trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.35294117647058826,
  "eval_steps": 500,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.029411764705882353,
      "grad_norm": 0.734375,
      "learning_rate": 2e-05,
      "loss": 2.0227,
      "step": 1
    },
    {
      "epoch": 0.058823529411764705,
      "grad_norm": 0.59765625,
      "learning_rate": 4e-05,
      "loss": 1.6615,
      "step": 2
    },
    {
      "epoch": 0.08823529411764706,
      "grad_norm": 0.6640625,
      "learning_rate": 6e-05,
      "loss": 1.9138,
      "step": 3
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 0.9375,
      "learning_rate": 8e-05,
      "loss": 2.2948,
      "step": 4
    },
    {
      "epoch": 0.14705882352941177,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0001,
      "loss": 2.0388,
      "step": 5
    },
    {
      "epoch": 0.17647058823529413,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00012,
      "loss": 1.9193,
      "step": 6
    },
    {
      "epoch": 0.20588235294117646,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00014,
      "loss": 1.7604,
      "step": 7
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 0.734375,
      "learning_rate": 0.00016,
      "loss": 1.6406,
      "step": 8
    },
    {
      "epoch": 0.2647058823529412,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00018,
      "loss": 1.528,
      "step": 9
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002,
      "loss": 1.4681,
      "step": 10
    },
    {
      "epoch": 0.3235294117647059,
      "grad_norm": 1.015625,
      "learning_rate": 0.00019914448613738106,
      "loss": 1.4515,
      "step": 11
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00019659258262890683,
      "loss": 1.4636,
      "step": 12
    }
  ],
  "logging_steps": 1,
  "max_steps": 34,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 12,
  "total_flos": 3.5234198927179776e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}