trainer_state.json

{
  "best_metric": 7.237969398498535,
  "best_model_checkpoint": "out/checkpoint-9",
  "epoch": 0.03769633507853403,
  "eval_steps": 1,
  "global_step": 9,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 17.85124397277832,
      "learning_rate": 0.0002,
      "loss": 10.6824,
      "step": 1
    },
    {
      "epoch": 0.0,
      "eval_loss": 9.331878662109375,
      "eval_runtime": 735.4997,
      "eval_samples_per_second": 0.305,
      "eval_steps_per_second": 0.305,
      "step": 1
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.460932731628418,
      "learning_rate": 0.0002,
      "loss": 9.4215,
      "step": 2
    },
    {
      "epoch": 0.01,
      "eval_loss": 8.825074195861816,
      "eval_runtime": 735.7631,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.690059661865234,
      "learning_rate": 0.0002,
      "loss": 7.7604,
      "step": 3
    },
    {
      "epoch": 0.01,
      "eval_loss": 8.533729553222656,
      "eval_runtime": 735.7831,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 3
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.018722057342529,
      "learning_rate": 0.0002,
      "loss": 7.94,
      "step": 4
    },
    {
      "epoch": 0.02,
      "eval_loss": 8.247579574584961,
      "eval_runtime": 735.7455,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 4
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.668647289276123,
      "learning_rate": 0.0002,
      "loss": 7.3386,
      "step": 5
    },
    {
      "epoch": 0.02,
      "eval_loss": 8.042969703674316,
      "eval_runtime": 735.6593,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 5
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.495285034179688,
      "learning_rate": 0.0002,
      "loss": 8.5567,
      "step": 6
    },
    {
      "epoch": 0.03,
      "eval_loss": 7.848848819732666,
      "eval_runtime": 735.6542,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 6
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.991780281066895,
      "learning_rate": 0.0002,
      "loss": 7.1324,
      "step": 7
    },
    {
      "epoch": 0.03,
      "eval_loss": 7.553219795227051,
      "eval_runtime": 735.6545,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 7
    },
    {
      "epoch": 0.03,
      "grad_norm": Infinity,
      "learning_rate": 0.0002,
      "loss": 6.999,
      "step": 8
    },
    {
      "epoch": 0.03,
      "eval_loss": 7.553219795227051,
      "eval_runtime": 735.6924,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 8
    },
    {
      "epoch": 0.04,
      "grad_norm": 15.588254928588867,
      "learning_rate": 0.0002,
      "loss": 8.0144,
      "step": 9
    },
    {
      "epoch": 0.04,
      "eval_loss": 7.237969398498535,
      "eval_runtime": 735.7337,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 9
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1,
  "total_flos": 1719982274641920.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}