last-checkpoint/trainer_state.json

{
  "best_metric": 4.06580114364624,
  "best_model_checkpoint": "./ES_corlec/checkpoint-2000",
  "epoch": 0.3965107057890563,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1,
      "learning_rate": 9.914549390452318e-07,
      "loss": 4.6573,
      "step": 500
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.77213170787285e-07,
      "loss": 4.1911,
      "step": 1000
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.62971402529338e-07,
      "loss": 4.1368,
      "step": 1500
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.487296342713911e-07,
      "loss": 4.0957,
      "step": 2000
    },
    {
      "epoch": 0.4,
      "eval_loss": 4.06580114364624,
      "eval_runtime": 962.1135,
      "eval_samples_per_second": 42.075,
      "eval_steps_per_second": 2.631,
      "step": 2000
    }
  ],
  "max_steps": 35308,
  "num_train_epochs": 7,
  "total_flos": 2449612800000000.0,
  "trial_name": null,
  "trial_params": null
}