
  {
    "loss": 1.9059,
    "grad_norm": 3.4210987091064453,
    "learning_rate": 1.6894565844312637e-05,
    "epoch": 2.033426183844011,
    "step": 366
  },
  {
    "loss": 1.155,
    "grad_norm": 4.348776817321777,
    "learning_rate": 1.6881418322254806e-05,
    "epoch": 2.0389972144846795,
    "step": 367
  },
  {
    "loss": 1.4513,
    "grad_norm": 4.143118858337402,
    "learning_rate": 1.6868270800196976e-05,
    "epoch": 2.0445682451253484,
    "step": 368
  },
  {
    "loss": 1.8148,
    "grad_norm": 4.118925094604492,
    "learning_rate": 1.6855123278139145e-05,
    "epoch": 2.050139275766017,
    "step": 369
  },
  {
    "loss": 1.6325,
    "grad_norm": 4.060324668884277,
    "learning_rate": 1.6841975756081315e-05,
    "epoch": 2.0557103064066853,
    "step": 370
  },
  {
    "loss": 1.694,
    "grad_norm": 4.604481220245361,
    "learning_rate": 1.6828828234023484e-05,
    "epoch": 2.0612813370473537,
    "step": 371
  },
  {
    "loss": 1.4905,
    "grad_norm": 5.273688316345215,
    "learning_rate": 1.681568071196565e-05,
    "epoch": 2.066852367688022,
    "step": 372
  },
  {
    "loss": 1.1557,
    "grad_norm": 6.0254387855529785,
    "learning_rate": 1.680253318990782e-05,
    "epoch": 2.0724233983286906,
    "step": 373
  },
  {
    "loss": 0.999,
    "grad_norm": 5.017882823944092,
    "learning_rate": 1.678938566784999e-05,
    "epoch": 2.0779944289693595,
    "step": 374
  },
  {
    "loss": 1.4159,
    "grad_norm": 6.874935626983643,
    "learning_rate": 1.6776238145792162e-05,
    "epoch": 2.083565459610028,
    "step": 375
  },
  {
    "loss": 0.9789,
    "grad_norm": 6.245709419250488,
    "learning_rate": 1.676309062373433e-05,
    "epoch": 2.0891364902506964,
    "step": 376
  },
  {
    "loss": 1.3929,
    "grad_norm": 6.976832866668701,
    "learning_rate": 1.6749943101676498e-05,
    "epoch": 2.094707520891365,
    "step": 377
  },
  {
    "loss": 1.5721,
    "grad_norm": 7.426636695861816,
    "learning_rate": 1.6736795579618668e-05,
    "epoch": 2.1002785515320332,
    "step": 378
  },
  {
    "loss": 1.4603,
    "grad_norm": 8.876333236694336,
    "learning_rate": 1.6723648057560837e-05,
    "epoch": 2.105849582172702,
    "step": 379
  },
  {
    "loss": 1.2115,
    "grad_norm": 5.889682769775391,
    "learning_rate": 1.6710500535503007e-05,
    "epoch": 2.1114206128133706,
    "step": 380
  },
  {
    "loss": 1.1689,
    "grad_norm": 6.435322284698486,
    "learning_rate": 1.6697353013445176e-05,
    "epoch": 2.116991643454039,
    "step": 381
  },
  {
    "loss": 1.1904,
    "grad_norm": 6.061446666717529,
    "learning_rate": 1.6684205491387342e-05,
    "epoch": 2.1225626740947074,
    "step": 382
  },
  {
    "loss": 1.3799,
    "grad_norm": 7.56770658493042,
    "learning_rate": 1.6671057969329512e-05,
    "epoch": 2.128133704735376,
    "step": 383
  },
  {
    "loss": 1.5787,
    "grad_norm": 8.942233085632324,
    "learning_rate": 1.665791044727168e-05,
    "epoch": 2.1337047353760448,
    "step": 384
  },
  {
    "loss": 1.4084,
    "grad_norm": 7.448763847351074,
    "learning_rate": 1.664476292521385e-05,
    "epoch": 2.139275766016713,
    "step": 385
  },
  {
    "loss": 1.3685,
    "grad_norm": 5.792154312133789,
    "learning_rate": 1.663161540315602e-05,
    "epoch": 2.1448467966573816,
    "step": 386
  },
  {
    "loss": 1.5465,
    "grad_norm": 7.226157188415527,
    "learning_rate": 1.661846788109819e-05,
    "epoch": 2.15041782729805,
    "step": 387
  },
  {
    "loss": 1.1914,
    "grad_norm": 5.6042022705078125,
    "learning_rate": 1.660532035904036e-05,
    "epoch": 2.1559888579387185,
    "step": 388
  },
  {
    "loss": 1.6443,
    "grad_norm": 5.619427680969238,
    "learning_rate": 1.6592172836982525e-05,
    "epoch": 2.1615598885793874,
    "step": 389
  },
  {
    "loss": 1.5371,
    "grad_norm": 4.770148754119873,
    "learning_rate": 1.65790253149247e-05,
    "epoch": 2.167130919220056,
    "step": 390
  },
  {
    "loss": 1.5124,
    "grad_norm": 7.61703634262085,
    "learning_rate": 1.6565877792866868e-05,
    "epoch": 2.1727019498607243,
    "step": 391
  },
  {
    "loss": 1.6248,
    "grad_norm": 4.498234272003174,
    "learning_rate": 1.6552730270809037e-05,
    "epoch": 2.1782729805013927,
    "step": 392
  },
  {
    "loss": 1.4621,
    "grad_norm": 4.0563063621521,
    "learning_rate": 1.6539582748751204e-05,
    "epoch": 2.183844011142061,
    "step": 393
  },
  {
    "loss": 1.4315,
    "grad_norm": 6.069952964782715,
    "learning_rate": 1.6526435226693373e-05,
    "epoch": 2.1894150417827296,
    "step": 394
  },
  {
    "loss": 1.4308,
    "grad_norm": 6.728673458099365,
    "learning_rate": 1.6513287704635543e-05,
    "epoch": 2.1949860724233985,
    "step": 395
  },
  {
    "loss": 1.2975,
    "grad_norm": 14.551620483398438,
    "learning_rate": 1.6500140182577712e-05,
    "epoch": 2.200557103064067,
    "step": 396
  },
  {
    "loss": 1.4624,
    "grad_norm": 6.782831192016602,
    "learning_rate": 1.648699266051988e-05,
    "epoch": 2.2061281337047354,
    "step": 397
  },
  {
    "loss": 1.5891,
    "grad_norm": 6.513261795043945,
    "learning_rate": 1.647384513846205e-05,
    "epoch": 2.211699164345404,
    "step": 398
  },
  {
    "loss": 1.3152,
    "grad_norm": 6.3476433753967285,
    "learning_rate": 1.646069761640422e-05,
    "epoch": 2.2172701949860723,
    "step": 399
  },
  {
    "loss": 1.3129,
    "grad_norm": 4.936390399932861,
    "learning_rate": 1.6447550094346387e-05,
    "epoch": 2.222841225626741,
    "step": 400
  },
  {
    "eval_loss": 2.531832218170166,
    "eval_runtime": 35.95,
    "eval_samples_per_second": 39.944,
    "eval_steps_per_second": 2.003,
    "epoch": 2.222841225626741,
    "step": 400
  },
  {
    "loss": 1.2283,
    "grad_norm": 8.302631378173828,
    "learning_rate": 1.6434402572288556e-05,
    "epoch": 2.2284122562674096,
    "step": 401
  },
  {
    "loss": 1.1884,
    "grad_norm": 5.8890886306762695,
    "learning_rate": 1.642125505023073e-05,
    "epoch": 2.233983286908078,
    "step": 402
  },
  {
    "loss": 1.3971,
    "grad_norm": 6.417287349700928,
    "learning_rate": 1.6408107528172895e-05,
    "epoch": 2.2395543175487465,
    "step": 403
  },
  {
    "loss": 1.5501,
    "grad_norm": 6.351545810699463,
    "learning_rate": 1.6394960006115065e-05,
    "epoch": 2.245125348189415,
    "step": 404
  },
  {
    "loss": 1.1685,
    "grad_norm": 5.121798992156982,
    "learning_rate": 1.6381812484057234e-05,
    "epoch": 2.2506963788300833,
    "step": 405
  },
  {
    "loss": 1.3617,
    "grad_norm": 5.293002128601074,
    "learning_rate": 1.6368664961999404e-05,
    "epoch": 2.256267409470752,
    "step": 406
  },
  {
    "loss": 1.3164,
    "grad_norm": 6.6434431076049805,
    "learning_rate": 1.6355517439941573e-05,
    "epoch": 2.2618384401114207,
    "step": 407
  },
  {
    "loss": 1.4339,
    "grad_norm": 6.383541584014893,
    "learning_rate": 1.6342369917883743e-05,
    "epoch": 2.267409470752089,
    "step": 408
  },
  {
    "loss": 1.3699,
    "grad_norm": 5.989224433898926,
    "learning_rate": 1.6329222395825913e-05,
    "epoch": 2.2729805013927575,
    "step": 409
  },
  {
    "loss": 1.4938,
    "grad_norm": 6.49315881729126,
    "learning_rate": 1.631607487376808e-05,
    "epoch": 2.2785515320334264,
    "step": 410
  },
  {
    "loss": 1.0902,
    "grad_norm": 4.942923069000244,
    "learning_rate": 1.6302927351710248e-05,
    "epoch": 2.284122562674095,
    "step": 411
  },
  {
    "loss": 1.0282,
    "grad_norm": 5.219899654388428,
    "learning_rate": 1.6289779829652418e-05,
    "epoch": 2.2896935933147633,
    "step": 412
  },
  {
    "loss": 1.3465,
    "grad_norm": 5.91557502746582,
    "learning_rate": 1.6276632307594587e-05,
    "epoch": 2.2952646239554317,
    "step": 413
  },
  {
    "loss": 1.4312,
    "grad_norm": 7.332894325256348,
    "learning_rate": 1.6263484785536757e-05,
    "epoch": 2.3008356545961,
    "step": 414
  },
  {
    "loss": 1.1921,
    "grad_norm": 6.784351825714111,
    "learning_rate": 1.6250337263478926e-05,
    "epoch": 2.3064066852367686,
    "step": 415
  },
  {
    "loss": 1.3644,
    "grad_norm": 6.222668647766113,
    "learning_rate": 1.6237189741421096e-05,
    "epoch": 2.3119777158774375,
    "step": 416
  },
  {
    "loss": 1.3318,
    "grad_norm": 6.7379841804504395,
    "learning_rate": 1.6224042219363265e-05,
    "epoch": 2.317548746518106,
    "step": 417
  },
  {
    "loss": 1.3955,
    "grad_norm": 7.218482494354248,
    "learning_rate": 1.6210894697305435e-05,
    "epoch": 2.3231197771587744,
    "step": 418
  },
  {
    "loss": 1.5949,
    "grad_norm": 6.676080226898193,
    "learning_rate": 1.6197747175247604e-05,
    "epoch": 2.328690807799443,
    "step": 419
  },
  {
    "loss": 1.2428,
    "grad_norm": 6.974861145019531,
    "learning_rate": 1.618459965318977e-05,
    "epoch": 2.3342618384401113,
    "step": 420
  },
  {
    "loss": 1.2438,
    "grad_norm": 7.018064975738525,
    "learning_rate": 1.617145213113194e-05,
    "epoch": 2.33983286908078,
    "step": 421
  },
  {
    "loss": 1.4979,
    "grad_norm": 6.781156063079834,
    "learning_rate": 1.615830460907411e-05,
    "epoch": 2.3454038997214486,
    "step": 422
  },
  {
    "loss": 1.4914,
    "grad_norm": 6.291943550109863,
    "learning_rate": 1.614515708701628e-05,
    "epoch": 2.350974930362117,
    "step": 423
  },
  {
    "loss": 1.1937,
    "grad_norm": 6.769220352172852,
    "learning_rate": 1.613200956495845e-05,
    "epoch": 2.3565459610027855,
    "step": 424
  },
  {
    "loss": 1.4428,
    "grad_norm": 7.461434841156006,
    "learning_rate": 1.6118862042900618e-05,
    "epoch": 2.362116991643454,
    "step": 425
  },
  {
    "loss": 1.0756,
    "grad_norm": 5.971315860748291,
    "learning_rate": 1.6105714520842788e-05,
    "epoch": 2.3676880222841223,
    "step": 426
  },
  {
    "loss": 1.1709,
    "grad_norm": 6.632075786590576,
    "learning_rate": 1.6092566998784954e-05,
    "epoch": 2.3732590529247912,
    "step": 427
  },
  {
    "loss": 1.2953,
    "grad_norm": 6.03197717666626,
    "learning_rate": 1.6079419476727123e-05,
    "epoch": 2.3788300835654597,
    "step": 428
  },
  {
    "loss": 1.1653,
    "grad_norm": 7.393289089202881,
    "learning_rate": 1.6066271954669296e-05,
    "epoch": 2.384401114206128,
    "step": 429
  },
  {
    "loss": 1.542,
    "grad_norm": 9.518671989440918,
    "learning_rate": 1.6053124432611462e-05,
    "epoch": 2.3899721448467965,
    "step": 430
  },
  {
    "loss": 1.0957,
    "grad_norm": 7.086347579956055,
    "learning_rate": 1.6039976910553632e-05,
    "epoch": 2.3955431754874654,
    "step": 431
  },
  {
    "loss": 1.1408,
    "grad_norm": 5.21544885635376,
    "learning_rate": 1.60268293884958e-05,
    "epoch": 2.401114206128134,
    "step": 432
  },
  {
    "loss": 1.1708,
    "grad_norm": 7.537359237670898,
    "learning_rate": 1.601368186643797e-05,
    "epoch": 2.4066852367688023,
    "step": 433
  },
  {
    "loss": 1.101,
    "grad_norm": 4.926475524902344,
    "learning_rate": 1.600053434438014e-05,
    "epoch": 2.4122562674094707,
    "step": 434
  },
  {
    "loss": 1.3898,
    "grad_norm": 5.6016740798950195,
    "learning_rate": 1.598738682232231e-05,
    "epoch": 2.417827298050139,
    "step": 435
  },
  {
    "loss": 1.4717,
    "grad_norm": 7.16878604888916,
    "learning_rate": 1.597423930026448e-05,
    "epoch": 2.4233983286908076,
    "step": 436
  },
  {
    "loss": 1.6173,
    "grad_norm": 6.310802459716797,
    "learning_rate": 1.5961091778206646e-05,
    "epoch": 2.4289693593314765,
    "step": 437
  },
  {
    "loss": 1.6172,
    "grad_norm": 8.035069465637207,
    "learning_rate": 1.5947944256148815e-05,
    "epoch": 2.434540389972145,
    "step": 438
  },
  {
    "loss": 1.4479,
    "grad_norm": 7.806406497955322,
    "learning_rate": 1.5934796734090985e-05,
    "epoch": 2.4401114206128134,
    "step": 439
  },
  {
    "loss": 1.3459,
    "grad_norm": 5.882315635681152,
    "learning_rate": 1.5921649212033154e-05,
    "epoch": 2.445682451253482,
    "step": 440
  },
  {
    "loss": 1.2195,
    "grad_norm": 5.817505359649658,
    "learning_rate": 1.5908501689975324e-05,
    "epoch": 2.4512534818941503,
    "step": 441
  },
  {
    "loss": 1.3043,
    "grad_norm": 7.497400283813477,
    "learning_rate": 1.5895354167917493e-05,
    "epoch": 2.456824512534819,
    "step": 442
  },
  {
    "loss": 1.42,
    "grad_norm": 5.955392837524414,
    "learning_rate": 1.5882206645859663e-05,
    "epoch": 2.4623955431754876,
    "step": 443
  },
  {
    "loss": 1.4764,
    "grad_norm": 8.848158836364746,
    "learning_rate": 1.5869059123801832e-05,
    "epoch": 2.467966573816156,
    "step": 444
  },
  {
    "loss": 1.4508,
    "grad_norm": 6.384143829345703,
    "learning_rate": 1.5855911601744002e-05,
    "epoch": 2.4735376044568245,
    "step": 445
  },
  {
    "loss": 1.3499,
    "grad_norm": 7.251498699188232,
    "learning_rate": 1.584276407968617e-05,
    "epoch": 2.479108635097493,
    "step": 446
  },
  {
    "loss": 1.297,
    "grad_norm": 8.700945854187012,
    "learning_rate": 1.5829616557628337e-05,
    "epoch": 2.4846796657381613,
    "step": 447
  },
  {
    "loss": 1.1607,
    "grad_norm": 8.17098617553711,
    "learning_rate": 1.5816469035570507e-05,
    "epoch": 2.4902506963788302,
    "step": 448
  },
  {
    "loss": 1.5328,
    "grad_norm": 6.918285846710205,
    "learning_rate": 1.5803321513512676e-05,
    "epoch": 2.4958217270194987,
    "step": 449
  },
  {
    "loss": 1.6258,
    "grad_norm": 6.7390851974487305,
    "learning_rate": 1.5790173991454846e-05,
    "epoch": 2.501392757660167,
    "step": 450
  },
  {
    "eval_loss": 2.571645498275757,
    "eval_runtime": 35.9556,
    "eval_samples_per_second": 39.938,
    "eval_steps_per_second": 2.002,
    "epoch": 2.501392757660167,
    "step": 450
  },
  {
    "loss": 1.5923,
    "grad_norm": 6.522182941436768,
    "learning_rate": 1.5777026469397015e-05,
    "epoch": 2.5069637883008355,
    "step": 451
  },
  {
    "loss": 1.2816,
    "grad_norm": 5.984560489654541,
    "learning_rate": 1.5763878947339185e-05,
    "epoch": 2.5125348189415044,
    "step": 452
  },
  {
    "loss": 1.2029,
    "grad_norm": 8.060498237609863,
    "learning_rate": 1.5750731425281354e-05,
    "epoch": 2.518105849582173,
    "step": 453
  },
  {
    "loss": 1.2117,
    "grad_norm": 6.93899393081665,
    "learning_rate": 1.573758390322352e-05,
    "epoch": 2.5236768802228413,
    "step": 454
  },
  {
    "loss": 1.4347,
    "grad_norm": 6.21560525894165,
    "learning_rate": 1.572443638116569e-05,
    "epoch": 2.5292479108635098,
    "step": 455
  },
  {
    "loss": 1.3394,
    "grad_norm": 7.837366580963135,
    "learning_rate": 1.5711288859107863e-05,
    "epoch": 2.534818941504178,
    "step": 456
  },
  {
    "loss": 1.4262,
    "grad_norm": 7.609643936157227,
    "learning_rate": 1.5698141337050033e-05,
    "epoch": 2.5403899721448466,
    "step": 457
  },
  {
    "loss": 1.3738,
    "grad_norm": 6.487556457519531,
    "learning_rate": 1.56849938149922e-05,
    "epoch": 2.545961002785515,
    "step": 458
  },
  {
    "loss": 1.4021,
    "grad_norm": 6.344869136810303,
    "learning_rate": 1.5671846292934368e-05,
    "epoch": 2.551532033426184,
    "step": 459
  },
  {
    "loss": 1.3887,
    "grad_norm": 6.960203170776367,
    "learning_rate": 1.5658698770876538e-05,
    "epoch": 2.5571030640668524,
    "step": 460
  },
  {
    "loss": 1.2997,
    "grad_norm": 11.57795524597168,
    "learning_rate": 1.5645551248818707e-05,
    "epoch": 2.562674094707521,
    "step": 461
  },
  {
    "loss": 1.5967,
    "grad_norm": 6.889705181121826,
    "learning_rate": 1.5632403726760877e-05,
    "epoch": 2.5682451253481893,
    "step": 462
  },
  {
    "loss": 1.2643,
    "grad_norm": 8.502350807189941,
    "learning_rate": 1.5619256204703046e-05,
    "epoch": 2.573816155988858,
    "step": 463
  },
  {
    "loss": 1.3686,
    "grad_norm": 8.704366683959961,
    "learning_rate": 1.5606108682645216e-05,
    "epoch": 2.5793871866295266,
    "step": 464
  },
  {
    "loss": 0.9961,
    "grad_norm": 8.154948234558105,
    "learning_rate": 1.5592961160587382e-05,
    "epoch": 2.584958217270195,
    "step": 465
  },
  {
    "loss": 1.0603,
    "grad_norm": 5.729700088500977,
    "learning_rate": 1.557981363852955e-05,
    "epoch": 2.5905292479108635,
    "step": 466
  },
  {
    "loss": 1.6641,
    "grad_norm": 7.716269493103027,
    "learning_rate": 1.556666611647172e-05,
    "epoch": 2.596100278551532,
    "step": 467
  },
  {
    "loss": 1.2886,
    "grad_norm": 11.220166206359863,
    "learning_rate": 1.555351859441389e-05,
    "epoch": 2.6016713091922004,
    "step": 468
  },
  {
    "loss": 1.2922,
    "grad_norm": 7.163726329803467,
    "learning_rate": 1.554037107235606e-05,
    "epoch": 2.6072423398328692,
    "step": 469
  },
  {
    "loss": 1.1046,
    "grad_norm": 7.28581428527832,
    "learning_rate": 1.552722355029823e-05,
    "epoch": 2.6128133704735377,
    "step": 470
  },
  {
    "loss": 1.6142,
    "grad_norm": 9.65365219116211,
    "learning_rate": 1.5514076028240396e-05,
    "epoch": 2.618384401114206,
    "step": 471
  },
  {
    "loss": 1.5575,
    "grad_norm": 6.458492279052734,
    "learning_rate": 1.550092850618257e-05,
    "epoch": 2.6239554317548746,
    "step": 472
  },
  {
    "loss": 1.3655,
    "grad_norm": 7.325246810913086,
    "learning_rate": 1.5487780984124738e-05,
    "epoch": 2.6295264623955434,
    "step": 473
  },
  {
    "loss": 1.3344,
    "grad_norm": 7.81355619430542,
    "learning_rate": 1.5474633462066908e-05,
    "epoch": 2.635097493036212,
    "step": 474
  },
  {
    "loss": 1.2505,
    "grad_norm": 7.347303867340088,
    "learning_rate": 1.5461485940009074e-05,
    "epoch": 2.6406685236768803,
    "step": 475
  },
  {
    "loss": 1.1988,
    "grad_norm": 7.306774616241455,
    "learning_rate": 1.5448338417951243e-05,
    "epoch": 2.6462395543175488,
    "step": 476
  },
  {
    "loss": 1.4075,
    "grad_norm": 7.261951446533203,
    "learning_rate": 1.5435190895893413e-05,
    "epoch": 2.651810584958217,
    "step": 477
  },
  {
    "loss": 1.3235,
    "grad_norm": 8.138806343078613,
    "learning_rate": 1.5422043373835582e-05,
    "epoch": 2.6573816155988856,
    "step": 478
  },
  {
    "loss": 1.4297,
    "grad_norm": 7.515624046325684,
    "learning_rate": 1.5408895851777752e-05,
    "epoch": 2.662952646239554,
    "step": 479
  },
  {
    "loss": 1.0187,
    "grad_norm": 7.298752307891846,
    "learning_rate": 1.539574832971992e-05,
    "epoch": 2.668523676880223,
    "step": 480
  },
  {
    "loss": 1.1512,
    "grad_norm": 7.08530855178833,
    "learning_rate": 1.538260080766209e-05,
    "epoch": 2.6740947075208914,
    "step": 481
  },
  {
    "loss": 0.9209,
    "grad_norm": 8.528051376342773,
    "learning_rate": 1.5369453285604257e-05,
    "epoch": 2.67966573816156,
    "step": 482
  },
  {
    "loss": 1.6726,
    "grad_norm": 6.991207122802734,
    "learning_rate": 1.535630576354643e-05,
    "epoch": 2.6852367688022283,
    "step": 483
  },
  {
    "loss": 1.6101,
    "grad_norm": 6.910933971405029,
    "learning_rate": 1.53431582414886e-05,
    "epoch": 2.690807799442897,
    "step": 484
  },
  {
    "loss": 1.0596,
    "grad_norm": 6.858171463012695,
    "learning_rate": 1.5330010719430766e-05,
    "epoch": 2.6963788300835656,
    "step": 485
  },
  {
    "loss": 1.3009,
    "grad_norm": 7.1738409996032715,
    "learning_rate": 1.5316863197372935e-05,
    "epoch": 2.701949860724234,
    "step": 486
  },
  {
    "loss": 1.1306,
    "grad_norm": 6.751303672790527,
    "learning_rate": 1.5303715675315105e-05,
    "epoch": 2.7075208913649025,
    "step": 487
  },
  {
    "loss": 1.6064,
    "grad_norm": 7.458596706390381,
    "learning_rate": 1.5290568153257274e-05,
    "epoch": 2.713091922005571,
    "step": 488
  },
  {
    "loss": 1.3423,
    "grad_norm": 4.847519397735596,
    "learning_rate": 1.5277420631199444e-05,
    "epoch": 2.7186629526462394,
    "step": 489
  },
  {
    "loss": 1.0908,
    "grad_norm": 6.585028648376465,
    "learning_rate": 1.5264273109141613e-05,
    "epoch": 2.724233983286908,
    "step": 490
  },
  {
    "loss": 1.6632,
    "grad_norm": 5.222984790802002,
    "learning_rate": 1.5251125587083783e-05,
    "epoch": 2.7298050139275767,
    "step": 491
  },
  {
    "loss": 1.3113,
    "grad_norm": 6.947058200836182,
    "learning_rate": 1.523797806502595e-05,
    "epoch": 2.735376044568245,
    "step": 492
  },
  {
    "loss": 1.0863,
    "grad_norm": 5.885672569274902,
    "learning_rate": 1.522483054296812e-05,
    "epoch": 2.7409470752089136,
    "step": 493
  },
  {
    "loss": 1.1982,
    "grad_norm": 7.9502034187316895,
    "learning_rate": 1.521168302091029e-05,
    "epoch": 2.7465181058495824,
    "step": 494
  },
  {
    "loss": 1.3941,
    "grad_norm": 5.9523773193359375,
    "learning_rate": 1.5198535498852457e-05,
    "epoch": 2.752089136490251,
    "step": 495
  },
  {
    "loss": 1.3251,
    "grad_norm": 7.984345436096191,
    "learning_rate": 1.5185387976794627e-05,
    "epoch": 2.7576601671309193,
    "step": 496
  },
  {
    "loss": 0.8109,
    "grad_norm": 8.467183113098145,
    "learning_rate": 1.5172240454736796e-05,
    "epoch": 2.7632311977715878,
    "step": 497
  },
  {
    "loss": 1.1339,
    "grad_norm": 7.878790378570557,
    "learning_rate": 1.5159092932678966e-05,
    "epoch": 2.768802228412256,
    "step": 498
  },
  {
    "loss": 1.1736,
    "grad_norm": 5.638209819793701,
    "learning_rate": 1.5145945410621134e-05,
    "epoch": 2.7743732590529246,
    "step": 499
  },
  {
    "loss": 1.3546,
    "grad_norm": 7.818211078643799,
    "learning_rate": 1.5132797888563303e-05,
    "epoch": 2.779944289693593,
    "step": 500
  },
  {
    "eval_loss": 2.6166257858276367,
    "eval_runtime": 35.971,
    "eval_samples_per_second": 39.921,
    "eval_steps_per_second": 2.002,
    "epoch": 2.779944289693593,
    "step": 500
  },
  {
    "loss": 1.4636,
    "grad_norm": 6.118830680847168,
    "learning_rate": 1.5119650366505473e-05,
    "epoch": 2.785515320334262,
    "step": 501
  },
  {
    "loss": 1.5519,
    "grad_norm": 7.9165778160095215,
    "learning_rate": 1.510650284444764e-05,
    "epoch": 2.7910863509749304,
    "step": 502
  },
  {
    "loss": 1.5206,
    "grad_norm": 6.975761413574219,
    "learning_rate": 1.5093355322389812e-05,
    "epoch": 2.796657381615599,
    "step": 503
  },
  {
    "loss": 1.0665,
    "grad_norm": 9.277933120727539,
    "learning_rate": 1.5080207800331981e-05,
    "epoch": 2.8022284122562673,
    "step": 504
  },
  {
    "loss": 1.2801,
    "grad_norm": 8.121682167053223,
    "learning_rate": 1.5067060278274151e-05,
    "epoch": 2.807799442896936,
    "step": 505
  },
  {
    "loss": 1.565,
    "grad_norm": 8.76021957397461,
    "learning_rate": 1.5053912756216319e-05,
    "epoch": 2.8133704735376046,
    "step": 506
  },
  {
    "loss": 1.3502,
    "grad_norm": 8.618566513061523,
    "learning_rate": 1.5040765234158488e-05,
    "epoch": 2.818941504178273,
    "step": 507
  },
  {
    "loss": 1.5859,
    "grad_norm": 8.027894020080566,
    "learning_rate": 1.5027617712100658e-05,
    "epoch": 2.8245125348189415,
    "step": 508
  },
  {
    "loss": 1.4159,
    "grad_norm": 7.063473701477051,
    "learning_rate": 1.5014470190042826e-05,
    "epoch": 2.83008356545961,
    "step": 509
  },
  {
    "loss": 1.5672,
    "grad_norm": 6.095931053161621,
    "learning_rate": 1.5001322667984995e-05,
    "epoch": 2.8356545961002784,
    "step": 510
  },
  {
    "loss": 1.2551,
    "grad_norm": 6.445271968841553,
    "learning_rate": 1.4988175145927165e-05,
    "epoch": 2.841225626740947,
    "step": 511
  },
  {
    "loss": 0.9671,
    "grad_norm": 7.601891040802002,
    "learning_rate": 1.4975027623869334e-05,
    "epoch": 2.8467966573816157,
    "step": 512
  },
  {
    "loss": 1.4462,
    "grad_norm": 8.017728805541992,
    "learning_rate": 1.4961880101811502e-05,
    "epoch": 2.852367688022284,
    "step": 513
  },
  {
    "loss": 1.2006,
    "grad_norm": 6.753676891326904,
    "learning_rate": 1.4948732579753672e-05,
    "epoch": 2.8579387186629526,
    "step": 514
  },
  {
    "loss": 0.9354,
    "grad_norm": 6.220627784729004,
    "learning_rate": 1.4935585057695843e-05,
    "epoch": 2.863509749303621,
    "step": 515
  },
  {
    "loss": 1.5554,
    "grad_norm": 7.825878620147705,
    "learning_rate": 1.4922437535638009e-05,
    "epoch": 2.86908077994429,
    "step": 516
  },
  {
    "loss": 0.9281,
    "grad_norm": 7.7669548988342285,
    "learning_rate": 1.490929001358018e-05,
    "epoch": 2.8746518105849583,
    "step": 517
  },
  {
    "loss": 1.1678,
    "grad_norm": 6.18816614151001,
    "learning_rate": 1.489614249152235e-05,
    "epoch": 2.8802228412256268,
    "step": 518
  },
  {
    "loss": 0.8378,
    "grad_norm": 11.241938591003418,
    "learning_rate": 1.4882994969464517e-05,
    "epoch": 2.885793871866295,
    "step": 519
  },
  {
    "loss": 1.6602,
    "grad_norm": 6.708087921142578,
    "learning_rate": 1.4869847447406687e-05,
    "epoch": 2.8913649025069637,
    "step": 520
  },
  {
    "loss": 1.5575,
    "grad_norm": 8.96353530883789,
    "learning_rate": 1.4856699925348856e-05,
    "epoch": 2.896935933147632,
    "step": 521
  },
  {
    "loss": 1.3553,
    "grad_norm": 7.286456108093262,
    "learning_rate": 1.4843552403291026e-05,
    "epoch": 2.902506963788301,
    "step": 522
  },
  {
    "loss": 1.3618,
    "grad_norm": 6.448929309844971,
    "learning_rate": 1.4830404881233194e-05,
    "epoch": 2.9080779944289694,
    "step": 523
  },
  {
    "loss": 1.0911,
    "grad_norm": 6.1524739265441895,
    "learning_rate": 1.4817257359175363e-05,
    "epoch": 2.913649025069638,
    "step": 524
  },
  {
    "loss": 1.2465,
    "grad_norm": 6.833171367645264,
    "learning_rate": 1.4804109837117533e-05,
    "epoch": 2.9192200557103063,
    "step": 525
  },
  {
    "loss": 0.9937,
    "grad_norm": 8.745670318603516,
    "learning_rate": 1.47909623150597e-05,
    "epoch": 2.924791086350975,
    "step": 526
  },
  {
    "loss": 1.3931,
    "grad_norm": 6.3659186363220215,
    "learning_rate": 1.477781479300187e-05,
    "epoch": 2.9303621169916436,
    "step": 527
  },
  {
    "loss": 1.029,
    "grad_norm": 8.309256553649902,
    "learning_rate": 1.476466727094404e-05,
    "epoch": 2.935933147632312,
    "step": 528
  },
  {
    "loss": 1.383,
    "grad_norm": 7.611057758331299,
    "learning_rate": 1.4751519748886211e-05,
    "epoch": 2.9415041782729805,
    "step": 529
  },
  {
    "loss": 1.4833,
    "grad_norm": 9.441068649291992,
    "learning_rate": 1.4738372226828379e-05,
    "epoch": 2.947075208913649,
    "step": 530
  },
  {
    "loss": 1.2739,
    "grad_norm": 7.198431968688965,
    "learning_rate": 1.4725224704770548e-05,
    "epoch": 2.9526462395543174,
    "step": 531
  },
  {
    "loss": 1.4294,
    "grad_norm": 8.88117790222168,
    "learning_rate": 1.4712077182712718e-05,
    "epoch": 2.958217270194986,
    "step": 532
  },
  {
    "loss": 1.0204,
    "grad_norm": 9.982294082641602,
    "learning_rate": 1.4698929660654886e-05,
    "epoch": 2.9637883008356547,
    "step": 533
  },
  {
    "loss": 1.1488,
    "grad_norm": 8.535533905029297,
    "learning_rate": 1.4685782138597055e-05,
    "epoch": 2.969359331476323,
    "step": 534
  },
  {
    "loss": 1.49,
    "grad_norm": 6.813885688781738,
    "learning_rate": 1.4672634616539225e-05,
    "epoch": 2.9749303621169916,
    "step": 535
  },
  {
    "loss": 1.113,
    "grad_norm": 9.557439804077148,
    "learning_rate": 1.4659487094481394e-05,
    "epoch": 2.98050139275766,
    "step": 536
  },
  {
    "loss": 1.5004,
    "grad_norm": 6.406128883361816,
    "learning_rate": 1.4646339572423562e-05,
    "epoch": 2.986072423398329,
    "step": 537
  },
  {
    "loss": 1.1722,
    "grad_norm": 7.9670915603637695,
    "learning_rate": 1.4633192050365732e-05,
    "epoch": 2.9916434540389973,
    "step": 538
  },
  {
    "loss": 1.5033,
    "grad_norm": 9.402728080749512,
    "learning_rate": 1.4620044528307901e-05,
    "epoch": 2.997214484679666,
    "step": 539
  },
  {
    "loss": 1.279,
    "grad_norm": 7.38714075088501,
    "learning_rate": 1.4606897006250069e-05,
    "epoch": 3.0,
    "step": 540
  },
  {
    "loss": 0.6426,
    "grad_norm": 7.639667510986328,
    "learning_rate": 1.4593749484192238e-05,
    "epoch": 3.0055710306406684,
    "step": 541
  },
  {
    "loss": 0.8426,
    "grad_norm": 7.864633560180664,
    "learning_rate": 1.458060196213441e-05,
    "epoch": 3.011142061281337,
    "step": 542
  },
  {
    "loss": 1.0651,
    "grad_norm": 6.637276649475098,
    "learning_rate": 1.4567454440076576e-05,
    "epoch": 3.0167130919220058,
    "step": 543
  },
  {
    "loss": 1.0804,
    "grad_norm": 7.148686408996582,
    "learning_rate": 1.4554306918018747e-05,
    "epoch": 3.022284122562674,
    "step": 544
  },
  {
    "loss": 1.0182,
    "grad_norm": 6.767364501953125,
    "learning_rate": 1.4541159395960917e-05,
    "epoch": 3.0278551532033426,
    "step": 545
  },
  {
    "loss": 0.8165,
    "grad_norm": 6.77062463760376,
    "learning_rate": 1.4528011873903086e-05,
    "epoch": 3.033426183844011,
    "step": 546
  },
  {
    "loss": 0.9941,
    "grad_norm": 8.067922592163086,
    "learning_rate": 1.4514864351845254e-05,
    "epoch": 3.0389972144846795,
    "step": 547
  },
  {
    "loss": 0.9579,
    "grad_norm": 8.817468643188477,
    "learning_rate": 1.4501716829787423e-05,
    "epoch": 3.0445682451253484,
    "step": 548
  },
  {
    "loss": 0.6023,
    "grad_norm": 8.70374870300293,
    "learning_rate": 1.4488569307729593e-05,
    "epoch": 3.050139275766017,
    "step": 549
  },
  {
    "loss": 1.1392,
    "grad_norm": 9.344374656677246,
    "learning_rate": 1.447542178567176e-05,
    "epoch": 3.0557103064066853,
    "step": 550
  },
  {
    "eval_loss": 3.0347440242767334,
    "eval_runtime": 35.9429,
    "eval_samples_per_second": 39.952,
    "eval_steps_per_second": 2.003,
    "epoch": 3.0557103064066853,
    "step": 550
  },
  {
    "loss": 0.663,
    "grad_norm": 10.07166862487793,
    "learning_rate": 1.446227426361393e-05,
    "epoch": 3.0612813370473537,
    "step": 551
  },
  {
    "loss": 1.1801,
    "grad_norm": 14.619653701782227,
    "learning_rate": 1.44491267415561e-05,
    "epoch": 3.066852367688022,
    "step": 552
  },
  {
    "loss": 0.9107,
    "grad_norm": 10.427509307861328,
    "learning_rate": 1.443597921949827e-05,
    "epoch": 3.0724233983286906,
    "step": 553
  },
  {
    "loss": 0.9474,
    "grad_norm": 8.392213821411133,
    "learning_rate": 1.4422831697440437e-05,
    "epoch": 3.0779944289693595,
    "step": 554
  },
  {
    "loss": 0.7972,
    "grad_norm": 13.848929405212402,
    "learning_rate": 1.4409684175382607e-05,
    "epoch": 3.083565459610028,
    "step": 555
  },
  {
    "loss": 0.7298,
    "grad_norm": 9.263422966003418,
    "learning_rate": 1.4396536653324778e-05,
    "epoch": 3.0891364902506964,
    "step": 556
  },
  {
    "loss": 0.5749,
    "grad_norm": 11.082460403442383,
    "learning_rate": 1.4383389131266946e-05,
    "epoch": 3.094707520891365,
    "step": 557
  },
  {
    "loss": 0.7301,
    "grad_norm": 7.7812604904174805,
    "learning_rate": 1.4370241609209115e-05,
    "epoch": 3.1002785515320332,
    "step": 558
  },
  {
    "loss": 0.5884,
    "grad_norm": 12.2935791015625,
    "learning_rate": 1.4357094087151285e-05,
    "epoch": 3.105849582172702,
    "step": 559
  },
  {
    "loss": 0.8034,
    "grad_norm": 8.129678726196289,
    "learning_rate": 1.4343946565093454e-05,
    "epoch": 3.1114206128133706,
    "step": 560
  },
  {
    "loss": 0.6528,
    "grad_norm": 8.628301620483398,
    "learning_rate": 1.4330799043035622e-05,
    "epoch": 3.116991643454039,
    "step": 561
  },
  {
    "loss": 0.8483,
    "grad_norm": 10.514995574951172,
    "learning_rate": 1.4317651520977792e-05,
    "epoch": 3.1225626740947074,
    "step": 562
  },
  {
    "loss": 0.7009,
    "grad_norm": 8.187010765075684,
    "learning_rate": 1.4304503998919961e-05,
    "epoch": 3.128133704735376,
    "step": 563
  },
  {
    "loss": 0.8732,
    "grad_norm": 10.525712013244629,
    "learning_rate": 1.4291356476862129e-05,
    "epoch": 3.1337047353760448,
    "step": 564
  },
  {
    "loss": 0.8319,
    "grad_norm": 9.198347091674805,
    "learning_rate": 1.4278208954804298e-05,
    "epoch": 3.139275766016713,
    "step": 565
  },
  {
    "loss": 0.726,
    "grad_norm": 8.486757278442383,
    "learning_rate": 1.4265061432746468e-05,
    "epoch": 3.1448467966573816,
    "step": 566
  },
  {
    "loss": 0.4993,
    "grad_norm": 8.220407485961914,
    "learning_rate": 1.4251913910688636e-05,
    "epoch": 3.15041782729805,
    "step": 567
  },
  {
    "loss": 1.0212,
    "grad_norm": 7.644767761230469,
    "learning_rate": 1.4238766388630805e-05,
    "epoch": 3.1559888579387185,
    "step": 568
  },
  {
    "loss": 0.8306,
    "grad_norm": 11.287712097167969,
    "learning_rate": 1.4225618866572977e-05,
    "epoch": 3.1615598885793874,
    "step": 569
  },
  {
    "loss": 0.626,
    "grad_norm": 7.9160637855529785,
    "learning_rate": 1.4212471344515146e-05,
    "epoch": 3.167130919220056,
    "step": 570
  },
  {
    "loss": 0.75,
    "grad_norm": 11.988582611083984,
    "learning_rate": 1.4199323822457314e-05,
    "epoch": 3.1727019498607243,
    "step": 571
  },
  {
    "loss": 1.2685,
    "grad_norm": 9.961721420288086,
    "learning_rate": 1.4186176300399483e-05,
    "epoch": 3.1782729805013927,
    "step": 572
  },
  {
    "loss": 0.7429,
    "grad_norm": 12.098424911499023,
    "learning_rate": 1.4173028778341653e-05,
    "epoch": 3.183844011142061,
    "step": 573
  },
  {
    "loss": 1.0086,
    "grad_norm": 8.59049129486084,
    "learning_rate": 1.415988125628382e-05,
    "epoch": 3.1894150417827296,
    "step": 574
  },
  {
    "loss": 1.1215,
    "grad_norm": 10.50232219696045,
    "learning_rate": 1.414673373422599e-05,
    "epoch": 3.1949860724233985,
    "step": 575
  },
  {
    "loss": 0.6729,
    "grad_norm": 11.673900604248047,
    "learning_rate": 1.413358621216816e-05,
    "epoch": 3.200557103064067,
    "step": 576
  },
  {
    "loss": 1.2036,
    "grad_norm": 6.419600009918213,
    "learning_rate": 1.412043869011033e-05,
    "epoch": 3.2061281337047354,
    "step": 577
  },
  {
    "loss": 0.6877,
    "grad_norm": 10.218490600585938,
    "learning_rate": 1.4107291168052497e-05,
    "epoch": 3.211699164345404,
    "step": 578
  },
  {
    "loss": 0.5637,
    "grad_norm": 5.7183918952941895,
    "learning_rate": 1.4094143645994667e-05,
    "epoch": 3.2172701949860723,
    "step": 579
  },
  {
    "loss": 0.7498,
    "grad_norm": 11.460823059082031,
    "learning_rate": 1.4080996123936836e-05,
    "epoch": 3.222841225626741,
    "step": 580
  },
  {
    "loss": 0.6792,
    "grad_norm": 8.623233795166016,
    "learning_rate": 1.4067848601879004e-05,
    "epoch": 3.2284122562674096,
    "step": 581
  },
  {
    "loss": 0.6752,
    "grad_norm": 11.339884757995605,
    "learning_rate": 1.4054701079821174e-05,
    "epoch": 3.233983286908078,
    "step": 582
  },
  {
    "loss": 0.8586,
    "grad_norm": 12.452316284179688,
    "learning_rate": 1.4041553557763345e-05,
    "epoch": 3.2395543175487465,
    "step": 583
  },
  {
    "loss": 0.8345,
    "grad_norm": 6.755831241607666,
    "learning_rate": 1.4028406035705514e-05,
    "epoch": 3.245125348189415,
    "step": 584
  },
  {
    "loss": 0.6932,
    "grad_norm": 9.68067741394043,
    "learning_rate": 1.4015258513647682e-05,
    "epoch": 3.2506963788300833,
    "step": 585
  },
  {
    "loss": 1.2071,
    "grad_norm": 11.948298454284668,
    "learning_rate": 1.4002110991589852e-05,
    "epoch": 3.256267409470752,
    "step": 586
  },
  {
    "loss": 0.7349,
    "grad_norm": 11.49226188659668,
    "learning_rate": 1.3988963469532021e-05,
    "epoch": 3.2618384401114207,
    "step": 587
  },
  {
    "loss": 0.7923,
    "grad_norm": 10.757736206054688,
    "learning_rate": 1.3975815947474189e-05,
    "epoch": 3.267409470752089,
    "step": 588
  },
  {
    "loss": 0.6857,
    "grad_norm": 8.46744441986084,
    "learning_rate": 1.3962668425416358e-05,
    "epoch": 3.2729805013927575,
    "step": 589
  },
  {
    "loss": 0.9153,
    "grad_norm": 6.472330093383789,
    "learning_rate": 1.3949520903358528e-05,
    "epoch": 3.2785515320334264,
    "step": 590
  },
  {
    "loss": 0.7542,
    "grad_norm": 12.151514053344727,
    "learning_rate": 1.3936373381300696e-05,
    "epoch": 3.284122562674095,
    "step": 591
  },
  {
    "loss": 0.9487,
    "grad_norm": 11.680760383605957,
    "learning_rate": 1.3923225859242865e-05,
    "epoch": 3.2896935933147633,
    "step": 592
  },
  {
    "loss": 0.6893,
    "grad_norm": 9.367558479309082,
    "learning_rate": 1.3910078337185035e-05,
    "epoch": 3.2952646239554317,
    "step": 593
  },
  {
    "loss": 0.7126,
    "grad_norm": 10.658570289611816,
    "learning_rate": 1.3896930815127206e-05,
    "epoch": 3.3008356545961,
    "step": 594
  },
  {
    "loss": 0.8014,
    "grad_norm": 8.675304412841797,
    "learning_rate": 1.3883783293069372e-05,
    "epoch": 3.3064066852367686,
    "step": 595
  },
  {
    "loss": 0.7078,
    "grad_norm": 6.470170974731445,
    "learning_rate": 1.3870635771011543e-05,
    "epoch": 3.3119777158774375,
    "step": 596
  },
  {
    "loss": 0.6612,
    "grad_norm": 7.141599178314209,
    "learning_rate": 1.3857488248953713e-05,
    "epoch": 3.317548746518106,
    "step": 597
  },
  {
    "loss": 0.8968,
    "grad_norm": 9.977639198303223,
    "learning_rate": 1.384434072689588e-05,
    "epoch": 3.3231197771587744,
    "step": 598
  },
  {
    "loss": 0.8395,
    "grad_norm": 10.208252906799316,
    "learning_rate": 1.383119320483805e-05,
    "epoch": 3.328690807799443,
    "step": 599
  },
  {
    "loss": 0.9248,
    "grad_norm": 9.933085441589355,
    "learning_rate": 1.381804568278022e-05,
    "epoch": 3.3342618384401113,
    "step": 600
  },
  {
    "eval_loss": 2.9689695835113525,
    "eval_runtime": 35.9526,
    "eval_samples_per_second": 39.941,
    "eval_steps_per_second": 2.003,
    "epoch": 3.3342618384401113,
    "step": 600
  },
  {
    "train_runtime": 2220.807,
    "train_samples_per_second": 5.944,
    "train_steps_per_second": 0.743,
    "total_flos": 4.35104765343744e+16,
    "train_loss": 1.654274252106746,
    "epoch": 3.3342618384401113,
    "step": 600
  }
]