checkpoint-15/trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 1.3786381483078003,
  "best_model_checkpoint": "./scheme-recommendation-lora/checkpoint-10",
  "epoch": 5.0,
  "eval_steps": 10,
  "global_step": 15,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.48101720213890076,
      "learning_rate": 0.0,
      "loss": 1.6163,
      "step": 1
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.4564454257488251,
      "learning_rate": 8e-05,
      "loss": 1.688,
      "step": 5
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.7571481466293335,
      "learning_rate": 0.00018,
      "loss": 1.5046,
      "step": 10
    },
    {
      "epoch": 3.3636363636363638,
      "eval_loss": 1.3786381483078003,
      "eval_runtime": 9.2062,
      "eval_samples_per_second": 0.869,
      "eval_steps_per_second": 0.434,
      "step": 10
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6026115417480469,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 1.1841,
      "step": 15
    }
  ],
  "logging_steps": 5,
  "max_steps": 15,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.00875657805824e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}