{
  "best_global_step": 2,
  "best_metric": 1.0,
  "best_model_checkpoint": "/predifine/results/experiment_Qwen_Qwen3_0.6B_9/checkpoint-2",
  "epoch": 2.0,
  "eval_steps": 1,
  "global_step": 2,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "accuracy": 0.625,
      "epoch": 1.0,
      "grad_norm": 19.023712158203125,
      "learning_rate": 0.0,
      "loss": 0.5563,
      "num_input_tokens_seen": 3384,
      "step": 1,
      "train_runtime": 1.4671,
      "train_tokens_per_second": 2306.516
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5,
      "eval_loss": 1.0934828519821167,
      "eval_runtime": 0.1697,
      "eval_samples_per_second": 11.783,
      "eval_steps_per_second": 5.892,
      "num_input_tokens_seen": 3384,
      "step": 1
    },
    {
      "accuracy": 0.625,
      "epoch": 2.0,
      "grad_norm": 18.854581832885742,
      "learning_rate": 0.0002,
      "loss": 0.5579,
      "num_input_tokens_seen": 6768,
      "step": 2,
      "train_runtime": 3.7099,
      "train_tokens_per_second": 1824.316
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 1.0,
      "eval_loss": 0.689201831817627,
      "eval_runtime": 0.1734,
      "eval_samples_per_second": 11.535,
      "eval_steps_per_second": 5.768,
      "num_input_tokens_seen": 6768,
      "step": 2
    }
  ],
  "logging_steps": 1,
  "max_steps": 3,
  "num_input_tokens_seen": 6768,
  "num_train_epochs": 3,
  "save_steps": 1,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 17979813789696.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}