{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 936,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "learning_rate": 4.9985919525527434e-05,
      "loss": 1.0499,
      "step": 10
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.994369396289063e-05,
      "loss": 0.7822,
      "step": 20
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987337087656614e-05,
      "loss": 0.575,
      "step": 30
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.977502948114772e-05,
      "loss": 0.5199,
      "step": 40
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.964878055211597e-05,
      "loss": 0.5358,
      "step": 50
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.949476630105669e-05,
      "loss": 0.508,
      "step": 60
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.9313160215468334e-05,
      "loss": 0.4915,
      "step": 70
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.910416686333906e-05,
      "loss": 0.4856,
      "step": 80
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.886802166271364e-05,
      "loss": 0.479,
      "step": 90
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.8604990616509616e-05,
      "loss": 0.4707,
      "step": 100
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.8315370012881514e-05,
      "loss": 0.4687,
      "step": 110
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.799948609147061e-05,
      "loss": 0.4834,
      "step": 120
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.765769467591625e-05,
      "loss": 0.4541,
      "step": 130
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.7290380773042575e-05,
      "loss": 0.4442,
      "step": 140
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.68979581391722e-05,
      "loss": 0.4528,
      "step": 150
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.6480868814055424e-05,
      "loss": 0.436,
      "step": 160
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.6039582622939854e-05,
      "loss": 0.4504,
      "step": 170
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.557459664734141e-05,
      "loss": 0.4348,
      "step": 180
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.5086434665112864e-05,
      "loss": 0.4283,
      "step": 190
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.457564656044056e-05,
      "loss": 0.4527,
      "step": 200
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.404280770443398e-05,
      "loss": 0.4525,
      "step": 210
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.348851830700593e-05,
      "loss": 0.4315,
      "step": 220
    },
    {
      "epoch": 0.98,
      "learning_rate": 4.2913402740773294e-05,
      "loss": 0.4309,
      "step": 230
    },
    {
      "epoch": 1.03,
      "learning_rate": 4.231810883773999e-05,
      "loss": 0.4207,
      "step": 240
    },
    {
      "epoch": 1.07,
      "learning_rate": 4.170330715955444e-05,
      "loss": 0.4189,
      "step": 250
    },
    {
      "epoch": 1.11,
      "learning_rate": 4.1069690242163484e-05,
      "loss": 0.4353,
      "step": 260
    },
    {
      "epoch": 1.15,
      "learning_rate": 4.0417971815713584e-05,
      "loss": 0.4054,
      "step": 270
    },
    {
      "epoch": 1.2,
      "learning_rate": 3.974888600057808e-05,
      "loss": 0.4152,
      "step": 280
    },
    {
      "epoch": 1.24,
      "learning_rate": 3.906318648041617e-05,
      "loss": 0.3879,
      "step": 290
    },
    {
      "epoch": 1.28,
      "learning_rate": 3.8361645653195026e-05,
      "loss": 0.4164,
      "step": 300
    },
    {
      "epoch": 1.32,
      "learning_rate": 3.764505376113138e-05,
      "loss": 0.406,
      "step": 310
    },
    {
      "epoch": 1.37,
      "learning_rate": 3.69142180005327e-05,
      "loss": 0.4076,
      "step": 320
    },
    {
      "epoch": 1.41,
      "learning_rate": 3.6169961612540645e-05,
      "loss": 0.3953,
      "step": 330
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.5413122955801005e-05,
      "loss": 0.4105,
      "step": 340
    },
    {
      "epoch": 1.5,
      "learning_rate": 3.4644554562104634e-05,
      "loss": 0.4014,
      "step": 350
    },
    {
      "epoch": 1.54,
      "learning_rate": 3.386512217606339e-05,
      "loss": 0.4234,
      "step": 360
    },
    {
      "epoch": 1.58,
      "learning_rate": 3.307570377990245e-05,
      "loss": 0.4012,
      "step": 370
    },
    {
      "epoch": 1.62,
      "learning_rate": 3.227718860446782e-05,
      "loss": 0.3979,
      "step": 380
    },
    {
      "epoch": 1.67,
      "learning_rate": 3.147047612756302e-05,
      "loss": 0.3877,
      "step": 390
    },
    {
      "epoch": 1.71,
      "learning_rate": 3.065647506074306e-05,
      "loss": 0.4024,
      "step": 400
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.983610232570728e-05,
      "loss": 0.3994,
      "step": 410
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.9010282021444008e-05,
      "loss": 0.4052,
      "step": 420
    },
    {
      "epoch": 1.84,
      "learning_rate": 2.8179944383290274e-05,
      "loss": 0.4038,
      "step": 430
    },
    {
      "epoch": 1.88,
      "learning_rate": 2.7346024735079486e-05,
      "loss": 0.3999,
      "step": 440
    },
    {
      "epoch": 1.92,
      "learning_rate": 2.6509462435557152e-05,
      "loss": 0.3848,
      "step": 450
    },
    {
      "epoch": 1.97,
      "learning_rate": 2.5671199820251534e-05,
      "loss": 0.4126,
      "step": 460
    },
    {
      "epoch": 2.01,
      "learning_rate": 2.48321811399911e-05,
      "loss": 0.3996,
      "step": 470
    },
    {
      "epoch": 2.05,
      "learning_rate": 2.399335149726463e-05,
      "loss": 0.3672,
      "step": 480
    },
    {
      "epoch": 2.09,
      "learning_rate": 2.3155655781621793e-05,
      "loss": 0.3728,
      "step": 490
    },
    {
      "epoch": 2.14,
      "learning_rate": 2.2320037605313808e-05,
      "loss": 0.3901,
      "step": 500
    },
    {
      "epoch": 2.18,
      "learning_rate": 2.148743824037269e-05,
      "loss": 0.3848,
      "step": 510
    },
    {
      "epoch": 2.22,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 0.3729,
      "step": 520
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.98350429737465e-05,
      "loss": 0.3709,
      "step": 530
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.9017108392811065e-05,
      "loss": 0.3811,
      "step": 540
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.820591316807939e-05,
      "loss": 0.3639,
      "step": 550
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.740237106064383e-05,
      "loss": 0.3693,
      "step": 560
    },
    {
      "epoch": 2.44,
      "learning_rate": 1.6607387210834887e-05,
      "loss": 0.3769,
      "step": 570
    },
    {
      "epoch": 2.48,
      "learning_rate": 1.582185711863681e-05,
      "loss": 0.3621,
      "step": 580
    },
    {
      "epoch": 2.52,
      "learning_rate": 1.5046665634962476e-05,
      "loss": 0.3585,
      "step": 590
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.4282685964923642e-05,
      "loss": 0.3705,
      "step": 600
    },
    {
      "epoch": 2.61,
      "learning_rate": 1.3530778684219648e-05,
      "loss": 0.3662,
      "step": 610
    },
    {
      "epoch": 2.65,
      "learning_rate": 1.2791790769752232e-05,
      "loss": 0.3831,
      "step": 620
    },
    {
      "epoch": 2.69,
      "learning_rate": 1.2066554645558578e-05,
      "loss": 0.3874,
      "step": 630
    },
    {
      "epoch": 2.74,
      "learning_rate": 1.1355887245137383e-05,
      "loss": 0.3563,
      "step": 640
    },
    {
      "epoch": 2.78,
      "learning_rate": 1.0660589091223855e-05,
      "loss": 0.3612,
      "step": 650
    },
    {
      "epoch": 2.82,
      "learning_rate": 9.981443394050525e-06,
      "loss": 0.3721,
      "step": 660
    },
    {
      "epoch": 2.86,
      "learning_rate": 9.3192151691096e-06,
      "loss": 0.3744,
      "step": 670
    },
    {
      "epoch": 2.91,
      "learning_rate": 8.67465037541038e-06,
      "loss": 0.3553,
      "step": 680
    },
    {
      "epoch": 2.95,
      "learning_rate": 8.048475075202727e-06,
      "loss": 0.3685,
      "step": 690
    },
    {
      "epoch": 2.99,
      "learning_rate": 7.441394616113062e-06,
      "loss": 0.3731,
      "step": 700
    },
    {
      "epoch": 3.03,
      "learning_rate": 6.854092836613948e-06,
      "loss": 0.3529,
      "step": 710
    },
    {
      "epoch": 3.08,
      "learning_rate": 6.28723129572247e-06,
      "loss": 0.3542,
      "step": 720
    },
    {
      "epoch": 3.12,
      "learning_rate": 5.741448527795137e-06,
      "loss": 0.3587,
      "step": 730
    },
    {
      "epoch": 3.16,
      "learning_rate": 5.217359323258459e-06,
      "loss": 0.3583,
      "step": 740
    },
    {
      "epoch": 3.21,
      "learning_rate": 4.715554036085673e-06,
      "loss": 0.3589,
      "step": 750
    },
    {
      "epoch": 3.25,
      "learning_rate": 4.236597918799709e-06,
      "loss": 0.3424,
      "step": 760
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.7810304857511914e-06,
      "loss": 0.3455,
      "step": 770
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.3493649053890326e-06,
      "loss": 0.3667,
      "step": 780
    },
    {
      "epoch": 3.38,
      "learning_rate": 2.942087422208051e-06,
      "loss": 0.3579,
      "step": 790
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.5596568090246548e-06,
      "loss": 0.342,
      "step": 800
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.2025038501977486e-06,
      "loss": 0.3542,
      "step": 810
    },
    {
      "epoch": 3.5,
      "learning_rate": 1.8710308563769124e-06,
      "loss": 0.3385,
      "step": 820
    },
    {
      "epoch": 3.55,
      "learning_rate": 1.5656112113243721e-06,
      "loss": 0.3564,
      "step": 830
    },
    {
      "epoch": 3.59,
      "learning_rate": 1.286588951321363e-06,
      "loss": 0.357,
      "step": 840
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.034278377632636e-06,
      "loss": 0.3492,
      "step": 850
    },
    {
      "epoch": 3.68,
      "learning_rate": 8.089637024655483e-07,
      "loss": 0.3547,
      "step": 860
    },
    {
      "epoch": 3.72,
      "learning_rate": 6.108987288226536e-07,
      "loss": 0.3401,
      "step": 870
    },
    {
      "epoch": 3.76,
      "learning_rate": 4.403065646083809e-07,
      "loss": 0.3417,
      "step": 880
    },
    {
      "epoch": 3.8,
      "learning_rate": 2.973793713118039e-07,
      "loss": 0.3709,
      "step": 890
    },
    {
      "epoch": 3.85,
      "learning_rate": 1.8227814754865068e-07,
      "loss": 0.3702,
      "step": 900
    },
    {
      "epoch": 3.89,
      "learning_rate": 9.513254770636137e-08,
      "loss": 0.3401,
      "step": 910
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.604073589645596e-08,
      "loss": 0.3566,
      "step": 920
    },
    {
      "epoch": 3.97,
      "learning_rate": 5.069275378746796e-09,
      "loss": 0.3785,
      "step": 930
    },
    {
      "epoch": 4.0,
      "step": 936,
      "total_flos": 4.98594351337046e+17,
      "train_loss": 0.40969020433914966,
      "train_runtime": 4321.1174,
      "train_samples_per_second": 6.918,
      "train_steps_per_second": 0.217
    }
  ],
  "logging_steps": 10,
  "max_steps": 936,
  "num_train_epochs": 4,
  "save_steps": 200,
  "total_flos": 4.98594351337046e+17,
  "trial_name": null,
  "trial_params": null
}