trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.526315789473684,
  "eval_steps": 500,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.21,
      "grad_norm": 7.908194508354831,
      "learning_rate": 1e-05,
      "loss": 1.6447,
      "step": 1
    },
    {
      "epoch": 0.84,
      "eval_loss": 1.3962677717208862,
      "eval_runtime": 36.3272,
      "eval_samples_per_second": 10.075,
      "eval_steps_per_second": 0.633,
      "step": 4
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.7740427914859738,
      "learning_rate": 1.5877852522924733e-05,
      "loss": 1.4922,
      "step": 5
    },
    {
      "epoch": 1.89,
      "eval_loss": 1.5709341764450073,
      "eval_runtime": 35.8444,
      "eval_samples_per_second": 10.211,
      "eval_steps_per_second": 0.642,
      "step": 9
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.1451956932311864,
      "learning_rate": 1.9098300562505266e-06,
      "loss": 1.1792,
      "step": 10
    },
    {
      "epoch": 2.53,
      "eval_loss": 1.5887579917907715,
      "eval_runtime": 36.3881,
      "eval_samples_per_second": 10.058,
      "eval_steps_per_second": 0.632,
      "step": 12
    },
    {
      "epoch": 2.53,
      "step": 12,
      "total_flos": 2460210954240.0,
      "train_loss": 1.2966619332631428,
      "train_runtime": 333.6869,
      "train_samples_per_second": 2.724,
      "train_steps_per_second": 0.036
    }
  ],
  "logging_steps": 5,
  "max_steps": 12,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2460210954240.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}