{
  "best_metric": 1.3614939451217651,
  "best_model_checkpoint": "outputs/checkpoint-488",
  "epoch": 13.992831541218639,
  "eval_steps": 500,
  "global_step": 488,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5734767025089605,
      "grad_norm": 1.2556052207946777,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.9206,
      "step": 20
    },
    {
      "epoch": 0.974910394265233,
      "eval_loss": 2.7616007328033447,
      "eval_runtime": 12.2966,
      "eval_samples_per_second": 30.252,
      "eval_steps_per_second": 3.822,
      "step": 34
    },
    {
      "epoch": 1.146953405017921,
      "grad_norm": 1.4213184118270874,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.8502,
      "step": 40
    },
    {
      "epoch": 1.7204301075268817,
      "grad_norm": 1.9485336542129517,
      "learning_rate": 1.2e-05,
      "loss": 2.6654,
      "step": 60
    },
    {
      "epoch": 1.978494623655914,
      "eval_loss": 2.177459478378296,
      "eval_runtime": 12.2962,
      "eval_samples_per_second": 30.253,
      "eval_steps_per_second": 3.822,
      "step": 69
    },
    {
      "epoch": 2.293906810035842,
      "grad_norm": 0.9282850623130798,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.2325,
      "step": 80
    },
    {
      "epoch": 2.867383512544803,
      "grad_norm": 0.6272071003913879,
      "learning_rate": 2e-05,
      "loss": 1.9127,
      "step": 100
    },
    {
      "epoch": 2.982078853046595,
      "eval_loss": 1.8282170295715332,
      "eval_runtime": 12.2659,
      "eval_samples_per_second": 30.328,
      "eval_steps_per_second": 3.832,
      "step": 104
    },
    {
      "epoch": 3.4408602150537635,
      "grad_norm": 0.6469711065292358,
      "learning_rate": 1.9882804237803487e-05,
      "loss": 1.833,
      "step": 120
    },
    {
      "epoch": 3.985663082437276,
      "eval_loss": 1.7144227027893066,
      "eval_runtime": 12.2541,
      "eval_samples_per_second": 30.357,
      "eval_steps_per_second": 3.835,
      "step": 139
    },
    {
      "epoch": 4.014336917562724,
      "grad_norm": 0.7177844643592834,
      "learning_rate": 1.9533963920549307e-05,
      "loss": 1.7656,
      "step": 140
    },
    {
      "epoch": 4.587813620071684,
      "grad_norm": 0.7868255972862244,
      "learning_rate": 1.8961655569610557e-05,
      "loss": 1.7057,
      "step": 160
    },
    {
      "epoch": 4.989247311827957,
      "eval_loss": 1.6268185377120972,
      "eval_runtime": 12.2554,
      "eval_samples_per_second": 30.354,
      "eval_steps_per_second": 3.835,
      "step": 174
    },
    {
      "epoch": 5.161290322580645,
      "grad_norm": 0.7630313038825989,
      "learning_rate": 1.8179293607667177e-05,
      "loss": 1.6298,
      "step": 180
    },
    {
      "epoch": 5.734767025089606,
      "grad_norm": 0.8428413271903992,
      "learning_rate": 1.720521593600787e-05,
      "loss": 1.5832,
      "step": 200
    },
    {
      "epoch": 5.992831541218638,
      "eval_loss": 1.5533965826034546,
      "eval_runtime": 12.2514,
      "eval_samples_per_second": 30.364,
      "eval_steps_per_second": 3.836,
      "step": 209
    },
    {
      "epoch": 6.308243727598566,
      "grad_norm": 0.9982613921165466,
      "learning_rate": 1.6062254109666383e-05,
      "loss": 1.5144,
      "step": 220
    },
    {
      "epoch": 6.881720430107527,
      "grad_norm": 0.988570511341095,
      "learning_rate": 1.477719818512263e-05,
      "loss": 1.4884,
      "step": 240
    },
    {
      "epoch": 6.996415770609319,
      "eval_loss": 1.4935128688812256,
      "eval_runtime": 12.2503,
      "eval_samples_per_second": 30.367,
      "eval_steps_per_second": 3.837,
      "step": 244
    },
    {
      "epoch": 7.455197132616488,
      "grad_norm": 0.9964269399642944,
      "learning_rate": 1.3380168784085028e-05,
      "loss": 1.4513,
      "step": 260
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.4478424787521362,
      "eval_runtime": 12.2711,
      "eval_samples_per_second": 30.315,
      "eval_steps_per_second": 3.83,
      "step": 279
    },
    {
      "epoch": 8.028673835125447,
      "grad_norm": 1.040726900100708,
      "learning_rate": 1.1903911091646684e-05,
      "loss": 1.3805,
      "step": 280
    },
    {
      "epoch": 8.602150537634408,
      "grad_norm": 1.3048664331436157,
      "learning_rate": 1.0383027336900356e-05,
      "loss": 1.3677,
      "step": 300
    },
    {
      "epoch": 8.974910394265233,
      "eval_loss": 1.4132319688796997,
      "eval_runtime": 12.2651,
      "eval_samples_per_second": 30.33,
      "eval_steps_per_second": 3.832,
      "step": 313
    },
    {
      "epoch": 9.175627240143369,
      "grad_norm": 1.1397103071212769,
      "learning_rate": 8.853165746015997e-06,
      "loss": 1.3506,
      "step": 320
    },
    {
      "epoch": 9.74910394265233,
      "grad_norm": 1.4186557531356812,
      "learning_rate": 7.350184978033386e-06,
      "loss": 1.3173,
      "step": 340
    },
    {
      "epoch": 9.978494623655914,
      "eval_loss": 1.3905054330825806,
      "eval_runtime": 12.26,
      "eval_samples_per_second": 30.343,
      "eval_steps_per_second": 3.834,
      "step": 348
    },
    {
      "epoch": 10.32258064516129,
      "grad_norm": 1.2632255554199219,
      "learning_rate": 5.9093136282866014e-06,
      "loss": 1.3071,
      "step": 360
    },
    {
      "epoch": 10.89605734767025,
      "grad_norm": 1.3758857250213623,
      "learning_rate": 4.56432449998779e-06,
      "loss": 1.2863,
      "step": 380
    },
    {
      "epoch": 10.982078853046595,
      "eval_loss": 1.375404953956604,
      "eval_runtime": 12.2609,
      "eval_samples_per_second": 30.34,
      "eval_steps_per_second": 3.833,
      "step": 383
    },
    {
      "epoch": 11.469534050179211,
      "grad_norm": 1.365923285484314,
      "learning_rate": 3.3467429983443477e-06,
      "loss": 1.2785,
      "step": 400
    },
    {
      "epoch": 11.985663082437275,
      "eval_loss": 1.366598129272461,
      "eval_runtime": 12.2562,
      "eval_samples_per_second": 30.352,
      "eval_steps_per_second": 3.835,
      "step": 418
    },
    {
      "epoch": 12.043010752688172,
      "grad_norm": 1.328456163406372,
      "learning_rate": 2.2851082017805704e-06,
      "loss": 1.2641,
      "step": 420
    },
    {
      "epoch": 12.616487455197133,
      "grad_norm": 1.3476920127868652,
      "learning_rate": 1.4043039301279904e-06,
      "loss": 1.261,
      "step": 440
    },
    {
      "epoch": 12.989247311827956,
      "eval_loss": 1.3626736402511597,
      "eval_runtime": 12.2637,
      "eval_samples_per_second": 30.333,
      "eval_steps_per_second": 3.832,
      "step": 453
    },
    {
      "epoch": 13.189964157706093,
      "grad_norm": 1.4261298179626465,
      "learning_rate": 7.249754889790539e-07,
      "loss": 1.2505,
      "step": 460
    },
    {
      "epoch": 13.763440860215054,
      "grad_norm": 1.580108642578125,
      "learning_rate": 2.6304576122221035e-07,
      "loss": 1.2546,
      "step": 480
    },
    {
      "epoch": 13.992831541218639,
      "eval_loss": 1.3614939451217651,
      "eval_runtime": 12.2686,
      "eval_samples_per_second": 30.321,
      "eval_steps_per_second": 3.831,
      "step": 488
    }
  ],
  "logging_steps": 20,
  "max_steps": 510,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 2.5689245217644544e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}