trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9987368421052631,
  "eval_steps": 100,
  "global_step": 593,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17,
      "grad_norm": 0.5140721201896667,
      "learning_rate": 1.972335510674149e-05,
      "loss": 1.7203,
      "step": 100
    },
    {
      "epoch": 0.17,
      "eval_loss": 1.0655403137207031,
      "eval_runtime": 108.9551,
      "eval_samples_per_second": 4.589,
      "eval_steps_per_second": 1.147,
      "step": 100
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6345012784004211,
      "learning_rate": 1.678421935868953e-05,
      "loss": 1.073,
      "step": 200
    },
    {
      "epoch": 0.34,
      "eval_loss": 1.0231871604919434,
      "eval_runtime": 109.7226,
      "eval_samples_per_second": 4.557,
      "eval_steps_per_second": 1.139,
      "step": 200
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.5556415915489197,
      "learning_rate": 1.1555611635532276e-05,
      "loss": 1.0453,
      "step": 300
    },
    {
      "epoch": 0.51,
      "eval_loss": 1.007139801979065,
      "eval_runtime": 109.1597,
      "eval_samples_per_second": 4.58,
      "eval_steps_per_second": 1.145,
      "step": 300
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.6960105895996094,
      "learning_rate": 5.8020313170922526e-06,
      "loss": 1.0422,
      "step": 400
    },
    {
      "epoch": 0.67,
      "eval_loss": 0.9994553327560425,
      "eval_runtime": 109.203,
      "eval_samples_per_second": 4.579,
      "eval_steps_per_second": 1.145,
      "step": 400
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5549703240394592,
      "learning_rate": 1.46514039852793e-06,
      "loss": 1.0179,
      "step": 500
    },
    {
      "epoch": 0.84,
      "eval_loss": 0.9968265295028687,
      "eval_runtime": 109.0565,
      "eval_samples_per_second": 4.585,
      "eval_steps_per_second": 1.146,
      "step": 500
    }
  ],
  "logging_steps": 100,
  "max_steps": 593,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 1.001225674624598e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}