{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.986666666666667,
  "eval_steps": 500,
  "global_step": 84,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 4.364832401275635,
      "learning_rate": 4.997807075247147e-06,
      "logits/chosen": -2.3599021434783936,
      "logits/rejected": -2.3986287117004395,
      "logps/chosen": -1.2810232639312744,
      "logps/rejected": -1.6647472381591797,
      "loss": 1.334,
      "odds_ratio_loss": 11.882925987243652,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.12810233235359192,
      "rewards/margins": 0.03837240859866142,
      "rewards/rejected": -0.16647472977638245,
      "sft_loss": 0.14570708572864532,
      "step": 10
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 3.797199010848999,
      "learning_rate": 4.7392794005985324e-06,
      "logits/chosen": -2.176560878753662,
      "logits/rejected": -2.2147622108459473,
      "logps/chosen": -1.1564085483551025,
      "logps/rejected": -1.6085771322250366,
      "loss": 1.2054,
      "odds_ratio_loss": 10.778497695922852,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.11564085632562637,
      "rewards/margins": 0.04521685466170311,
      "rewards/rejected": -0.160857692360878,
      "sft_loss": 0.1275218427181244,
      "step": 20
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 3.668826103210449,
      "learning_rate": 4.093559974371725e-06,
      "logits/chosen": -2.3584961891174316,
      "logits/rejected": -2.3714914321899414,
      "logps/chosen": -1.0279743671417236,
      "logps/rejected": -1.439477562904358,
      "loss": 1.0773,
      "odds_ratio_loss": 9.591609001159668,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.10279743373394012,
      "rewards/margins": 0.041150324046611786,
      "rewards/rejected": -0.1439477652311325,
      "sft_loss": 0.11812162399291992,
      "step": 30
    },
    {
      "epoch": 1.4222222222222223,
      "grad_norm": 2.7432360649108887,
      "learning_rate": 3.1722995515381644e-06,
      "logits/chosen": -2.2854158878326416,
      "logits/rejected": -2.3222243785858154,
      "logps/chosen": -0.8480159640312195,
      "logps/rejected": -1.3128085136413574,
      "loss": 0.8927,
      "odds_ratio_loss": 8.068161010742188,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": -0.08480159938335419,
      "rewards/margins": 0.0464792437851429,
      "rewards/rejected": -0.1312808394432068,
      "sft_loss": 0.08589236438274384,
      "step": 40
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 2.6973206996917725,
      "learning_rate": 2.134792428593971e-06,
      "logits/chosen": -2.3095250129699707,
      "logits/rejected": -2.336907386779785,
      "logps/chosen": -0.6922627687454224,
      "logps/rejected": -1.121113896369934,
      "loss": 0.7399,
      "odds_ratio_loss": 6.732758522033691,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.06922627240419388,
      "rewards/margins": 0.042885102331638336,
      "rewards/rejected": -0.11211137473583221,
      "sft_loss": 0.06660701334476471,
      "step": 50
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 1.8098381757736206,
      "learning_rate": 1.160433012552508e-06,
      "logits/chosen": -2.3899528980255127,
      "logits/rejected": -2.4369609355926514,
      "logps/chosen": -0.5784354209899902,
      "logps/rejected": -1.0445263385772705,
      "loss": 0.63,
      "odds_ratio_loss": 5.705307960510254,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.05784354358911514,
      "rewards/margins": 0.04660910367965698,
      "rewards/rejected": -0.10445265471935272,
      "sft_loss": 0.05942065268754959,
      "step": 60
    },
    {
      "epoch": 2.488888888888889,
      "grad_norm": 2.6440157890319824,
      "learning_rate": 4.1769689822475147e-07,
      "logits/chosen": -2.3521196842193604,
      "logits/rejected": -2.39290189743042,
      "logps/chosen": -0.5662254095077515,
      "logps/rejected": -1.0197746753692627,
      "loss": 0.6132,
      "odds_ratio_loss": 5.592169761657715,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.056622546166181564,
      "rewards/margins": 0.045354925096035004,
      "rewards/rejected": -0.10197745263576508,
      "sft_loss": 0.05396850034594536,
      "step": 70
    },
    {
      "epoch": 2.8444444444444446,
      "grad_norm": 2.6655380725860596,
      "learning_rate": 3.5009907323737826e-08,
      "logits/chosen": -2.3952314853668213,
      "logits/rejected": -2.4219796657562256,
      "logps/chosen": -0.5315676927566528,
      "logps/rejected": -0.9738686680793762,
      "loss": 0.5745,
      "odds_ratio_loss": 5.228327751159668,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.053156763315200806,
      "rewards/margins": 0.044230107218027115,
      "rewards/rejected": -0.09738686680793762,
      "sft_loss": 0.05164428427815437,
      "step": 80
    },
    {
      "epoch": 2.986666666666667,
      "step": 84,
      "total_flos": 2.2783332409344e+16,
      "train_loss": 0.8738818849836077,
      "train_runtime": 261.5721,
      "train_samples_per_second": 5.161,
      "train_steps_per_second": 0.321
    }
  ],
  "logging_steps": 10,
  "max_steps": 84,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2.2783332409344e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}