{
  "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": 8.48122634477978,
      "learning_rate": 2.6315789473684208e-08,
      "logits/chosen": -1.015625,
      "logits/rejected": -1.390625,
      "logps/chosen": -45.5,
      "logps/rejected": -80.5,
      "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": 7.666414237190791,
      "learning_rate": 2.631578947368421e-07,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.21875,
      "logps/chosen": -55.0,
      "logps/rejected": -57.0,
      "loss": 0.6925,
      "rewards/accuracies": 0.0833333358168602,
      "rewards/chosen": -0.00017452239990234375,
      "rewards/margins": -0.00244140625,
      "rewards/rejected": 0.00225830078125,
      "step": 10
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 6.673773390918632,
      "learning_rate": 4.970588235294118e-07,
      "logits/chosen": -1.09375,
      "logits/rejected": -1.21875,
      "logps/chosen": -53.75,
      "logps/rejected": -56.75,
      "loss": 0.69,
      "rewards/accuracies": 0.20000000298023224,
      "rewards/chosen": -0.01470947265625,
      "rewards/margins": 0.0024871826171875,
      "rewards/rejected": -0.0172119140625,
      "step": 20
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 7.379593048881886,
      "learning_rate": 4.676470588235294e-07,
      "logits/chosen": -1.1796875,
      "logits/rejected": -1.1015625,
      "logps/chosen": -60.0,
      "logps/rejected": -57.0,
      "loss": 0.6804,
      "rewards/accuracies": 0.375,
      "rewards/chosen": -0.06689453125,
      "rewards/margins": 0.030029296875,
      "rewards/rejected": -0.09716796875,
      "step": 30
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 9.322532970919411,
      "learning_rate": 4.38235294117647e-07,
      "logits/chosen": -1.3359375,
      "logits/rejected": -1.265625,
      "logps/chosen": -63.25,
      "logps/rejected": -56.5,
      "loss": 0.661,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": -0.0185546875,
      "rewards/margins": 0.06494140625,
      "rewards/rejected": -0.08349609375,
      "step": 40
    },
    {
      "epoch": 0.7936507936507936,
      "grad_norm": 7.879751058232231,
      "learning_rate": 4.0882352941176465e-07,
      "logits/chosen": -1.171875,
      "logits/rejected": -1.21875,
      "logps/chosen": -63.5,
      "logps/rejected": -55.5,
      "loss": 0.6611,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.0830078125,
      "rewards/margins": 0.1220703125,
      "rewards/rejected": -0.0390625,
      "step": 50
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 7.002381076132743,
      "learning_rate": 3.7941176470588235e-07,
      "logits/chosen": -1.25,
      "logits/rejected": -1.2265625,
      "logps/chosen": -62.5,
      "logps/rejected": -56.25,
      "loss": 0.645,
      "rewards/accuracies": 0.36250001192092896,
      "rewards/chosen": 0.1005859375,
      "rewards/margins": 0.12451171875,
      "rewards/rejected": -0.024169921875,
      "step": 60
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.2578125,
      "eval_logits/rejected": -1.171875,
      "eval_logps/chosen": -65.0,
      "eval_logps/rejected": -57.25,
      "eval_loss": 0.6340625286102295,
      "eval_rewards/accuracies": 0.3928571343421936,
      "eval_rewards/chosen": 0.02734375,
      "eval_rewards/margins": 0.1787109375,
      "eval_rewards/rejected": -0.1513671875,
      "eval_runtime": 12.2628,
      "eval_samples_per_second": 16.309,
      "eval_steps_per_second": 0.571,
      "step": 63
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 7.403427316810219,
      "learning_rate": 3.5e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.2109375,
      "logps/chosen": -58.75,
      "logps/rejected": -57.75,
      "loss": 0.6193,
      "rewards/accuracies": 0.36250001192092896,
      "rewards/chosen": 0.044677734375,
      "rewards/margins": 0.1630859375,
      "rewards/rejected": -0.119140625,
      "step": 70
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 7.698447091140595,
      "learning_rate": 3.205882352941177e-07,
      "logits/chosen": -1.1015625,
      "logits/rejected": -1.1640625,
      "logps/chosen": -56.5,
      "logps/rejected": -57.75,
      "loss": 0.5965,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.035400390625,
      "rewards/margins": 0.232421875,
      "rewards/rejected": -0.26953125,
      "step": 80
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 5.644436155568419,
      "learning_rate": 2.911764705882353e-07,
      "logits/chosen": -1.25,
      "logits/rejected": -1.234375,
      "logps/chosen": -57.25,
      "logps/rejected": -58.75,
      "loss": 0.5885,
      "rewards/accuracies": 0.42500001192092896,
      "rewards/chosen": 0.0103759765625,
      "rewards/margins": 0.29296875,
      "rewards/rejected": -0.283203125,
      "step": 90
    },
    {
      "epoch": 1.5873015873015874,
      "grad_norm": 7.133267052353533,
      "learning_rate": 2.6176470588235295e-07,
      "logits/chosen": -1.15625,
      "logits/rejected": -1.296875,
      "logps/chosen": -60.5,
      "logps/rejected": -61.5,
      "loss": 0.5796,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": -0.025146484375,
      "rewards/margins": 0.3671875,
      "rewards/rejected": -0.392578125,
      "step": 100
    },
    {
      "epoch": 1.746031746031746,
      "grad_norm": 3.8242948995128176,
      "learning_rate": 2.323529411764706e-07,
      "logits/chosen": -1.0234375,
      "logits/rejected": -1.171875,
      "logps/chosen": -61.75,
      "logps/rejected": -62.5,
      "loss": 0.5951,
      "rewards/accuracies": 0.42500001192092896,
      "rewards/chosen": -0.0732421875,
      "rewards/margins": 0.318359375,
      "rewards/rejected": -0.390625,
      "step": 110
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 5.60510875244782,
      "learning_rate": 2.0294117647058823e-07,
      "logits/chosen": -1.1328125,
      "logits/rejected": -1.2265625,
      "logps/chosen": -61.5,
      "logps/rejected": -67.0,
      "loss": 0.5691,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.12451171875,
      "rewards/margins": 0.35546875,
      "rewards/rejected": -0.478515625,
      "step": 120
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.21875,
      "eval_logits/rejected": -1.109375,
      "eval_logps/chosen": -66.5,
      "eval_logps/rejected": -60.0,
      "eval_loss": 0.6115624904632568,
      "eval_rewards/accuracies": 0.3928571343421936,
      "eval_rewards/chosen": -0.11279296875,
      "eval_rewards/margins": 0.30078125,
      "eval_rewards/rejected": -0.4140625,
      "eval_runtime": 14.9499,
      "eval_samples_per_second": 13.378,
      "eval_steps_per_second": 0.468,
      "step": 126
    },
    {
      "epoch": 2.0634920634920633,
      "grad_norm": 4.8582765968746875,
      "learning_rate": 1.7352941176470587e-07,
      "logits/chosen": -1.265625,
      "logits/rejected": -1.1875,
      "logps/chosen": -59.5,
      "logps/rejected": -65.0,
      "loss": 0.5577,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": -0.002899169921875,
      "rewards/margins": 0.408203125,
      "rewards/rejected": -0.41015625,
      "step": 130
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 3.9951840250758166,
      "learning_rate": 1.441176470588235e-07,
      "logits/chosen": -1.171875,
      "logits/rejected": -1.15625,
      "logps/chosen": -57.75,
      "logps/rejected": -58.0,
      "loss": 0.5505,
      "rewards/accuracies": 0.36250001192092896,
      "rewards/chosen": -0.052978515625,
      "rewards/margins": 0.33203125,
      "rewards/rejected": -0.38671875,
      "step": 140
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 6.94549887559355,
      "learning_rate": 1.1470588235294116e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.109375,
      "logps/chosen": -66.0,
      "logps/rejected": -57.75,
      "loss": 0.5602,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.048095703125,
      "rewards/margins": 0.3984375,
      "rewards/rejected": -0.4453125,
      "step": 150
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 4.543691810100308,
      "learning_rate": 8.529411764705883e-08,
      "logits/chosen": -1.1484375,
      "logits/rejected": -1.3046875,
      "logps/chosen": -61.25,
      "logps/rejected": -65.0,
      "loss": 0.5427,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.11962890625,
      "rewards/margins": 0.53515625,
      "rewards/rejected": -0.65234375,
      "step": 160
    },
    {
      "epoch": 2.6984126984126986,
      "grad_norm": 8.364914220576363,
      "learning_rate": 5.5882352941176474e-08,
      "logits/chosen": -1.171875,
      "logits/rejected": -1.15625,
      "logps/chosen": -57.0,
      "logps/rejected": -66.0,
      "loss": 0.5554,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": -0.1552734375,
      "rewards/margins": 0.412109375,
      "rewards/rejected": -0.56640625,
      "step": 170
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 5.985657271989901,
      "learning_rate": 2.6470588235294116e-08,
      "logits/chosen": -0.96875,
      "logits/rejected": -1.125,
      "logps/chosen": -61.25,
      "logps/rejected": -62.0,
      "loss": 0.5432,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.047607421875,
      "rewards/margins": 0.48828125,
      "rewards/rejected": -0.53515625,
      "step": 180
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -1.1953125,
      "eval_logits/rejected": -1.0703125,
      "eval_logps/chosen": -67.5,
      "eval_logps/rejected": -61.25,
      "eval_loss": 0.6068750023841858,
      "eval_rewards/accuracies": 0.375,
      "eval_rewards/chosen": -0.2021484375,
      "eval_rewards/margins": 0.34375,
      "eval_rewards/rejected": -0.546875,
      "eval_runtime": 14.9623,
      "eval_samples_per_second": 13.367,
      "eval_steps_per_second": 0.468,
      "step": 189
    },
    {
      "epoch": 3.0,
      "step": 189,
      "total_flos": 0.0,
      "train_loss": 0.601676173941799,
      "train_runtime": 1666.5095,
      "train_samples_per_second": 3.6,
      "train_steps_per_second": 0.113
    }
  ],
  "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
}