{
  "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": 2.956210802841563,
      "learning_rate": 2.6315789473684208e-08,
      "logits/chosen": -0.84765625,
      "logits/rejected": -0.84765625,
      "logps/chosen": -14.875,
      "logps/rejected": -14.875,
      "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": 0.0026515697903897284,
      "learning_rate": 2.631578947368421e-07,
      "logits/chosen": -0.8828125,
      "logits/rejected": -0.88671875,
      "logps/chosen": -15.9375,
      "logps/rejected": -15.8125,
      "loss": 0.6914,
      "rewards/accuracies": 0.02777777798473835,
      "rewards/chosen": 0.0008697509765625,
      "rewards/margins": -0.00043487548828125,
      "rewards/rejected": 0.00130462646484375,
      "step": 10
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 2.201768075236467,
      "learning_rate": 4.970588235294118e-07,
      "logits/chosen": -0.890625,
      "logits/rejected": -0.88671875,
      "logps/chosen": -15.625,
      "logps/rejected": -15.5625,
      "loss": 0.6894,
      "rewards/accuracies": 0.07500000298023224,
      "rewards/chosen": 0.0228271484375,
      "rewards/margins": 0.00860595703125,
      "rewards/rejected": 0.01422119140625,
      "step": 20
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 2.7819271307964892,
      "learning_rate": 4.676470588235294e-07,
      "logits/chosen": -0.890625,
      "logits/rejected": -0.875,
      "logps/chosen": -14.9375,
      "logps/rejected": -14.75,
      "loss": 0.689,
      "rewards/accuracies": 0.0625,
      "rewards/chosen": 0.0830078125,
      "rewards/margins": 0.0111083984375,
      "rewards/rejected": 0.07177734375,
      "step": 30
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 2.468434193598462,
      "learning_rate": 4.38235294117647e-07,
      "logits/chosen": -0.8984375,
      "logits/rejected": -0.91015625,
      "logps/chosen": -14.375,
      "logps/rejected": -14.8125,
      "loss": 0.6849,
      "rewards/accuracies": 0.11249999701976776,
      "rewards/chosen": 0.1552734375,
      "rewards/margins": 0.038330078125,
      "rewards/rejected": 0.11669921875,
      "step": 40
    },
    {
      "epoch": 0.7936507936507936,
      "grad_norm": 3.217396320056085,
      "learning_rate": 4.0882352941176465e-07,
      "logits/chosen": -0.87109375,
      "logits/rejected": -0.8828125,
      "logps/chosen": -13.125,
      "logps/rejected": -12.625,
      "loss": 0.6828,
      "rewards/accuracies": 0.03750000149011612,
      "rewards/chosen": 0.267578125,
      "rewards/margins": 0.01129150390625,
      "rewards/rejected": 0.255859375,
      "step": 50
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 3.5576289077982315,
      "learning_rate": 3.7941176470588235e-07,
      "logits/chosen": -0.85546875,
      "logits/rejected": -0.8671875,
      "logps/chosen": -12.125,
      "logps/rejected": -12.625,
      "loss": 0.6772,
      "rewards/accuracies": 0.02500000037252903,
      "rewards/chosen": 0.3359375,
      "rewards/margins": 0.0196533203125,
      "rewards/rejected": 0.31640625,
      "step": 60
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -0.875,
      "eval_logits/rejected": -0.87109375,
      "eval_logps/chosen": -10.25,
      "eval_logps/rejected": -10.625,
      "eval_loss": 0.6818749904632568,
      "eval_rewards/accuracies": 0.0357142873108387,
      "eval_rewards/chosen": 0.458984375,
      "eval_rewards/margins": -0.0004177093505859375,
      "eval_rewards/rejected": 0.4609375,
      "eval_runtime": 12.3547,
      "eval_samples_per_second": 16.188,
      "eval_steps_per_second": 0.567,
      "step": 63
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.6480653257577673,
      "learning_rate": 3.5e-07,
      "logits/chosen": -0.83984375,
      "logits/rejected": -0.8515625,
      "logps/chosen": -9.8125,
      "logps/rejected": -11.0,
      "loss": 0.6654,
      "rewards/accuracies": 0.0625,
      "rewards/chosen": 0.52734375,
      "rewards/margins": 0.05859375,
      "rewards/rejected": 0.466796875,
      "step": 70
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 1.8542768080035792,
      "learning_rate": 3.205882352941177e-07,
      "logits/chosen": -0.828125,
      "logits/rejected": -0.83203125,
      "logps/chosen": -10.1875,
      "logps/rejected": -11.0,
      "loss": 0.6656,
      "rewards/accuracies": 0.03750000149011612,
      "rewards/chosen": 0.49609375,
      "rewards/margins": 0.05419921875,
      "rewards/rejected": 0.44140625,
      "step": 80
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 1.4880802722309492,
      "learning_rate": 2.911764705882353e-07,
      "logits/chosen": -0.83984375,
      "logits/rejected": -0.84765625,
      "logps/chosen": -10.6875,
      "logps/rejected": -11.5,
      "loss": 0.6672,
      "rewards/accuracies": 0.10000000149011612,
      "rewards/chosen": 0.474609375,
      "rewards/margins": 0.10400390625,
      "rewards/rejected": 0.37109375,
      "step": 90
    },
    {
      "epoch": 1.5873015873015874,
      "grad_norm": 0.9541857190755878,
      "learning_rate": 2.6176470588235295e-07,
      "logits/chosen": -0.84765625,
      "logits/rejected": -0.84765625,
      "logps/chosen": -12.3125,
      "logps/rejected": -13.875,
      "loss": 0.667,
      "rewards/accuracies": 0.08749999850988388,
      "rewards/chosen": 0.34765625,
      "rewards/margins": 0.1240234375,
      "rewards/rejected": 0.224609375,
      "step": 100
    },
    {
      "epoch": 1.746031746031746,
      "grad_norm": 1.3739568831557003,
      "learning_rate": 2.323529411764706e-07,
      "logits/chosen": -0.875,
      "logits/rejected": -0.8671875,
      "logps/chosen": -12.0,
      "logps/rejected": -13.375,
      "loss": 0.6519,
      "rewards/accuracies": 0.10000000149011612,
      "rewards/chosen": 0.419921875,
      "rewards/margins": 0.1884765625,
      "rewards/rejected": 0.23046875,
      "step": 110
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 1.2013802625805676,
      "learning_rate": 2.0294117647058823e-07,
      "logits/chosen": -0.8359375,
      "logits/rejected": -0.84375,
      "logps/chosen": -12.25,
      "logps/rejected": -13.4375,
      "loss": 0.6564,
      "rewards/accuracies": 0.07500000298023224,
      "rewards/chosen": 0.36328125,
      "rewards/margins": 0.10498046875,
      "rewards/rejected": 0.2578125,
      "step": 120
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -0.85546875,
      "eval_logits/rejected": -0.85546875,
      "eval_logps/chosen": -11.9375,
      "eval_logps/rejected": -12.5,
      "eval_loss": 0.6790624856948853,
      "eval_rewards/accuracies": 0.0357142873108387,
      "eval_rewards/chosen": 0.2890625,
      "eval_rewards/margins": 0.01324462890625,
      "eval_rewards/rejected": 0.275390625,
      "eval_runtime": 14.8138,
      "eval_samples_per_second": 13.501,
      "eval_steps_per_second": 0.473,
      "step": 126
    },
    {
      "epoch": 2.0634920634920633,
      "grad_norm": 1.6810359268654005,
      "learning_rate": 1.7352941176470587e-07,
      "logits/chosen": -0.8359375,
      "logits/rejected": -0.8359375,
      "logps/chosen": -13.0,
      "logps/rejected": -15.0,
      "loss": 0.6466,
      "rewards/accuracies": 0.08749999850988388,
      "rewards/chosen": 0.255859375,
      "rewards/margins": 0.1630859375,
      "rewards/rejected": 0.09228515625,
      "step": 130
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 1.6049829222017113,
      "learning_rate": 1.441176470588235e-07,
      "logits/chosen": -0.83984375,
      "logits/rejected": -0.83984375,
      "logps/chosen": -13.1875,
      "logps/rejected": -13.0,
      "loss": 0.6546,
      "rewards/accuracies": 0.08749999850988388,
      "rewards/chosen": 0.30859375,
      "rewards/margins": 0.1064453125,
      "rewards/rejected": 0.203125,
      "step": 140
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 0.8745967010917558,
      "learning_rate": 1.1470588235294116e-07,
      "logits/chosen": -0.85546875,
      "logits/rejected": -0.84375,
      "logps/chosen": -12.1875,
      "logps/rejected": -13.1875,
      "loss": 0.647,
      "rewards/accuracies": 0.08749999850988388,
      "rewards/chosen": 0.28125,
      "rewards/margins": 0.099609375,
      "rewards/rejected": 0.1806640625,
      "step": 150
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 2.5161816747376817,
      "learning_rate": 8.529411764705883e-08,
      "logits/chosen": -0.87109375,
      "logits/rejected": -0.87890625,
      "logps/chosen": -13.625,
      "logps/rejected": -15.3125,
      "loss": 0.6455,
      "rewards/accuracies": 0.08749999850988388,
      "rewards/chosen": 0.2734375,
      "rewards/margins": 0.1630859375,
      "rewards/rejected": 0.11083984375,
      "step": 160
    },
    {
      "epoch": 2.6984126984126986,
      "grad_norm": 1.779552803700363,
      "learning_rate": 5.5882352941176474e-08,
      "logits/chosen": -0.8203125,
      "logits/rejected": -0.83203125,
      "logps/chosen": -12.1875,
      "logps/rejected": -13.75,
      "loss": 0.6576,
      "rewards/accuracies": 0.05000000074505806,
      "rewards/chosen": 0.2890625,
      "rewards/margins": 0.11328125,
      "rewards/rejected": 0.17578125,
      "step": 170
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 1.0681708171714699,
      "learning_rate": 2.6470588235294116e-08,
      "logits/chosen": -0.80859375,
      "logits/rejected": -0.796875,
      "logps/chosen": -13.0625,
      "logps/rejected": -14.375,
      "loss": 0.6566,
      "rewards/accuracies": 0.07500000298023224,
      "rewards/chosen": 0.2294921875,
      "rewards/margins": 0.09619140625,
      "rewards/rejected": 0.1328125,
      "step": 180
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -0.85546875,
      "eval_logits/rejected": -0.8515625,
      "eval_logps/chosen": -12.375,
      "eval_logps/rejected": -12.9375,
      "eval_loss": 0.6785937547683716,
      "eval_rewards/accuracies": 0.0357142873108387,
      "eval_rewards/chosen": 0.2470703125,
      "eval_rewards/margins": 0.016845703125,
      "eval_rewards/rejected": 0.23046875,
      "eval_runtime": 14.6047,
      "eval_samples_per_second": 13.694,
      "eval_steps_per_second": 0.479,
      "step": 189
    },
    {
      "epoch": 3.0,
      "step": 189,
      "total_flos": 0.0,
      "train_loss": 0.6665452422288359,
      "train_runtime": 1664.0362,
      "train_samples_per_second": 3.606,
      "train_steps_per_second": 0.114
    }
  ],
  "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
}