{
  "best_global_step": 1800,
  "best_metric": 2.1885855197906494,
  "best_model_checkpoint": "./outputs/checkpoint-1800",
  "epoch": 2.6058631921824107,
  "eval_steps": 200,
  "global_step": 2400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05428881650380022,
      "grad_norm": 0.3356933295726776,
      "learning_rate": 0.000196,
      "loss": 2.9766,
      "step": 50
    },
    {
      "epoch": 0.10857763300760044,
      "grad_norm": 0.5300453305244446,
      "learning_rate": 0.00019983906674307943,
      "loss": 2.4625,
      "step": 100
    },
    {
      "epoch": 0.16286644951140064,
      "grad_norm": 0.4830726385116577,
      "learning_rate": 0.00019934360560412188,
      "loss": 2.3984,
      "step": 150
    },
    {
      "epoch": 0.21715526601520088,
      "grad_norm": 0.6850009560585022,
      "learning_rate": 0.000198515209997625,
      "loss": 2.3727,
      "step": 200
    },
    {
      "epoch": 0.21715526601520088,
      "eval_loss": 2.348525047302246,
      "eval_runtime": 9.0868,
      "eval_samples_per_second": 66.25,
      "eval_steps_per_second": 2.091,
      "step": 200
    },
    {
      "epoch": 0.2714440825190011,
      "grad_norm": 0.6019847393035889,
      "learning_rate": 0.00019735665616115814,
      "loss": 2.3382,
      "step": 250
    },
    {
      "epoch": 0.3257328990228013,
      "grad_norm": 0.7136948704719543,
      "learning_rate": 0.00019587182680566505,
      "loss": 2.3352,
      "step": 300
    },
    {
      "epoch": 0.3800217155266015,
      "grad_norm": 0.6682165861129761,
      "learning_rate": 0.00019406569810316855,
      "loss": 2.3282,
      "step": 350
    },
    {
      "epoch": 0.43431053203040176,
      "grad_norm": 0.6841819882392883,
      "learning_rate": 0.00019194432300991294,
      "loss": 2.3176,
      "step": 400
    },
    {
      "epoch": 0.43431053203040176,
      "eval_loss": 2.2951860427856445,
      "eval_runtime": 7.346,
      "eval_samples_per_second": 81.949,
      "eval_steps_per_second": 2.586,
      "step": 400
    },
    {
      "epoch": 0.48859934853420195,
      "grad_norm": 0.6011854410171509,
      "learning_rate": 0.00018951481098083322,
      "loss": 2.2901,
      "step": 450
    },
    {
      "epoch": 0.5428881650380022,
      "grad_norm": 0.6126257181167603,
      "learning_rate": 0.00018678530414333524,
      "loss": 2.3023,
      "step": 500
    },
    {
      "epoch": 0.5971769815418024,
      "grad_norm": 0.6901931762695312,
      "learning_rate": 0.00018376495001023636,
      "loss": 2.2943,
      "step": 550
    },
    {
      "epoch": 0.6514657980456026,
      "grad_norm": 0.6403626203536987,
      "learning_rate": 0.00018046387082331516,
      "loss": 2.2696,
      "step": 600
    },
    {
      "epoch": 0.6514657980456026,
      "eval_loss": 2.2626566886901855,
      "eval_runtime": 7.335,
      "eval_samples_per_second": 82.072,
      "eval_steps_per_second": 2.59,
      "step": 600
    },
    {
      "epoch": 0.7057546145494028,
      "grad_norm": 0.6320544481277466,
      "learning_rate": 0.00017689312963021044,
      "loss": 2.2678,
      "step": 650
    },
    {
      "epoch": 0.760043431053203,
      "grad_norm": 0.7033542990684509,
      "learning_rate": 0.00017306469320835707,
      "loss": 2.2667,
      "step": 700
    },
    {
      "epoch": 0.8143322475570033,
      "grad_norm": 0.6568920612335205,
      "learning_rate": 0.00016899139196021397,
      "loss": 2.2727,
      "step": 750
    },
    {
      "epoch": 0.8686210640608035,
      "grad_norm": 0.7634288668632507,
      "learning_rate": 0.00016468687691418833,
      "loss": 2.2482,
      "step": 800
    },
    {
      "epoch": 0.8686210640608035,
      "eval_loss": 2.2396719455718994,
      "eval_runtime": 7.3506,
      "eval_samples_per_second": 81.898,
      "eval_steps_per_second": 2.585,
      "step": 800
    },
    {
      "epoch": 0.9229098805646037,
      "grad_norm": 0.6467435359954834,
      "learning_rate": 0.0001601655739753613,
      "loss": 2.2523,
      "step": 850
    },
    {
      "epoch": 0.9771986970684039,
      "grad_norm": 0.6962546706199646,
      "learning_rate": 0.0001554426355793363,
      "loss": 2.2509,
      "step": 900
    },
    {
      "epoch": 1.0314875135722041,
      "grad_norm": 0.6648530960083008,
      "learning_rate": 0.0001505338899112344,
      "loss": 2.2103,
      "step": 950
    },
    {
      "epoch": 1.0857763300760044,
      "grad_norm": 0.9022650122642517,
      "learning_rate": 0.00014545578786002178,
      "loss": 2.1862,
      "step": 1000
    },
    {
      "epoch": 1.0857763300760044,
      "eval_loss": 2.22625994682312,
      "eval_runtime": 7.3813,
      "eval_samples_per_second": 81.558,
      "eval_steps_per_second": 2.574,
      "step": 1000
    },
    {
      "epoch": 1.1400651465798046,
      "grad_norm": 0.788972795009613,
      "learning_rate": 0.00014022534788594312,
      "loss": 2.2099,
      "step": 1050
    },
    {
      "epoch": 1.1943539630836049,
      "grad_norm": 0.7912983894348145,
      "learning_rate": 0.00013486009898582889,
      "loss": 2.1799,
      "step": 1100
    },
    {
      "epoch": 1.248642779587405,
      "grad_norm": 0.766243040561676,
      "learning_rate": 0.00012937802194741994,
      "loss": 2.2058,
      "step": 1150
    },
    {
      "epoch": 1.3029315960912053,
      "grad_norm": 0.8190643191337585,
      "learning_rate": 0.00012379748908958562,
      "loss": 2.2021,
      "step": 1200
    },
    {
      "epoch": 1.3029315960912053,
      "eval_loss": 2.21370530128479,
      "eval_runtime": 7.3565,
      "eval_samples_per_second": 81.833,
      "eval_steps_per_second": 2.583,
      "step": 1200
    },
    {
      "epoch": 1.3572204125950054,
      "grad_norm": 0.824234664440155,
      "learning_rate": 0.00011813720269038712,
      "loss": 2.1931,
      "step": 1250
    },
    {
      "epoch": 1.4115092290988056,
      "grad_norm": 0.8461474776268005,
      "learning_rate": 0.00011241613230933439,
      "loss": 2.1755,
      "step": 1300
    },
    {
      "epoch": 1.4657980456026058,
      "grad_norm": 0.8048629760742188,
      "learning_rate": 0.00010665345121389123,
      "loss": 2.186,
      "step": 1350
    },
    {
      "epoch": 1.520086862106406,
      "grad_norm": 0.7822412848472595,
      "learning_rate": 0.00010086847212328548,
      "loss": 2.1949,
      "step": 1400
    },
    {
      "epoch": 1.520086862106406,
      "eval_loss": 2.2031383514404297,
      "eval_runtime": 7.3421,
      "eval_samples_per_second": 81.993,
      "eval_steps_per_second": 2.588,
      "step": 1400
    },
    {
      "epoch": 1.5743756786102063,
      "grad_norm": 0.8551859855651855,
      "learning_rate": 9.508058248496825e-05,
      "loss": 2.193,
      "step": 1450
    },
    {
      "epoch": 1.6286644951140063,
      "grad_norm": 0.8906627893447876,
      "learning_rate": 8.930917950063358e-05,
      "loss": 2.1834,
      "step": 1500
    },
    {
      "epoch": 1.6829533116178066,
      "grad_norm": 0.7810714840888977,
      "learning_rate": 8.357360511954789e-05,
      "loss": 2.1665,
      "step": 1550
    },
    {
      "epoch": 1.7372421281216068,
      "grad_norm": 0.8593874573707581,
      "learning_rate": 7.789308121704968e-05,
      "loss": 2.1698,
      "step": 1600
    },
    {
      "epoch": 1.7372421281216068,
      "eval_loss": 2.1958162784576416,
      "eval_runtime": 7.3308,
      "eval_samples_per_second": 82.119,
      "eval_steps_per_second": 2.592,
      "step": 1600
    },
    {
      "epoch": 1.791530944625407,
      "grad_norm": 0.8802404403686523,
      "learning_rate": 7.228664517545785e-05,
      "loss": 2.1577,
      "step": 1650
    },
    {
      "epoch": 1.8458197611292073,
      "grad_norm": 0.8591893911361694,
      "learning_rate": 6.677308608327948e-05,
      "loss": 2.161,
      "step": 1700
    },
    {
      "epoch": 1.9001085776330076,
      "grad_norm": 0.8191171884536743,
      "learning_rate": 6.137088176653491e-05,
      "loss": 2.178,
      "step": 1750
    },
    {
      "epoch": 1.9543973941368078,
      "grad_norm": 0.9448741674423218,
      "learning_rate": 5.609813686323086e-05,
      "loss": 2.1746,
      "step": 1800
    },
    {
      "epoch": 1.9543973941368078,
      "eval_loss": 2.1885855197906494,
      "eval_runtime": 7.3549,
      "eval_samples_per_second": 81.85,
      "eval_steps_per_second": 2.583,
      "step": 1800
    },
    {
      "epoch": 2.008686210640608,
      "grad_norm": 0.9358265399932861,
      "learning_rate": 5.097252214851365e-05,
      "loss": 2.165,
      "step": 1850
    },
    {
      "epoch": 2.0629750271444083,
      "grad_norm": 1.0949631929397583,
      "learning_rate": 4.601121531384579e-05,
      "loss": 2.1348,
      "step": 1900
    },
    {
      "epoch": 2.1172638436482085,
      "grad_norm": 1.3909953832626343,
      "learning_rate": 4.1230843398675555e-05,
      "loss": 2.1399,
      "step": 1950
    },
    {
      "epoch": 2.1715526601520088,
      "grad_norm": 1.2448049783706665,
      "learning_rate": 3.664742706752925e-05,
      "loss": 2.1417,
      "step": 2000
    },
    {
      "epoch": 2.1715526601520088,
      "eval_loss": 2.194857120513916,
      "eval_runtime": 7.3786,
      "eval_samples_per_second": 81.587,
      "eval_steps_per_second": 2.575,
      "step": 2000
    },
    {
      "epoch": 2.225841476655809,
      "grad_norm": 1.2955018281936646,
      "learning_rate": 3.227632691927414e-05,
      "loss": 2.1308,
      "step": 2050
    },
    {
      "epoch": 2.2801302931596092,
      "grad_norm": 1.0340995788574219,
      "learning_rate": 2.8132192008487768e-05,
      "loss": 2.129,
      "step": 2100
    },
    {
      "epoch": 2.3344191096634095,
      "grad_norm": 1.2585651874542236,
      "learning_rate": 2.4228910751455625e-05,
      "loss": 2.1403,
      "step": 2150
    },
    {
      "epoch": 2.3887079261672097,
      "grad_norm": 1.2523435354232788,
      "learning_rate": 2.0579564381328775e-05,
      "loss": 2.1197,
      "step": 2200
    },
    {
      "epoch": 2.3887079261672097,
      "eval_loss": 2.190831422805786,
      "eval_runtime": 7.3454,
      "eval_samples_per_second": 81.956,
      "eval_steps_per_second": 2.587,
      "step": 2200
    },
    {
      "epoch": 2.44299674267101,
      "grad_norm": 1.4523568153381348,
      "learning_rate": 1.7196383108428315e-05,
      "loss": 2.1359,
      "step": 2250
    },
    {
      "epoch": 2.49728555917481,
      "grad_norm": 1.3263523578643799,
      "learning_rate": 1.4090705132619631e-05,
      "loss": 2.1206,
      "step": 2300
    },
    {
      "epoch": 2.5515743756786105,
      "grad_norm": 1.4148005247116089,
      "learning_rate": 1.1272938645118803e-05,
      "loss": 2.1314,
      "step": 2350
    },
    {
      "epoch": 2.6058631921824107,
      "grad_norm": 1.9476280212402344,
      "learning_rate": 8.752526947077045e-06,
      "loss": 2.1254,
      "step": 2400
    },
    {
      "epoch": 2.6058631921824107,
      "eval_loss": 2.189617395401001,
      "eval_runtime": 7.3524,
      "eval_samples_per_second": 81.878,
      "eval_steps_per_second": 2.584,
      "step": 2400
    }
  ],
  "logging_steps": 50,
  "max_steps": 2763,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 3
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.723721683521372e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}