{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1085,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2304147465437788,
      "grad_norm": 0.1171875,
      "learning_rate": 4.9423963133640554e-05,
      "loss": 0.45,
      "step": 50
    },
    {
      "epoch": 0.4608294930875576,
      "grad_norm": 0.12060546875,
      "learning_rate": 4.8847926267281106e-05,
      "loss": 0.2531,
      "step": 100
    },
    {
      "epoch": 0.6912442396313364,
      "grad_norm": 0.1259765625,
      "learning_rate": 4.827188940092166e-05,
      "loss": 0.2317,
      "step": 150
    },
    {
      "epoch": 0.9216589861751152,
      "grad_norm": 0.11865234375,
      "learning_rate": 4.7695852534562216e-05,
      "loss": 0.2204,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.21880939602851868,
      "eval_runtime": 52.8765,
      "eval_samples_per_second": 14.827,
      "eval_steps_per_second": 0.473,
      "step": 217
    },
    {
      "epoch": 1.1520737327188941,
      "grad_norm": 0.1201171875,
      "learning_rate": 4.711981566820277e-05,
      "loss": 0.2114,
      "step": 250
    },
    {
      "epoch": 1.3824884792626728,
      "grad_norm": 0.1083984375,
      "learning_rate": 4.654377880184332e-05,
      "loss": 0.2109,
      "step": 300
    },
    {
      "epoch": 1.6129032258064515,
      "grad_norm": 0.11865234375,
      "learning_rate": 4.596774193548387e-05,
      "loss": 0.2045,
      "step": 350
    },
    {
      "epoch": 1.8433179723502304,
      "grad_norm": 0.158203125,
      "learning_rate": 4.539170506912442e-05,
      "loss": 0.21,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.20793424546718597,
      "eval_runtime": 52.6064,
      "eval_samples_per_second": 14.903,
      "eval_steps_per_second": 0.475,
      "step": 434
    },
    {
      "epoch": 2.0737327188940093,
      "grad_norm": 0.12109375,
      "learning_rate": 4.4815668202764974e-05,
      "loss": 0.2032,
      "step": 450
    },
    {
      "epoch": 2.3041474654377883,
      "grad_norm": 0.1201171875,
      "learning_rate": 4.423963133640553e-05,
      "loss": 0.2001,
      "step": 500
    },
    {
      "epoch": 2.5345622119815667,
      "grad_norm": 0.125,
      "learning_rate": 4.366359447004609e-05,
      "loss": 0.1953,
      "step": 550
    },
    {
      "epoch": 2.7649769585253456,
      "grad_norm": 0.130859375,
      "learning_rate": 4.308755760368664e-05,
      "loss": 0.1956,
      "step": 600
    },
    {
      "epoch": 2.9953917050691246,
      "grad_norm": 0.1220703125,
      "learning_rate": 4.2511520737327194e-05,
      "loss": 0.1978,
      "step": 650
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.20394934713840485,
      "eval_runtime": 52.684,
      "eval_samples_per_second": 14.881,
      "eval_steps_per_second": 0.475,
      "step": 651
    },
    {
      "epoch": 3.225806451612903,
      "grad_norm": 0.1376953125,
      "learning_rate": 4.1935483870967746e-05,
      "loss": 0.1875,
      "step": 700
    },
    {
      "epoch": 3.456221198156682,
      "grad_norm": 0.119140625,
      "learning_rate": 4.13594470046083e-05,
      "loss": 0.1883,
      "step": 750
    },
    {
      "epoch": 3.686635944700461,
      "grad_norm": 0.1376953125,
      "learning_rate": 4.078341013824885e-05,
      "loss": 0.1925,
      "step": 800
    },
    {
      "epoch": 3.9170506912442398,
      "grad_norm": 0.1357421875,
      "learning_rate": 4.02073732718894e-05,
      "loss": 0.1903,
      "step": 850
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.20200659334659576,
      "eval_runtime": 53.2269,
      "eval_samples_per_second": 14.729,
      "eval_steps_per_second": 0.47,
      "step": 868
    },
    {
      "epoch": 4.147465437788019,
      "grad_norm": 0.140625,
      "learning_rate": 3.963133640552996e-05,
      "loss": 0.1838,
      "step": 900
    },
    {
      "epoch": 4.377880184331797,
      "grad_norm": 0.1455078125,
      "learning_rate": 3.905529953917051e-05,
      "loss": 0.182,
      "step": 950
    },
    {
      "epoch": 4.6082949308755765,
      "grad_norm": 0.1396484375,
      "learning_rate": 3.847926267281106e-05,
      "loss": 0.1827,
      "step": 1000
    },
    {
      "epoch": 4.838709677419355,
      "grad_norm": 0.1416015625,
      "learning_rate": 3.7903225806451614e-05,
      "loss": 0.1808,
      "step": 1050
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.2014361470937729,
      "eval_runtime": 54.1943,
      "eval_samples_per_second": 14.466,
      "eval_steps_per_second": 0.461,
      "step": 1085
    }
  ],
  "logging_steps": 50,
  "max_steps": 4340,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 7.632310487786455e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}