{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 189,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015873015873015872,
      "grad_norm": 15.546492141333783,
      "learning_rate": 2.6315789473684208e-08,
      "logits/chosen": -1.4921875,
      "logits/rejected": -1.3828125,
      "logps/chosen": -83.0,
      "logps/rejected": -108.5,
      "loss": 0.6914,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.15873015873015872,
      "grad_norm": 14.774196452895438,
      "learning_rate": 2.631578947368421e-07,
      "logits/chosen": -1.2734375,
      "logits/rejected": -1.1953125,
      "logps/chosen": -107.5,
      "logps/rejected": -90.0,
      "loss": 0.6918,
      "rewards/accuracies": 0.2222222238779068,
      "rewards/chosen": -0.0020751953125,
      "rewards/margins": -0.00139617919921875,
      "rewards/rejected": -0.00069427490234375,
      "step": 10
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 14.3389141291959,
      "learning_rate": 4.970588235294118e-07,
      "logits/chosen": -1.28125,
      "logits/rejected": -1.2578125,
      "logps/chosen": -87.0,
      "logps/rejected": -102.5,
      "loss": 0.6842,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0015411376953125,
      "rewards/margins": 0.021240234375,
      "rewards/rejected": -0.019775390625,
      "step": 20
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 14.661136683478352,
      "learning_rate": 4.676470588235294e-07,
      "logits/chosen": -1.3125,
      "logits/rejected": -1.21875,
      "logps/chosen": -94.5,
      "logps/rejected": -100.0,
      "loss": 0.6687,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.036865234375,
      "rewards/margins": 0.07470703125,
      "rewards/rejected": -0.037841796875,
      "step": 30
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 15.043075717803145,
      "learning_rate": 4.38235294117647e-07,
      "logits/chosen": -1.3359375,
      "logits/rejected": -1.4140625,
      "logps/chosen": -94.0,
      "logps/rejected": -112.5,
      "loss": 0.6159,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.06005859375,
      "rewards/margins": 0.1669921875,
      "rewards/rejected": -0.10693359375,
      "step": 40
    },
    {
      "epoch": 0.7936507936507936,
      "grad_norm": 20.286017694579517,
      "learning_rate": 4.0882352941176465e-07,
      "logits/chosen": -1.359375,
      "logits/rejected": -1.3515625,
      "logps/chosen": -103.0,
      "logps/rejected": -115.0,
      "loss": 0.6568,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.00506591796875,
      "rewards/margins": 0.2099609375,
      "rewards/rejected": -0.2158203125,
      "step": 50
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 15.604577240602536,
      "learning_rate": 3.7941176470588235e-07,
      "logits/chosen": -1.3984375,
      "logits/rejected": -1.421875,
      "logps/chosen": -98.0,
      "logps/rejected": -117.0,
      "loss": 0.6149,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.0322265625,
      "rewards/margins": 0.1865234375,
      "rewards/rejected": -0.154296875,
      "step": 60
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.3359375,
      "eval_logits/rejected": -1.3046875,
      "eval_logps/chosen": -91.0,
      "eval_logps/rejected": -111.5,
      "eval_loss": 0.6164844036102295,
      "eval_rewards/accuracies": 0.7142857313156128,
      "eval_rewards/chosen": 0.007598876953125,
      "eval_rewards/margins": 0.224609375,
      "eval_rewards/rejected": -0.2177734375,
      "eval_runtime": 12.239,
      "eval_samples_per_second": 16.341,
      "eval_steps_per_second": 0.572,
      "step": 63
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 13.062522185911641,
      "learning_rate": 3.5e-07,
      "logits/chosen": -1.375,
      "logits/rejected": -1.375,
      "logps/chosen": -93.0,
      "logps/rejected": -108.0,
      "loss": 0.5617,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.1533203125,
      "rewards/margins": 0.375,
      "rewards/rejected": -0.220703125,
      "step": 70
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 12.793133064546527,
      "learning_rate": 3.205882352941177e-07,
      "logits/chosen": -1.2421875,
      "logits/rejected": -1.359375,
      "logps/chosen": -82.5,
      "logps/rejected": -112.5,
      "loss": 0.5127,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.1455078125,
      "rewards/margins": 0.515625,
      "rewards/rejected": -0.37109375,
      "step": 80
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 13.067935081505256,
      "learning_rate": 2.911764705882353e-07,
      "logits/chosen": -1.484375,
      "logits/rejected": -1.3203125,
      "logps/chosen": -100.0,
      "logps/rejected": -98.0,
      "loss": 0.5069,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.283203125,
      "rewards/margins": 0.5390625,
      "rewards/rejected": -0.2578125,
      "step": 90
    },
    {
      "epoch": 1.5873015873015874,
      "grad_norm": 15.317610443936815,
      "learning_rate": 2.6176470588235295e-07,
      "logits/chosen": -1.3671875,
      "logits/rejected": -1.375,
      "logps/chosen": -90.0,
      "logps/rejected": -111.0,
      "loss": 0.4966,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.2353515625,
      "rewards/margins": 0.5546875,
      "rewards/rejected": -0.318359375,
      "step": 100
    },
    {
      "epoch": 1.746031746031746,
      "grad_norm": 12.447441760307086,
      "learning_rate": 2.323529411764706e-07,
      "logits/chosen": -1.4296875,
      "logits/rejected": -1.296875,
      "logps/chosen": -97.0,
      "logps/rejected": -115.0,
      "loss": 0.4934,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.1435546875,
      "rewards/margins": 0.56640625,
      "rewards/rejected": -0.423828125,
      "step": 110
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 13.19848987425151,
      "learning_rate": 2.0294117647058823e-07,
      "logits/chosen": -1.34375,
      "logits/rejected": -1.34375,
      "logps/chosen": -94.0,
      "logps/rejected": -117.5,
      "loss": 0.4805,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.10205078125,
      "rewards/margins": 0.60546875,
      "rewards/rejected": -0.50390625,
      "step": 120
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.34375,
      "eval_logits/rejected": -1.328125,
      "eval_logps/chosen": -92.0,
      "eval_logps/rejected": -113.0,
      "eval_loss": 0.5889843702316284,
      "eval_rewards/accuracies": 0.6428571343421936,
      "eval_rewards/chosen": -0.0751953125,
      "eval_rewards/margins": 0.31640625,
      "eval_rewards/rejected": -0.392578125,
      "eval_runtime": 14.7924,
      "eval_samples_per_second": 13.52,
      "eval_steps_per_second": 0.473,
      "step": 126
    },
    {
      "epoch": 2.0634920634920633,
      "grad_norm": 10.98673276740895,
      "learning_rate": 1.7352941176470587e-07,
      "logits/chosen": -1.390625,
      "logits/rejected": -1.3984375,
      "logps/chosen": -101.0,
      "logps/rejected": -116.0,
      "loss": 0.4562,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.193359375,
      "rewards/margins": 0.67578125,
      "rewards/rejected": -0.482421875,
      "step": 130
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 9.416416295176981,
      "learning_rate": 1.441176470588235e-07,
      "logits/chosen": -1.4140625,
      "logits/rejected": -1.3828125,
      "logps/chosen": -100.5,
      "logps/rejected": -107.5,
      "loss": 0.4104,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.2099609375,
      "rewards/margins": 0.76171875,
      "rewards/rejected": -0.55078125,
      "step": 140
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 11.858508121730686,
      "learning_rate": 1.1470588235294116e-07,
      "logits/chosen": -1.3671875,
      "logits/rejected": -1.3515625,
      "logps/chosen": -93.5,
      "logps/rejected": -113.5,
      "loss": 0.413,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.2177734375,
      "rewards/margins": 0.7265625,
      "rewards/rejected": -0.51171875,
      "step": 150
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 9.556697117699082,
      "learning_rate": 8.529411764705883e-08,
      "logits/chosen": -1.296875,
      "logits/rejected": -1.4296875,
      "logps/chosen": -90.5,
      "logps/rejected": -108.5,
      "loss": 0.4203,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.1337890625,
      "rewards/margins": 0.82421875,
      "rewards/rejected": -0.69140625,
      "step": 160
    },
    {
      "epoch": 2.6984126984126986,
      "grad_norm": 10.065550572065934,
      "learning_rate": 5.5882352941176474e-08,
      "logits/chosen": -1.3671875,
      "logits/rejected": -1.3671875,
      "logps/chosen": -98.5,
      "logps/rejected": -109.5,
      "loss": 0.4442,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.09814453125,
      "rewards/margins": 0.734375,
      "rewards/rejected": -0.63671875,
      "step": 170
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 11.316830865980743,
      "learning_rate": 2.6470588235294116e-08,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.34375,
      "logps/chosen": -102.0,
      "logps/rejected": -104.5,
      "loss": 0.4264,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.10498046875,
      "rewards/margins": 0.6328125,
      "rewards/rejected": -0.52734375,
      "step": 180
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -1.3515625,
      "eval_logits/rejected": -1.3359375,
      "eval_logps/chosen": -92.5,
      "eval_logps/rejected": -114.0,
      "eval_loss": 0.5811718702316284,
      "eval_rewards/accuracies": 0.6428571343421936,
      "eval_rewards/chosen": -0.1357421875,
      "eval_rewards/margins": 0.34765625,
      "eval_rewards/rejected": -0.484375,
      "eval_runtime": 14.7749,
      "eval_samples_per_second": 13.537,
      "eval_steps_per_second": 0.474,
      "step": 189
    },
    {
      "epoch": 3.0,
      "step": 189,
      "total_flos": 0.0,
      "train_loss": 0.5234840029761905,
      "train_runtime": 1680.5719,
      "train_samples_per_second": 3.57,
      "train_steps_per_second": 0.112
    }
  ],
  "logging_steps": 10,
  "max_steps": 189,
  "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": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}