checkpoint-15/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.857142857142857,
  "eval_steps": 500,
  "global_step": 15,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.19,
      "grad_norm": 6.906898498535156,
      "learning_rate": 0.0001,
      "loss": 5.0884,
      "step": 1
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.150900840759277,
      "learning_rate": 0.0002,
      "loss": 4.6143,
      "step": 2
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.832034111022949,
      "learning_rate": 0.00018461538461538463,
      "loss": 4.4776,
      "step": 3
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.760493755340576,
      "learning_rate": 0.00016923076923076923,
      "loss": 4.4845,
      "step": 4
    },
    {
      "epoch": 0.95,
      "grad_norm": NaN,
      "learning_rate": 0.00016923076923076923,
      "loss": 4.8731,
      "step": 5
    },
    {
      "epoch": 1.14,
      "grad_norm": 5.546196937561035,
      "learning_rate": 0.00015384615384615385,
      "loss": 4.0771,
      "step": 6
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.478511810302734,
      "learning_rate": 0.00013846153846153847,
      "loss": 4.4343,
      "step": 7
    },
    {
      "epoch": 1.52,
      "grad_norm": 6.995128154754639,
      "learning_rate": 0.0001230769230769231,
      "loss": 3.6025,
      "step": 8
    },
    {
      "epoch": 1.71,
      "grad_norm": 6.940845012664795,
      "learning_rate": 0.0001076923076923077,
      "loss": 3.7716,
      "step": 9
    },
    {
      "epoch": 1.9,
      "grad_norm": 4.949404239654541,
      "learning_rate": 9.230769230769232e-05,
      "loss": 3.7752,
      "step": 10
    },
    {
      "epoch": 2.1,
      "grad_norm": 5.935202121734619,
      "learning_rate": 7.692307692307693e-05,
      "loss": 3.5705,
      "step": 11
    },
    {
      "epoch": 2.29,
      "grad_norm": 5.354796886444092,
      "learning_rate": 6.153846153846155e-05,
      "loss": 3.4207,
      "step": 12
    },
    {
      "epoch": 2.48,
      "grad_norm": 5.133244514465332,
      "learning_rate": 4.615384615384616e-05,
      "loss": 3.5064,
      "step": 13
    },
    {
      "epoch": 2.67,
      "grad_norm": 5.422821521759033,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 3.3041,
      "step": 14
    },
    {
      "epoch": 2.86,
      "grad_norm": 6.032769203186035,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 3.2407,
      "step": 15
    }
  ],
  "logging_steps": 1,
  "max_steps": 15,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 203616748830720.0,
  "train_batch_size": 5,
  "trial_name": null,
  "trial_params": null
}