{
  "best_global_step": 200,
  "best_metric": 0.8899220824241638,
  "best_model_checkpoint": "./Qwen3-8B-Deview-Finetune-v3/checkpoint-200",
  "epoch": 3.0,
  "eval_steps": 100,
  "global_step": 213,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.2387383307516575,
      "epoch": 0.35587188612099646,
      "grad_norm": 0.2482031136751175,
      "learning_rate": 0.00019665805565657603,
      "loss": 1.4302,
      "mean_token_accuracy": 0.6765058571100235,
      "num_tokens": 359256.0,
      "step": 25
    },
    {
      "entropy": 1.024711874872446,
      "epoch": 0.7117437722419929,
      "grad_norm": 0.18657556176185608,
      "learning_rate": 0.00018017857708534107,
      "loss": 1.0322,
      "mean_token_accuracy": 0.7361943054199219,
      "num_tokens": 723580.0,
      "step": 50
    },
    {
      "entropy": 0.9622754953878442,
      "epoch": 1.0569395017793595,
      "grad_norm": 0.17924001812934875,
      "learning_rate": 0.00015218482555779165,
      "loss": 0.9524,
      "mean_token_accuracy": 0.7510290666646564,
      "num_tokens": 1079672.0,
      "step": 75
    },
    {
      "entropy": 0.9302532352507115,
      "epoch": 1.4128113879003559,
      "grad_norm": 0.21390686929225922,
      "learning_rate": 0.00011669692350409223,
      "loss": 0.9225,
      "mean_token_accuracy": 0.7568456518650055,
      "num_tokens": 1439597.0,
      "step": 100
    },
    {
      "epoch": 1.4128113879003559,
      "eval_entropy": 0.9099734690454271,
      "eval_loss": 0.9155102372169495,
      "eval_mean_token_accuracy": 0.7592034753825929,
      "eval_num_tokens": 1439597.0,
      "eval_runtime": 93.9465,
      "eval_samples_per_second": 3.002,
      "eval_steps_per_second": 0.383,
      "step": 100
    },
    {
      "entropy": 0.9195852947235107,
      "epoch": 1.7686832740213523,
      "grad_norm": 0.21934926509857178,
      "learning_rate": 7.881121218829787e-05,
      "loss": 0.9131,
      "mean_token_accuracy": 0.75831638738513,
      "num_tokens": 1805353.0,
      "step": 125
    },
    {
      "entropy": 0.895305324921903,
      "epoch": 2.113879003558719,
      "grad_norm": 0.23493237793445587,
      "learning_rate": 4.3968377103865024e-05,
      "loss": 0.8869,
      "mean_token_accuracy": 0.7629970355132192,
      "num_tokens": 2153810.0,
      "step": 150
    },
    {
      "entropy": 0.8703042688965797,
      "epoch": 2.469750889679715,
      "grad_norm": 0.2501010298728943,
      "learning_rate": 1.7172122835211337e-05,
      "loss": 0.8631,
      "mean_token_accuracy": 0.768850556910038,
      "num_tokens": 2521409.0,
      "step": 175
    },
    {
      "entropy": 0.8889072737097741,
      "epoch": 2.8256227758007118,
      "grad_norm": 0.22853726148605347,
      "learning_rate": 2.2706018074875045e-06,
      "loss": 0.8853,
      "mean_token_accuracy": 0.762938199788332,
      "num_tokens": 2888143.0,
      "step": 200
    },
    {
      "epoch": 2.8256227758007118,
      "eval_entropy": 0.8701072997517056,
      "eval_loss": 0.8899220824241638,
      "eval_mean_token_accuracy": 0.7636316882239448,
      "eval_num_tokens": 2888143.0,
      "eval_runtime": 93.9318,
      "eval_samples_per_second": 3.002,
      "eval_steps_per_second": 0.383,
      "step": 200
    }
  ],
  "logging_steps": 25,
  "max_steps": 213,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3925707456359629e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}