checkpoint-186/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.904761904761905,
  "eval_steps": 500,
  "global_step": 186,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.79,
      "learning_rate": 0.00019281437125748504,
      "loss": 0.7324,
      "step": 25
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.000162874251497006,
      "loss": 0.1678,
      "step": 50
    },
    {
      "epoch": 2.38,
      "learning_rate": 0.00013293413173652696,
      "loss": 0.1251,
      "step": 75
    },
    {
      "epoch": 3.17,
      "learning_rate": 0.0001029940119760479,
      "loss": 0.1184,
      "step": 100
    },
    {
      "epoch": 3.97,
      "learning_rate": 7.305389221556886e-05,
      "loss": 0.1148,
      "step": 125
    },
    {
      "epoch": 4.76,
      "learning_rate": 4.311377245508982e-05,
      "loss": 0.112,
      "step": 150
    },
    {
      "epoch": 5.56,
      "learning_rate": 1.317365269461078e-05,
      "loss": 0.1107,
      "step": 175
    }
  ],
  "logging_steps": 25,
  "max_steps": 186,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 1.764247973564252e+17,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}