{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.12,
  "eval_steps": 4,
  "global_step": 39,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "grad_norm": 17.890169143676758,
      "learning_rate": 2e-05,
      "loss": 6.6299,
      "step": 1
    },
    {
      "epoch": 0.08,
      "eval_loss": 6.932005405426025,
      "eval_runtime": 5.2841,
      "eval_samples_per_second": 2.271,
      "eval_steps_per_second": 1.135,
      "step": 1
    },
    {
      "epoch": 0.16,
      "grad_norm": 17.0613956451416,
      "learning_rate": 4e-05,
      "loss": 6.8062,
      "step": 2
    },
    {
      "epoch": 0.24,
      "grad_norm": 19.356693267822266,
      "learning_rate": 6e-05,
      "loss": 7.0697,
      "step": 3
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.688406944274902,
      "learning_rate": 8e-05,
      "loss": 5.9686,
      "step": 4
    },
    {
      "epoch": 0.32,
      "eval_loss": 4.446321964263916,
      "eval_runtime": 5.3077,
      "eval_samples_per_second": 2.261,
      "eval_steps_per_second": 1.13,
      "step": 4
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.240882873535156,
      "learning_rate": 0.0001,
      "loss": 4.5161,
      "step": 5
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.287125587463379,
      "learning_rate": 0.00012,
      "loss": 2.5736,
      "step": 6
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.680168628692627,
      "learning_rate": 0.00014,
      "loss": 0.9119,
      "step": 7
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.75272536277771,
      "learning_rate": 0.00016,
      "loss": 0.5956,
      "step": 8
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.5577284693717957,
      "eval_runtime": 5.323,
      "eval_samples_per_second": 2.254,
      "eval_steps_per_second": 1.127,
      "step": 8
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.569386959075928,
      "learning_rate": 0.00018,
      "loss": 0.6499,
      "step": 9
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.337709426879883,
      "learning_rate": 0.0002,
      "loss": 1.6675,
      "step": 10
    },
    {
      "epoch": 0.88,
      "grad_norm": 3.6579396724700928,
      "learning_rate": 0.00019981755542233177,
      "loss": 0.5612,
      "step": 11
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.5605533123016357,
      "learning_rate": 0.0001992708874098054,
      "loss": 0.4848,
      "step": 12
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.8369883894920349,
      "eval_runtime": 5.327,
      "eval_samples_per_second": 2.253,
      "eval_steps_per_second": 1.126,
      "step": 12
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.139214515686035,
      "learning_rate": 0.00019836199069471437,
      "loss": 0.9289,
      "step": 13
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.3758115768432617,
      "learning_rate": 0.0001970941817426052,
      "loss": 0.5154,
      "step": 14
    },
    {
      "epoch": 1.2,
      "grad_norm": 2.3552050590515137,
      "learning_rate": 0.00019547208665085457,
      "loss": 0.5028,
      "step": 15
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.0085458755493164,
      "learning_rate": 0.0001935016242685415,
      "loss": 0.4913,
      "step": 16
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.4895554482936859,
      "eval_runtime": 5.325,
      "eval_samples_per_second": 2.254,
      "eval_steps_per_second": 1.127,
      "step": 16
    },
    {
      "epoch": 1.36,
      "grad_norm": 3.799877643585205,
      "learning_rate": 0.00019118998459920902,
      "loss": 0.3925,
      "step": 17
    },
    {
      "epoch": 1.44,
      "grad_norm": 6.162234783172607,
      "learning_rate": 0.000188545602565321,
      "loss": 1.0242,
      "step": 18
    },
    {
      "epoch": 1.52,
      "grad_norm": 4.3233642578125,
      "learning_rate": 0.00018557812723014476,
      "loss": 0.7239,
      "step": 19
    },
    {
      "epoch": 1.6,
      "grad_norm": 4.661636829376221,
      "learning_rate": 0.00018229838658936564,
      "loss": 0.671,
      "step": 20
    },
    {
      "epoch": 1.6,
      "eval_loss": 0.5175419449806213,
      "eval_runtime": 5.3241,
      "eval_samples_per_second": 2.254,
      "eval_steps_per_second": 1.127,
      "step": 20
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.8649834394454956,
      "learning_rate": 0.00017871834806090501,
      "loss": 0.3227,
      "step": 21
    },
    {
      "epoch": 1.76,
      "grad_norm": 18.722929000854492,
      "learning_rate": 0.00017485107481711012,
      "loss": 1.5111,
      "step": 22
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.523069381713867,
      "learning_rate": 0.00017071067811865476,
      "loss": 0.5847,
      "step": 23
    },
    {
      "epoch": 1.92,
      "grad_norm": 125.49153900146484,
      "learning_rate": 0.00016631226582407952,
      "loss": 2.6136,
      "step": 24
    },
    {
      "epoch": 1.92,
      "eval_loss": 2.3445661067962646,
      "eval_runtime": 5.3223,
      "eval_samples_per_second": 2.255,
      "eval_steps_per_second": 1.127,
      "step": 24
    },
    {
      "epoch": 2.0,
      "grad_norm": 92.8938980102539,
      "learning_rate": 0.00016167188726285434,
      "loss": 2.5399,
      "step": 25
    },
    {
      "epoch": 2.08,
      "grad_norm": 18.453845977783203,
      "learning_rate": 0.00015680647467311557,
      "loss": 0.6644,
      "step": 26
    },
    {
      "epoch": 2.16,
      "grad_norm": 7.2362141609191895,
      "learning_rate": 0.00015173378141776568,
      "loss": 0.3526,
      "step": 27
    },
    {
      "epoch": 2.24,
      "grad_norm": 8.623927116394043,
      "learning_rate": 0.00014647231720437686,
      "loss": 0.6383,
      "step": 28
    },
    {
      "epoch": 2.24,
      "eval_loss": 0.5194393992424011,
      "eval_runtime": 5.3225,
      "eval_samples_per_second": 2.255,
      "eval_steps_per_second": 1.127,
      "step": 28
    },
    {
      "epoch": 2.32,
      "grad_norm": 4.0871782302856445,
      "learning_rate": 0.0001410412805452757,
      "loss": 0.5128,
      "step": 29
    },
    {
      "epoch": 2.4,
      "grad_norm": 4.378565311431885,
      "learning_rate": 0.00013546048870425356,
      "loss": 0.4853,
      "step": 30
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.0163638591766357,
      "learning_rate": 0.00012975030538552032,
      "loss": 0.4083,
      "step": 31
    },
    {
      "epoch": 2.56,
      "grad_norm": 12.063375473022461,
      "learning_rate": 0.0001239315664287558,
      "loss": 0.5776,
      "step": 32
    },
    {
      "epoch": 2.56,
      "eval_loss": 0.5652761459350586,
      "eval_runtime": 5.3235,
      "eval_samples_per_second": 2.254,
      "eval_steps_per_second": 1.127,
      "step": 32
    },
    {
      "epoch": 2.64,
      "grad_norm": 3.3182127475738525,
      "learning_rate": 0.0001180255037813906,
      "loss": 0.5234,
      "step": 33
    },
    {
      "epoch": 2.72,
      "grad_norm": 2.9432356357574463,
      "learning_rate": 0.0001120536680255323,
      "loss": 0.4264,
      "step": 34
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.9505524635314941,
      "learning_rate": 0.00010603784974222861,
      "loss": 0.4241,
      "step": 35
    },
    {
      "epoch": 2.88,
      "grad_norm": 3.3271913528442383,
      "learning_rate": 0.0001,
      "loss": 0.4913,
      "step": 36
    },
    {
      "epoch": 2.88,
      "eval_loss": 0.47906360030174255,
      "eval_runtime": 5.323,
      "eval_samples_per_second": 2.254,
      "eval_steps_per_second": 1.127,
      "step": 36
    },
    {
      "epoch": 2.96,
      "grad_norm": 12.475479125976562,
      "learning_rate": 9.396215025777139e-05,
      "loss": 0.5668,
      "step": 37
    },
    {
      "epoch": 3.04,
      "grad_norm": 3.1950840950012207,
      "learning_rate": 8.79463319744677e-05,
      "loss": 0.3439,
      "step": 38
    },
    {
      "epoch": 3.12,
      "grad_norm": 12.156915664672852,
      "learning_rate": 8.197449621860943e-05,
      "loss": 0.4084,
      "step": 39
    }
  ],
  "logging_steps": 1,
  "max_steps": 62,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 13,
  "total_flos": 2.539161524581171e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}