{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 2,
  "global_step": 56,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 268.4626770019531,
      "learning_rate": 0.0005621725453848409,
      "loss": 4.6194,
      "step": 2
    },
    {
      "epoch": 0.07142857142857142,
      "eval_loss": 4.2445478439331055,
      "eval_runtime": 108.1522,
      "eval_samples_per_second": 3.236,
      "eval_steps_per_second": 0.055,
      "step": 2
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 71.9693374633789,
      "learning_rate": 0.0005413513400002172,
      "loss": 6.5579,
      "step": 4
    },
    {
      "epoch": 0.14285714285714285,
      "eval_loss": 4.125729560852051,
      "eval_runtime": 146.7071,
      "eval_samples_per_second": 2.386,
      "eval_steps_per_second": 0.041,
      "step": 4
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 5.648405075073242,
      "learning_rate": 0.0005205301346155935,
      "loss": 4.9826,
      "step": 6
    },
    {
      "epoch": 0.21428571428571427,
      "eval_loss": 5.672520160675049,
      "eval_runtime": 142.0721,
      "eval_samples_per_second": 2.464,
      "eval_steps_per_second": 0.042,
      "step": 6
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 6.27989387512207,
      "learning_rate": 0.0004997089292309697,
      "loss": 5.9002,
      "step": 8
    },
    {
      "epoch": 0.2857142857142857,
      "eval_loss": 4.104394912719727,
      "eval_runtime": 135.6009,
      "eval_samples_per_second": 2.581,
      "eval_steps_per_second": 0.044,
      "step": 8
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.1307026743888855,
      "learning_rate": 0.000478887723846346,
      "loss": 4.8597,
      "step": 10
    },
    {
      "epoch": 0.35714285714285715,
      "eval_loss": 4.095502853393555,
      "eval_runtime": 111.3151,
      "eval_samples_per_second": 3.144,
      "eval_steps_per_second": 0.054,
      "step": 10
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.01885412447154522,
      "learning_rate": 0.00045806651846172226,
      "loss": 4.8531,
      "step": 12
    },
    {
      "epoch": 0.42857142857142855,
      "eval_loss": 4.094283580780029,
      "eval_runtime": 109.2586,
      "eval_samples_per_second": 3.203,
      "eval_steps_per_second": 0.055,
      "step": 12
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.010488088242709637,
      "learning_rate": 0.0004372453130770985,
      "loss": 4.8523,
      "step": 14
    },
    {
      "epoch": 0.5,
      "eval_loss": 4.094061851501465,
      "eval_runtime": 137.8714,
      "eval_samples_per_second": 2.539,
      "eval_steps_per_second": 0.044,
      "step": 14
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.006233076564967632,
      "learning_rate": 0.00041642410769247477,
      "loss": 4.8521,
      "step": 16
    },
    {
      "epoch": 0.5714285714285714,
      "eval_loss": 4.093998432159424,
      "eval_runtime": 129.4726,
      "eval_samples_per_second": 2.703,
      "eval_steps_per_second": 0.046,
      "step": 16
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 0.02763601951301098,
      "learning_rate": 0.00039560290230785105,
      "loss": 4.8521,
      "step": 18
    },
    {
      "epoch": 0.6428571428571429,
      "eval_loss": 4.09397554397583,
      "eval_runtime": 109.2369,
      "eval_samples_per_second": 3.204,
      "eval_steps_per_second": 0.055,
      "step": 18
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.005364938639104366,
      "learning_rate": 0.00037478169692322733,
      "loss": 4.8521,
      "step": 20
    },
    {
      "epoch": 0.7142857142857143,
      "eval_loss": 4.093966960906982,
      "eval_runtime": 129.3687,
      "eval_samples_per_second": 2.705,
      "eval_steps_per_second": 0.046,
      "step": 20
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 0.0038224116433411837,
      "learning_rate": 0.00035396049153860355,
      "loss": 4.8521,
      "step": 22
    },
    {
      "epoch": 0.7857142857142857,
      "eval_loss": 4.093982219696045,
      "eval_runtime": 110.6155,
      "eval_samples_per_second": 3.164,
      "eval_steps_per_second": 0.054,
      "step": 22
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.005476065911352634,
      "learning_rate": 0.0003331392861539798,
      "loss": 4.8521,
      "step": 24
    },
    {
      "epoch": 0.8571428571428571,
      "eval_loss": 4.093994140625,
      "eval_runtime": 108.8427,
      "eval_samples_per_second": 3.216,
      "eval_steps_per_second": 0.055,
      "step": 24
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 0.00341013353317976,
      "learning_rate": 0.00031231808076935606,
      "loss": 4.8521,
      "step": 26
    },
    {
      "epoch": 0.9285714285714286,
      "eval_loss": 4.093983173370361,
      "eval_runtime": 109.5893,
      "eval_samples_per_second": 3.194,
      "eval_steps_per_second": 0.055,
      "step": 26
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.004943197127431631,
      "learning_rate": 0.00029149687538473234,
      "loss": 4.3067,
      "step": 28
    },
    {
      "epoch": 1.0,
      "eval_loss": 4.093968868255615,
      "eval_runtime": 109.4688,
      "eval_samples_per_second": 3.197,
      "eval_steps_per_second": 0.055,
      "step": 28
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 0.002467899350449443,
      "learning_rate": 0.0002706756700001086,
      "loss": 4.8521,
      "step": 30
    },
    {
      "epoch": 1.0714285714285714,
      "eval_loss": 4.093968868255615,
      "eval_runtime": 108.5053,
      "eval_samples_per_second": 3.226,
      "eval_steps_per_second": 0.055,
      "step": 30
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.004230957478284836,
      "learning_rate": 0.00024985446461548485,
      "loss": 4.8521,
      "step": 32
    },
    {
      "epoch": 1.1428571428571428,
      "eval_loss": 4.093967914581299,
      "eval_runtime": 121.2409,
      "eval_samples_per_second": 2.887,
      "eval_steps_per_second": 0.049,
      "step": 32
    },
    {
      "epoch": 1.2142857142857142,
      "grad_norm": 0.0021761604584753513,
      "learning_rate": 0.00022903325923086113,
      "loss": 4.8521,
      "step": 34
    },
    {
      "epoch": 1.2142857142857142,
      "eval_loss": 4.093955993652344,
      "eval_runtime": 150.1791,
      "eval_samples_per_second": 2.331,
      "eval_steps_per_second": 0.04,
      "step": 34
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 0.0025435429997742176,
      "learning_rate": 0.00020821205384623738,
      "loss": 4.8521,
      "step": 36
    },
    {
      "epoch": 1.2857142857142856,
      "eval_loss": 4.093955039978027,
      "eval_runtime": 138.5452,
      "eval_samples_per_second": 2.526,
      "eval_steps_per_second": 0.043,
      "step": 36
    },
    {
      "epoch": 1.3571428571428572,
      "grad_norm": 0.0025120778009295464,
      "learning_rate": 0.00018739084846161366,
      "loss": 4.852,
      "step": 38
    },
    {
      "epoch": 1.3571428571428572,
      "eval_loss": 4.093952178955078,
      "eval_runtime": 159.7134,
      "eval_samples_per_second": 2.191,
      "eval_steps_per_second": 0.038,
      "step": 38
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.0023238961584866047,
      "learning_rate": 0.0001665696430769899,
      "loss": 4.852,
      "step": 40
    },
    {
      "epoch": 1.4285714285714286,
      "eval_loss": 4.093949317932129,
      "eval_runtime": 143.6861,
      "eval_samples_per_second": 2.436,
      "eval_steps_per_second": 0.042,
      "step": 40
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.0012704160762950778,
      "learning_rate": 0.00014574843769236617,
      "loss": 4.852,
      "step": 42
    },
    {
      "epoch": 1.5,
      "eval_loss": 4.0939459800720215,
      "eval_runtime": 122.8463,
      "eval_samples_per_second": 2.849,
      "eval_steps_per_second": 0.049,
      "step": 42
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 0.0016051293350756168,
      "learning_rate": 0.00012492723230774242,
      "loss": 4.852,
      "step": 44
    },
    {
      "epoch": 1.5714285714285714,
      "eval_loss": 4.093944549560547,
      "eval_runtime": 107.8466,
      "eval_samples_per_second": 3.245,
      "eval_steps_per_second": 0.056,
      "step": 44
    },
    {
      "epoch": 1.6428571428571428,
      "grad_norm": 0.0015636960742995143,
      "learning_rate": 0.00010410602692311869,
      "loss": 4.852,
      "step": 46
    },
    {
      "epoch": 1.6428571428571428,
      "eval_loss": 4.093943119049072,
      "eval_runtime": 129.3014,
      "eval_samples_per_second": 2.707,
      "eval_steps_per_second": 0.046,
      "step": 46
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.001410445780493319,
      "learning_rate": 8.328482153849495e-05,
      "loss": 4.8521,
      "step": 48
    },
    {
      "epoch": 1.7142857142857144,
      "eval_loss": 4.093941688537598,
      "eval_runtime": 109.6222,
      "eval_samples_per_second": 3.193,
      "eval_steps_per_second": 0.055,
      "step": 48
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.001393631100654602,
      "learning_rate": 6.246361615387121e-05,
      "loss": 4.852,
      "step": 50
    },
    {
      "epoch": 1.7857142857142856,
      "eval_loss": 4.093940734863281,
      "eval_runtime": 122.8998,
      "eval_samples_per_second": 2.848,
      "eval_steps_per_second": 0.049,
      "step": 50
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.0009566438966430724,
      "learning_rate": 4.164241076924747e-05,
      "loss": 4.852,
      "step": 52
    },
    {
      "epoch": 1.8571428571428572,
      "eval_loss": 4.093940258026123,
      "eval_runtime": 134.9009,
      "eval_samples_per_second": 2.594,
      "eval_steps_per_second": 0.044,
      "step": 52
    },
    {
      "epoch": 1.9285714285714286,
      "grad_norm": 0.0013086905237287283,
      "learning_rate": 2.0821205384623736e-05,
      "loss": 4.852,
      "step": 54
    },
    {
      "epoch": 1.9285714285714286,
      "eval_loss": 4.093940258026123,
      "eval_runtime": 126.2369,
      "eval_samples_per_second": 2.773,
      "eval_steps_per_second": 0.048,
      "step": 54
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.0012161381309852004,
      "learning_rate": 0.0,
      "loss": 4.3066,
      "step": 56
    },
    {
      "epoch": 2.0,
      "eval_loss": 4.093940258026123,
      "eval_runtime": 110.6877,
      "eval_samples_per_second": 3.162,
      "eval_steps_per_second": 0.054,
      "step": 56
    }
  ],
  "logging_steps": 2,
  "max_steps": 56,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 407052651766068.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "_wandb": {},
    "assignments": {},
    "decay": 0.01,
    "learning_rate": 0.0005829937507694647,
    "metric": "eval/loss",
    "per_device_train_batch_size": 128
  }
}