{
  "best_metric": 1.1112251281738281,
  "best_model_checkpoint": "models/qwen2.5-3b-sft-10/checkpoint-776",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 3880,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002577319587628866,
      "grad_norm": 10.882415929571213,
      "learning_rate": 5.1546391752577325e-08,
      "loss": 1.538,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.1319564580917358,
      "eval_runtime": 18.609,
      "eval_samples_per_second": 74.319,
      "eval_steps_per_second": 1.182,
      "step": 388
    },
    {
      "epoch": 1.2886597938144329,
      "grad_norm": 0.7448965371149199,
      "learning_rate": 1.994927876137743e-05,
      "loss": 1.1917,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.1112251281738281,
      "eval_runtime": 18.2135,
      "eval_samples_per_second": 75.933,
      "eval_steps_per_second": 1.208,
      "step": 776
    },
    {
      "epoch": 2.5773195876288657,
      "grad_norm": 0.6867587126100023,
      "learning_rate": 1.8522168236559693e-05,
      "loss": 0.9549,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.137048363685608,
      "eval_runtime": 18.2225,
      "eval_samples_per_second": 75.895,
      "eval_steps_per_second": 1.207,
      "step": 1164
    },
    {
      "epoch": 3.865979381443299,
      "grad_norm": 0.7401490876847191,
      "learning_rate": 1.539952602786511e-05,
      "loss": 0.755,
      "step": 1500
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.2109209299087524,
      "eval_runtime": 18.2407,
      "eval_samples_per_second": 75.82,
      "eval_steps_per_second": 1.206,
      "step": 1552
    },
    {
      "epoch": 5.0,
      "eval_loss": 1.3351761102676392,
      "eval_runtime": 18.2438,
      "eval_samples_per_second": 75.807,
      "eval_steps_per_second": 1.206,
      "step": 1940
    },
    {
      "epoch": 5.154639175257732,
      "grad_norm": 0.9037511466097634,
      "learning_rate": 1.1202618772525679e-05,
      "loss": 0.5356,
      "step": 2000
    },
    {
      "epoch": 6.0,
      "eval_loss": 1.5199735164642334,
      "eval_runtime": 18.4271,
      "eval_samples_per_second": 75.052,
      "eval_steps_per_second": 1.194,
      "step": 2328
    },
    {
      "epoch": 6.443298969072165,
      "grad_norm": 0.8676017322399043,
      "learning_rate": 6.766443976832518e-06,
      "loss": 0.3548,
      "step": 2500
    },
    {
      "epoch": 7.0,
      "eval_loss": 1.74420166015625,
      "eval_runtime": 18.2732,
      "eval_samples_per_second": 75.685,
      "eval_steps_per_second": 1.204,
      "step": 2716
    },
    {
      "epoch": 7.731958762886598,
      "grad_norm": 0.8393961410905794,
      "learning_rate": 2.9736027245152276e-06,
      "loss": 0.2395,
      "step": 3000
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.9801589250564575,
      "eval_runtime": 18.2678,
      "eval_samples_per_second": 75.707,
      "eval_steps_per_second": 1.204,
      "step": 3104
    },
    {
      "epoch": 9.0,
      "eval_loss": 2.1654229164123535,
      "eval_runtime": 18.4713,
      "eval_samples_per_second": 74.873,
      "eval_steps_per_second": 1.191,
      "step": 3492
    },
    {
      "epoch": 9.02061855670103,
      "grad_norm": 0.6887675413824407,
      "learning_rate": 5.787013946911546e-07,
      "loss": 0.1724,
      "step": 3500
    },
    {
      "epoch": 10.0,
      "eval_loss": 2.2557177543640137,
      "eval_runtime": 18.1106,
      "eval_samples_per_second": 76.364,
      "eval_steps_per_second": 1.215,
      "step": 3880
    },
    {
      "epoch": 10.0,
      "step": 3880,
      "total_flos": 737402935050240.0,
      "train_loss": 0.556043832511017,
      "train_runtime": 12515.2297,
      "train_samples_per_second": 19.797,
      "train_steps_per_second": 0.31
    }
  ],
  "logging_steps": 500,
  "max_steps": 3880,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 737402935050240.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}