{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 5,
  "global_step": 72,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 5.283575057983398,
      "learning_rate": 0.0,
      "loss": 2.367,
      "step": 1
    },
    {
      "epoch": 0.05555555555555555,
      "eval_loss": 3.829460859298706,
      "eval_runtime": 4.378,
      "eval_samples_per_second": 51.394,
      "eval_steps_per_second": 17.131,
      "step": 1
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 4.849005699157715,
      "learning_rate": 2e-05,
      "loss": 2.0856,
      "step": 2
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 4.854206562042236,
      "learning_rate": 4e-05,
      "loss": 2.0361,
      "step": 3
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 3.646277666091919,
      "learning_rate": 6e-05,
      "loss": 1.9353,
      "step": 4
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 2.636997699737549,
      "learning_rate": 8e-05,
      "loss": 1.5523,
      "step": 5
    },
    {
      "epoch": 0.2777777777777778,
      "eval_loss": 2.0394411087036133,
      "eval_runtime": 4.4306,
      "eval_samples_per_second": 50.784,
      "eval_steps_per_second": 16.928,
      "step": 5
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 2.822960376739502,
      "learning_rate": 0.0001,
      "loss": 1.1471,
      "step": 6
    },
    {
      "epoch": 0.3888888888888889,
      "grad_norm": 3.2218048572540283,
      "learning_rate": 0.00012,
      "loss": 0.7826,
      "step": 7
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 2.1202049255371094,
      "learning_rate": 0.00014,
      "loss": 0.4402,
      "step": 8
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.287557601928711,
      "learning_rate": 0.00016,
      "loss": 0.4659,
      "step": 9
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 1.2889245748519897,
      "learning_rate": 0.00018,
      "loss": 0.3407,
      "step": 10
    },
    {
      "epoch": 0.5555555555555556,
      "eval_loss": 1.2009233236312866,
      "eval_runtime": 4.4308,
      "eval_samples_per_second": 50.781,
      "eval_steps_per_second": 16.927,
      "step": 10
    },
    {
      "epoch": 0.6111111111111112,
      "grad_norm": 0.6914752721786499,
      "learning_rate": 0.0002,
      "loss": 0.2716,
      "step": 11
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.010902762413025,
      "learning_rate": 0.00019987165071710527,
      "loss": 0.2239,
      "step": 12
    },
    {
      "epoch": 0.7222222222222222,
      "grad_norm": 0.5965589284896851,
      "learning_rate": 0.00019948693233918952,
      "loss": 0.2037,
      "step": 13
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.7514087557792664,
      "learning_rate": 0.00019884683243281116,
      "loss": 0.241,
      "step": 14
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.3119024634361267,
      "learning_rate": 0.00019795299412524945,
      "loss": 0.1686,
      "step": 15
    },
    {
      "epoch": 0.8333333333333334,
      "eval_loss": 1.0128521919250488,
      "eval_runtime": 4.4341,
      "eval_samples_per_second": 50.744,
      "eval_steps_per_second": 16.915,
      "step": 15
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.727888286113739,
      "learning_rate": 0.00019680771188662044,
      "loss": 0.1748,
      "step": 16
    },
    {
      "epoch": 0.9444444444444444,
      "grad_norm": 0.32004445791244507,
      "learning_rate": 0.00019541392564000488,
      "loss": 0.1383,
      "step": 17
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3259457051753998,
      "learning_rate": 0.00019377521321470805,
      "loss": 0.1288,
      "step": 18
    },
    {
      "epoch": 1.0555555555555556,
      "grad_norm": 0.3080686926841736,
      "learning_rate": 0.00019189578116202307,
      "loss": 0.1145,
      "step": 19
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.4244043529033661,
      "learning_rate": 0.00018978045395707418,
      "loss": 0.1512,
      "step": 20
    },
    {
      "epoch": 1.1111111111111112,
      "eval_loss": 1.1556396484375,
      "eval_runtime": 4.4392,
      "eval_samples_per_second": 50.685,
      "eval_steps_per_second": 16.895,
      "step": 20
    },
    {
      "epoch": 1.1666666666666667,
      "grad_norm": 0.3573606610298157,
      "learning_rate": 0.00018743466161445823,
      "loss": 0.1075,
      "step": 21
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 0.8563918471336365,
      "learning_rate": 0.00018486442574947511,
      "loss": 0.1301,
      "step": 22
    },
    {
      "epoch": 1.2777777777777777,
      "grad_norm": 0.387724369764328,
      "learning_rate": 0.00018207634412072764,
      "loss": 0.1264,
      "step": 23
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.4080502390861511,
      "learning_rate": 0.00017907757369376985,
      "loss": 0.1209,
      "step": 24
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 0.3518580198287964,
      "learning_rate": 0.0001758758122692791,
      "loss": 0.1278,
      "step": 25
    },
    {
      "epoch": 1.3888888888888888,
      "eval_loss": 0.7397511005401611,
      "eval_runtime": 4.4247,
      "eval_samples_per_second": 50.851,
      "eval_steps_per_second": 16.95,
      "step": 25
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 0.1988414078950882,
      "learning_rate": 0.000172479278722912,
      "loss": 0.0998,
      "step": 26
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.33381035923957825,
      "learning_rate": 0.00016889669190756868,
      "loss": 0.128,
      "step": 27
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 0.21273471415042877,
      "learning_rate": 0.00016513724827222227,
      "loss": 0.0824,
      "step": 28
    },
    {
      "epoch": 1.6111111111111112,
      "grad_norm": 0.26343834400177,
      "learning_rate": 0.0001612105982547663,
      "loss": 0.1177,
      "step": 29
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.1714707314968109,
      "learning_rate": 0.00015712682150947923,
      "loss": 0.0805,
      "step": 30
    },
    {
      "epoch": 1.6666666666666665,
      "eval_loss": 0.6418749690055847,
      "eval_runtime": 4.4192,
      "eval_samples_per_second": 50.914,
      "eval_steps_per_second": 16.971,
      "step": 30
    },
    {
      "epoch": 1.7222222222222223,
      "grad_norm": 0.22125771641731262,
      "learning_rate": 0.00015289640103269625,
      "loss": 0.0894,
      "step": 31
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.24318796396255493,
      "learning_rate": 0.00014853019625310813,
      "loss": 0.0872,
      "step": 32
    },
    {
      "epoch": 1.8333333333333335,
      "grad_norm": 0.2300858497619629,
      "learning_rate": 0.00014403941515576344,
      "loss": 0.0632,
      "step": 33
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 0.20156258344650269,
      "learning_rate": 0.00013943558551133186,
      "loss": 0.0367,
      "step": 34
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.28669273853302,
      "learning_rate": 0.00013473052528448201,
      "loss": 0.0691,
      "step": 35
    },
    {
      "epoch": 1.9444444444444444,
      "eval_loss": 0.6279710531234741,
      "eval_runtime": 4.428,
      "eval_samples_per_second": 50.813,
      "eval_steps_per_second": 16.938,
      "step": 35
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2706594467163086,
      "learning_rate": 0.00012993631229733582,
      "loss": 0.0698,
      "step": 36
    },
    {
      "epoch": 2.0555555555555554,
      "grad_norm": 0.1888062208890915,
      "learning_rate": 0.00012506525322587207,
      "loss": 0.0479,
      "step": 37
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 0.1502252072095871,
      "learning_rate": 0.00012012985200886602,
      "loss": 0.0446,
      "step": 38
    },
    {
      "epoch": 2.1666666666666665,
      "grad_norm": 0.1221298798918724,
      "learning_rate": 0.00011514277775045768,
      "loss": 0.0359,
      "step": 39
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.1640322506427765,
      "learning_rate": 0.00011011683219874323,
      "loss": 0.0532,
      "step": 40
    },
    {
      "epoch": 2.2222222222222223,
      "eval_loss": 0.6401044726371765,
      "eval_runtime": 4.4435,
      "eval_samples_per_second": 50.636,
      "eval_steps_per_second": 16.879,
      "step": 40
    },
    {
      "epoch": 2.2777777777777777,
      "grad_norm": 0.14193332195281982,
      "learning_rate": 0.00010506491688387127,
      "loss": 0.057,
      "step": 41
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.2185075283050537,
      "learning_rate": 0.0001,
      "loss": 0.0665,
      "step": 42
    },
    {
      "epoch": 2.388888888888889,
      "grad_norm": 0.10428009927272797,
      "learning_rate": 9.493508311612874e-05,
      "loss": 0.0227,
      "step": 43
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 0.13296478986740112,
      "learning_rate": 8.98831678012568e-05,
      "loss": 0.0446,
      "step": 44
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.2428351491689682,
      "learning_rate": 8.485722224954237e-05,
      "loss": 0.0401,
      "step": 45
    },
    {
      "epoch": 2.5,
      "eval_loss": 0.6121364235877991,
      "eval_runtime": 4.4316,
      "eval_samples_per_second": 50.772,
      "eval_steps_per_second": 16.924,
      "step": 45
    },
    {
      "epoch": 2.5555555555555554,
      "grad_norm": 0.11428028345108032,
      "learning_rate": 7.987014799113397e-05,
      "loss": 0.0571,
      "step": 46
    },
    {
      "epoch": 2.611111111111111,
      "grad_norm": 0.16103307902812958,
      "learning_rate": 7.493474677412794e-05,
      "loss": 0.06,
      "step": 47
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.15212145447731018,
      "learning_rate": 7.006368770266421e-05,
      "loss": 0.0565,
      "step": 48
    },
    {
      "epoch": 2.7222222222222223,
      "grad_norm": 0.13669654726982117,
      "learning_rate": 6.526947471551798e-05,
      "loss": 0.0528,
      "step": 49
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.14751426875591278,
      "learning_rate": 6.0564414488668165e-05,
      "loss": 0.0367,
      "step": 50
    },
    {
      "epoch": 2.7777777777777777,
      "eval_loss": 0.6128059029579163,
      "eval_runtime": 4.4242,
      "eval_samples_per_second": 50.857,
      "eval_steps_per_second": 16.952,
      "step": 50
    },
    {
      "epoch": 2.8333333333333335,
      "grad_norm": 0.1321621984243393,
      "learning_rate": 5.596058484423656e-05,
      "loss": 0.0515,
      "step": 51
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 0.12428299337625504,
      "learning_rate": 5.146980374689192e-05,
      "loss": 0.0323,
      "step": 52
    },
    {
      "epoch": 2.9444444444444446,
      "grad_norm": 0.12496347725391388,
      "learning_rate": 4.710359896730379e-05,
      "loss": 0.0477,
      "step": 53
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.13071990013122559,
      "learning_rate": 4.287317849052075e-05,
      "loss": 0.0476,
      "step": 54
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 0.14479975402355194,
      "learning_rate": 3.878940174523371e-05,
      "loss": 0.064,
      "step": 55
    },
    {
      "epoch": 3.0555555555555554,
      "eval_loss": 0.6650803685188293,
      "eval_runtime": 4.4195,
      "eval_samples_per_second": 50.911,
      "eval_steps_per_second": 16.97,
      "step": 55
    },
    {
      "epoch": 3.111111111111111,
      "grad_norm": 0.10720603168010712,
      "learning_rate": 3.4862751727777797e-05,
      "loss": 0.0342,
      "step": 56
    },
    {
      "epoch": 3.1666666666666665,
      "grad_norm": 0.11034850031137466,
      "learning_rate": 3.110330809243134e-05,
      "loss": 0.0427,
      "step": 57
    },
    {
      "epoch": 3.2222222222222223,
      "grad_norm": 0.10966861248016357,
      "learning_rate": 2.7520721277088024e-05,
      "loss": 0.0391,
      "step": 58
    },
    {
      "epoch": 3.2777777777777777,
      "grad_norm": 0.15885663032531738,
      "learning_rate": 2.4124187730720917e-05,
      "loss": 0.0418,
      "step": 59
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.11797765642404556,
      "learning_rate": 2.092242630623016e-05,
      "loss": 0.0315,
      "step": 60
    },
    {
      "epoch": 3.3333333333333335,
      "eval_loss": 0.67604660987854,
      "eval_runtime": 4.4237,
      "eval_samples_per_second": 50.863,
      "eval_steps_per_second": 16.954,
      "step": 60
    },
    {
      "epoch": 3.388888888888889,
      "grad_norm": 0.0905529335141182,
      "learning_rate": 1.7923655879272393e-05,
      "loss": 0.0273,
      "step": 61
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 0.16150006651878357,
      "learning_rate": 1.5135574250524897e-05,
      "loss": 0.0388,
      "step": 62
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.11532038450241089,
      "learning_rate": 1.2565338385541792e-05,
      "loss": 0.0448,
      "step": 63
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 0.11461252719163895,
      "learning_rate": 1.0219546042925843e-05,
      "loss": 0.0355,
      "step": 64
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 0.15628494322299957,
      "learning_rate": 8.10421883797694e-06,
      "loss": 0.0526,
      "step": 65
    },
    {
      "epoch": 3.611111111111111,
      "eval_loss": 0.6782650947570801,
      "eval_runtime": 4.4202,
      "eval_samples_per_second": 50.903,
      "eval_steps_per_second": 16.968,
      "step": 65
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.11209984123706818,
      "learning_rate": 6.22478678529197e-06,
      "loss": 0.0396,
      "step": 66
    },
    {
      "epoch": 3.7222222222222223,
      "grad_norm": 0.09433431923389435,
      "learning_rate": 4.586074359995119e-06,
      "loss": 0.0363,
      "step": 67
    },
    {
      "epoch": 3.7777777777777777,
      "grad_norm": 0.17248067259788513,
      "learning_rate": 3.1922881133795825e-06,
      "loss": 0.0469,
      "step": 68
    },
    {
      "epoch": 3.8333333333333335,
      "grad_norm": 0.1422806680202484,
      "learning_rate": 2.0470058747505516e-06,
      "loss": 0.0437,
      "step": 69
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.13716408610343933,
      "learning_rate": 1.1531675671888619e-06,
      "loss": 0.0412,
      "step": 70
    },
    {
      "epoch": 3.888888888888889,
      "eval_loss": 0.6792881488800049,
      "eval_runtime": 4.4297,
      "eval_samples_per_second": 50.794,
      "eval_steps_per_second": 16.931,
      "step": 70
    },
    {
      "epoch": 3.9444444444444446,
      "grad_norm": 0.14409439265727997,
      "learning_rate": 5.130676608104845e-07,
      "loss": 0.0449,
      "step": 71
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.13294203579425812,
      "learning_rate": 1.2834928289472416e-07,
      "loss": 0.0353,
      "step": 72
    }
  ],
  "logging_steps": 1,
  "max_steps": 72,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 18,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.533795369340109e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}