{
  "best_metric": 1.634413242340088,
  "best_model_checkpoint": "outputs/checkpoint-442",
  "epoch": 12.67383512544803,
  "eval_steps": 500,
  "global_step": 442,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5734767025089605,
      "grad_norm": 1.2666417360305786,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.9248,
      "step": 20
    },
    {
      "epoch": 0.974910394265233,
      "eval_loss": 2.8157026767730713,
      "eval_runtime": 11.7687,
      "eval_samples_per_second": 31.609,
      "eval_steps_per_second": 3.994,
      "step": 34
    },
    {
      "epoch": 1.146953405017921,
      "grad_norm": 1.3574857711791992,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.8919,
      "step": 40
    },
    {
      "epoch": 1.7204301075268817,
      "grad_norm": 1.58558189868927,
      "learning_rate": 6e-06,
      "loss": 2.8246,
      "step": 60
    },
    {
      "epoch": 1.978494623655914,
      "eval_loss": 2.589261293411255,
      "eval_runtime": 11.7715,
      "eval_samples_per_second": 31.602,
      "eval_steps_per_second": 3.993,
      "step": 69
    },
    {
      "epoch": 2.293906810035842,
      "grad_norm": 2.0228383541107178,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.6556,
      "step": 80
    },
    {
      "epoch": 2.867383512544803,
      "grad_norm": 1.7711297273635864,
      "learning_rate": 1e-05,
      "loss": 2.2948,
      "step": 100
    },
    {
      "epoch": 2.982078853046595,
      "eval_loss": 1.9966150522232056,
      "eval_runtime": 11.7703,
      "eval_samples_per_second": 31.605,
      "eval_steps_per_second": 3.993,
      "step": 104
    },
    {
      "epoch": 3.4408602150537635,
      "grad_norm": 0.7906980514526367,
      "learning_rate": 9.915855517973776e-06,
      "loss": 2.0075,
      "step": 120
    },
    {
      "epoch": 3.985663082437276,
      "eval_loss": 1.8547258377075195,
      "eval_runtime": 11.7654,
      "eval_samples_per_second": 31.618,
      "eval_steps_per_second": 3.995,
      "step": 139
    },
    {
      "epoch": 4.014336917562724,
      "grad_norm": 0.6804280281066895,
      "learning_rate": 9.666254189437286e-06,
      "loss": 1.9264,
      "step": 140
    },
    {
      "epoch": 4.587813620071684,
      "grad_norm": 0.6464373469352722,
      "learning_rate": 9.259597044191635e-06,
      "loss": 1.8719,
      "step": 160
    },
    {
      "epoch": 4.989247311827957,
      "eval_loss": 1.7840723991394043,
      "eval_runtime": 11.7636,
      "eval_samples_per_second": 31.623,
      "eval_steps_per_second": 3.995,
      "step": 174
    },
    {
      "epoch": 5.161290322580645,
      "grad_norm": 0.6501721739768982,
      "learning_rate": 8.709571264176408e-06,
      "loss": 1.8091,
      "step": 180
    },
    {
      "epoch": 5.734767025089606,
      "grad_norm": 0.6568534970283508,
      "learning_rate": 8.034689503135785e-06,
      "loss": 1.7829,
      "step": 200
    },
    {
      "epoch": 5.992831541218638,
      "eval_loss": 1.7349460124969482,
      "eval_runtime": 11.7693,
      "eval_samples_per_second": 31.608,
      "eval_steps_per_second": 3.993,
      "step": 209
    },
    {
      "epoch": 6.308243727598566,
      "grad_norm": 0.701804518699646,
      "learning_rate": 7.257666791554448e-06,
      "loss": 1.7286,
      "step": 220
    },
    {
      "epoch": 6.881720430107527,
      "grad_norm": 0.6902477741241455,
      "learning_rate": 6.4046559988678485e-06,
      "loss": 1.7219,
      "step": 240
    },
    {
      "epoch": 6.996415770609319,
      "eval_loss": 1.697705626487732,
      "eval_runtime": 11.7645,
      "eval_samples_per_second": 31.621,
      "eval_steps_per_second": 3.995,
      "step": 244
    },
    {
      "epoch": 7.455197132616488,
      "grad_norm": 0.6586928963661194,
      "learning_rate": 5.504367585601342e-06,
      "loss": 1.7072,
      "step": 260
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.670423984527588,
      "eval_runtime": 11.7671,
      "eval_samples_per_second": 31.614,
      "eval_steps_per_second": 3.994,
      "step": 279
    },
    {
      "epoch": 8.028673835125447,
      "grad_norm": 0.6181725263595581,
      "learning_rate": 4.587103272638339e-06,
      "loss": 1.6589,
      "step": 280
    },
    {
      "epoch": 8.602150537634408,
      "grad_norm": 0.7612842321395874,
      "learning_rate": 3.6837361521770056e-06,
      "loss": 1.6622,
      "step": 300
    },
    {
      "epoch": 8.974910394265233,
      "eval_loss": 1.6525731086730957,
      "eval_runtime": 11.7681,
      "eval_samples_per_second": 31.611,
      "eval_steps_per_second": 3.994,
      "step": 313
    },
    {
      "epoch": 9.175627240143369,
      "grad_norm": 0.7089374661445618,
      "learning_rate": 2.8246715675896354e-06,
      "loss": 1.6601,
      "step": 320
    },
    {
      "epoch": 9.74910394265233,
      "grad_norm": 0.8760582804679871,
      "learning_rate": 2.0388237366751005e-06,
      "loss": 1.6416,
      "step": 340
    },
    {
      "epoch": 9.978494623655914,
      "eval_loss": 1.6415046453475952,
      "eval_runtime": 11.7688,
      "eval_samples_per_second": 31.609,
      "eval_steps_per_second": 3.994,
      "step": 348
    },
    {
      "epoch": 10.32258064516129,
      "grad_norm": 0.7347344160079956,
      "learning_rate": 1.3526425629068968e-06,
      "loss": 1.6421,
      "step": 360
    },
    {
      "epoch": 10.89605734767025,
      "grad_norm": 0.7853028774261475,
      "learning_rate": 7.89223390062172e-07,
      "loss": 1.6276,
      "step": 380
    },
    {
      "epoch": 10.982078853046595,
      "eval_loss": 1.6362229585647583,
      "eval_runtime": 11.826,
      "eval_samples_per_second": 31.456,
      "eval_steps_per_second": 3.974,
      "step": 383
    },
    {
      "epoch": 11.469534050179211,
      "grad_norm": 0.7689109444618225,
      "learning_rate": 3.675296639259912e-07,
      "loss": 1.6334,
      "step": 400
    },
    {
      "epoch": 11.985663082437275,
      "eval_loss": 1.6345170736312866,
      "eval_runtime": 11.8501,
      "eval_samples_per_second": 31.392,
      "eval_steps_per_second": 3.966,
      "step": 418
    },
    {
      "epoch": 12.043010752688172,
      "grad_norm": 0.7731339335441589,
      "learning_rate": 1.0175466456213034e-07,
      "loss": 1.6217,
      "step": 420
    },
    {
      "epoch": 12.616487455197133,
      "grad_norm": 0.7765325307846069,
      "learning_rate": 8.437918333864537e-10,
      "loss": 1.6348,
      "step": 440
    },
    {
      "epoch": 12.67383512544803,
      "eval_loss": 1.634413242340088,
      "eval_runtime": 11.9143,
      "eval_samples_per_second": 31.223,
      "eval_steps_per_second": 3.945,
      "step": 442
    }
  ],
  "logging_steps": 20,
  "max_steps": 442,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 13,
  "save_steps": 500,
  "total_flos": 2.327705803685069e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}