checkpoint-8/trainer_state.json

{
  "best_metric": 1.9868038892745972,
  "best_model_checkpoint": "./out/checkpoint-8",
  "epoch": 0.49230769230769234,
  "eval_steps": 4,
  "global_step": 8,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.1256,
      "step": 1
    },
    {
      "epoch": 0.06,
      "eval_loss": 2.164146661758423,
      "eval_runtime": 4.1785,
      "eval_samples_per_second": 6.701,
      "eval_steps_per_second": 1.675,
      "step": 1
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.1478,
      "step": 2
    },
    {
      "epoch": 0.18,
      "learning_rate": 6e-06,
      "loss": 2.1527,
      "step": 3
    },
    {
      "epoch": 0.25,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.1049,
      "step": 4
    },
    {
      "epoch": 0.25,
      "eval_loss": 2.125430107116699,
      "eval_runtime": 4.2671,
      "eval_samples_per_second": 6.562,
      "eval_steps_per_second": 1.64,
      "step": 4
    },
    {
      "epoch": 0.31,
      "learning_rate": 1e-05,
      "loss": 2.0978,
      "step": 5
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.2e-05,
      "loss": 2.0632,
      "step": 6
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 2.0281,
      "step": 7
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.9826,
      "step": 8
    },
    {
      "epoch": 0.49,
      "eval_loss": 1.9868038892745972,
      "eval_runtime": 4.2896,
      "eval_samples_per_second": 6.527,
      "eval_steps_per_second": 1.632,
      "step": 8
    }
  ],
  "logging_steps": 1,
  "max_steps": 16,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 8,
  "total_flos": 2.078490581847245e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}