{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.971563981042654,
  "eval_steps": 128,
  "global_step": 104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018957345971563982,
      "grad_norm": 154.22432557968577,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 117.53560638427734,
      "logits/rejected": 126.8960952758789,
      "logps/chosen": -335.40118408203125,
      "logps/rejected": -439.16552734375,
      "loss": 0.7085,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.1895734597156398,
      "grad_norm": 173.54279362146667,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 135.02284240722656,
      "logits/rejected": 138.36712646484375,
      "logps/chosen": -395.9000549316406,
      "logps/rejected": -439.2119445800781,
      "loss": 0.7446,
      "rewards/accuracies": 0.4444444477558136,
      "rewards/chosen": 0.004824934061616659,
      "rewards/margins": -0.0009564608335494995,
      "rewards/rejected": 0.005781392101198435,
      "step": 10
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 213.0203224012726,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 122.93214416503906,
      "logits/rejected": 126.6201171875,
      "logps/chosen": -370.6175537109375,
      "logps/rejected": -422.05157470703125,
      "loss": 0.7117,
      "rewards/accuracies": 0.5562499761581421,
      "rewards/chosen": 0.08972366899251938,
      "rewards/margins": 0.19273246824741364,
      "rewards/rejected": -0.10300880670547485,
      "step": 20
    },
    {
      "epoch": 0.5687203791469194,
      "grad_norm": 354.71895126297414,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 145.0334014892578,
      "logits/rejected": 138.62002563476562,
      "logps/chosen": -400.11090087890625,
      "logps/rejected": -439.10186767578125,
      "loss": 0.6564,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.3822835385799408,
      "rewards/margins": 0.5726981163024902,
      "rewards/rejected": -0.9549816846847534,
      "step": 30
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 181.79345422502436,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 127.66829681396484,
      "logits/rejected": 116.15958404541016,
      "logps/chosen": -363.76531982421875,
      "logps/rejected": -384.00177001953125,
      "loss": 0.6169,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.47855812311172485,
      "rewards/margins": 0.8277664184570312,
      "rewards/rejected": -1.3063244819641113,
      "step": 40
    },
    {
      "epoch": 0.9478672985781991,
      "grad_norm": 199.34426276233623,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 143.59988403320312,
      "logits/rejected": 147.75946044921875,
      "logps/chosen": -429.3666076660156,
      "logps/rejected": -502.1111755371094,
      "loss": 0.5891,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.511139452457428,
      "rewards/margins": 0.7781005501747131,
      "rewards/rejected": -1.2892398834228516,
      "step": 50
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 75.80200961720344,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 135.1314239501953,
      "logits/rejected": 138.91156005859375,
      "logps/chosen": -407.7182922363281,
      "logps/rejected": -475.638916015625,
      "loss": 5136693.2,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.4856320321559906,
      "rewards/margins": 1.4780375957489014,
      "rewards/rejected": -1.9636695384979248,
      "step": 60
    },
    {
      "epoch": 1.3270142180094786,
      "grad_norm": 80.25707042426906,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 127.4847183227539,
      "logits/rejected": 129.38381958007812,
      "logps/chosen": -353.80865478515625,
      "logps/rejected": -439.21112060546875,
      "loss": 0.2855,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.6667600870132446,
      "rewards/margins": 1.9516220092773438,
      "rewards/rejected": -1.2848618030548096,
      "step": 70
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 86.1788898089996,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 114.5214614868164,
      "logits/rejected": 131.37765502929688,
      "logps/chosen": -336.91925048828125,
      "logps/rejected": -432.9645080566406,
      "loss": 0.6917,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.8365888595581055,
      "rewards/margins": 1.8229297399520874,
      "rewards/rejected": -0.9863409996032715,
      "step": 80
    },
    {
      "epoch": 1.7061611374407581,
      "grad_norm": 118.44223744236918,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 115.7225112915039,
      "logits/rejected": 128.92230224609375,
      "logps/chosen": -368.44451904296875,
      "logps/rejected": -458.6366271972656,
      "loss": 0.2874,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.8993278741836548,
      "rewards/margins": 1.9470703601837158,
      "rewards/rejected": -1.047742486000061,
      "step": 90
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 76.34921328754892,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 122.6693115234375,
      "logits/rejected": 122.82279968261719,
      "logps/chosen": -359.6036682128906,
      "logps/rejected": -428.95135498046875,
      "loss": 0.2687,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.1185719966888428,
      "rewards/margins": 1.7279402017593384,
      "rewards/rejected": -0.6093680262565613,
      "step": 100
    },
    {
      "epoch": 1.971563981042654,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 528876699520.6585,
      "train_runtime": 2168.869,
      "train_samples_per_second": 6.224,
      "train_steps_per_second": 0.048
    }
  ],
  "logging_steps": 10,
  "max_steps": 104,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}