{
  "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": 32.49287155827833,
      "learning_rate": 2.6315789473684208e-08,
      "logits/chosen": -2.453125,
      "logits/rejected": -1.4453125,
      "logps/chosen": -35.0,
      "logps/rejected": -71.5,
      "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": 35.56440458755336,
      "learning_rate": 2.631578947368421e-07,
      "logits/chosen": -1.890625,
      "logits/rejected": -0.7421875,
      "logps/chosen": -65.5,
      "logps/rejected": -58.75,
      "loss": 0.6871,
      "rewards/accuracies": 0.4583333432674408,
      "rewards/chosen": 0.00341796875,
      "rewards/margins": 0.0299072265625,
      "rewards/rejected": -0.0264892578125,
      "step": 10
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 27.925399110912338,
      "learning_rate": 4.970588235294118e-07,
      "logits/chosen": -1.640625,
      "logits/rejected": -0.69140625,
      "logps/chosen": -52.25,
      "logps/rejected": -74.5,
      "loss": 0.7455,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.00579833984375,
      "rewards/margins": 0.0537109375,
      "rewards/rejected": -0.0478515625,
      "step": 20
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 21.690722291304077,
      "learning_rate": 4.676470588235294e-07,
      "logits/chosen": -0.92578125,
      "logits/rejected": -1.0859375,
      "logps/chosen": -62.75,
      "logps/rejected": -57.75,
      "loss": 0.6262,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.10693359375,
      "rewards/margins": 0.1513671875,
      "rewards/rejected": -0.044189453125,
      "step": 30
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 27.607029623598933,
      "learning_rate": 4.38235294117647e-07,
      "logits/chosen": -1.5390625,
      "logits/rejected": -0.6015625,
      "logps/chosen": -52.0,
      "logps/rejected": -97.5,
      "loss": 0.6431,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.08740234375,
      "rewards/margins": 1.046875,
      "rewards/rejected": -0.9609375,
      "step": 40
    },
    {
      "epoch": 0.7936507936507936,
      "grad_norm": 89.33304447908654,
      "learning_rate": 4.0882352941176465e-07,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.0625,
      "logps/chosen": -75.0,
      "logps/rejected": -74.0,
      "loss": 0.742,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.2197265625,
      "rewards/margins": 0.0859375,
      "rewards/rejected": -0.306640625,
      "step": 50
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 28.864344086826605,
      "learning_rate": 3.7941176470588235e-07,
      "logits/chosen": -1.5,
      "logits/rejected": -0.96875,
      "logps/chosen": -59.75,
      "logps/rejected": -84.0,
      "loss": 0.6597,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.205078125,
      "rewards/margins": 0.478515625,
      "rewards/rejected": -0.2734375,
      "step": 60
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.3671875,
      "eval_logits/rejected": -0.71875,
      "eval_logps/chosen": -61.25,
      "eval_logps/rejected": -89.5,
      "eval_loss": 0.684765636920929,
      "eval_rewards/accuracies": 0.625,
      "eval_rewards/chosen": 0.0400390625,
      "eval_rewards/margins": 0.4375,
      "eval_rewards/rejected": -0.3984375,
      "eval_runtime": 18.8729,
      "eval_samples_per_second": 10.597,
      "eval_steps_per_second": 0.371,
      "step": 63
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 20.770261653452945,
      "learning_rate": 3.5e-07,
      "logits/chosen": -1.6171875,
      "logits/rejected": -0.87109375,
      "logps/chosen": -62.25,
      "logps/rejected": -85.0,
      "loss": 0.3381,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.578125,
      "rewards/margins": 1.953125,
      "rewards/rejected": -1.375,
      "step": 70
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 18.43753108370381,
      "learning_rate": 3.205882352941177e-07,
      "logits/chosen": -1.34375,
      "logits/rejected": -0.84375,
      "logps/chosen": -49.5,
      "logps/rejected": -99.0,
      "loss": 0.3254,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.3671875,
      "rewards/margins": 2.390625,
      "rewards/rejected": -2.015625,
      "step": 80
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 16.839062370433073,
      "learning_rate": 2.911764705882353e-07,
      "logits/chosen": -1.5078125,
      "logits/rejected": -0.78515625,
      "logps/chosen": -57.5,
      "logps/rejected": -72.5,
      "loss": 0.2715,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.84765625,
      "rewards/margins": 2.03125,
      "rewards/rejected": -1.1875,
      "step": 90
    },
    {
      "epoch": 1.5873015873015874,
      "grad_norm": 16.580044861649466,
      "learning_rate": 2.6176470588235295e-07,
      "logits/chosen": -2.09375,
      "logits/rejected": -1.3359375,
      "logps/chosen": -51.0,
      "logps/rejected": -99.5,
      "loss": 0.2519,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 0.345703125,
      "rewards/margins": 1.9375,
      "rewards/rejected": -1.59375,
      "step": 100
    },
    {
      "epoch": 1.746031746031746,
      "grad_norm": 16.88214421094986,
      "learning_rate": 2.323529411764706e-07,
      "logits/chosen": -2.3125,
      "logits/rejected": -1.171875,
      "logps/chosen": -55.75,
      "logps/rejected": -90.5,
      "loss": 0.2389,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.197265625,
      "rewards/margins": 1.9296875,
      "rewards/rejected": -1.734375,
      "step": 110
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 12.829485946331816,
      "learning_rate": 2.0294117647058823e-07,
      "logits/chosen": -1.8828125,
      "logits/rejected": -0.9453125,
      "logps/chosen": -50.5,
      "logps/rejected": -102.5,
      "loss": 0.2831,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.10888671875,
      "rewards/margins": 2.578125,
      "rewards/rejected": -2.46875,
      "step": 120
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.84375,
      "eval_logits/rejected": -1.171875,
      "eval_logps/chosen": -66.0,
      "eval_logps/rejected": -97.5,
      "eval_loss": 0.6284375190734863,
      "eval_rewards/accuracies": 0.6071428656578064,
      "eval_rewards/chosen": -0.453125,
      "eval_rewards/margins": 0.75,
      "eval_rewards/rejected": -1.203125,
      "eval_runtime": 23.6296,
      "eval_samples_per_second": 8.464,
      "eval_steps_per_second": 0.296,
      "step": 126
    },
    {
      "epoch": 2.0634920634920633,
      "grad_norm": 7.418629403738186,
      "learning_rate": 1.7352941176470587e-07,
      "logits/chosen": -2.109375,
      "logits/rejected": -0.73046875,
      "logps/chosen": -61.25,
      "logps/rejected": -106.0,
      "loss": 0.1749,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.4296875,
      "rewards/margins": 2.625,
      "rewards/rejected": -2.203125,
      "step": 130
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 5.917946365968759,
      "learning_rate": 1.441176470588235e-07,
      "logits/chosen": -2.390625,
      "logits/rejected": -1.2890625,
      "logps/chosen": -72.0,
      "logps/rejected": -100.0,
      "loss": 0.1132,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.6015625,
      "rewards/margins": 3.65625,
      "rewards/rejected": -3.0625,
      "step": 140
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 7.62179839987552,
      "learning_rate": 1.1470588235294116e-07,
      "logits/chosen": -2.125,
      "logits/rejected": -1.296875,
      "logps/chosen": -54.0,
      "logps/rejected": -92.5,
      "loss": 0.125,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.66796875,
      "rewards/margins": 3.171875,
      "rewards/rejected": -2.5,
      "step": 150
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 7.689817277750616,
      "learning_rate": 8.529411764705883e-08,
      "logits/chosen": -1.8671875,
      "logits/rejected": -2.3125,
      "logps/chosen": -50.25,
      "logps/rejected": -92.5,
      "loss": 0.118,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.3203125,
      "rewards/margins": 3.109375,
      "rewards/rejected": -2.78125,
      "step": 160
    },
    {
      "epoch": 2.6984126984126986,
      "grad_norm": 5.583079792162674,
      "learning_rate": 5.5882352941176474e-08,
      "logits/chosen": -2.21875,
      "logits/rejected": -1.1328125,
      "logps/chosen": -66.0,
      "logps/rejected": -114.5,
      "loss": 0.145,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.388671875,
      "rewards/margins": 3.671875,
      "rewards/rejected": -3.28125,
      "step": 170
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 10.79399873370834,
      "learning_rate": 2.6470588235294116e-08,
      "logits/chosen": -1.6640625,
      "logits/rejected": -1.5234375,
      "logps/chosen": -59.0,
      "logps/rejected": -87.5,
      "loss": 0.1329,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.310546875,
      "rewards/margins": 2.75,
      "rewards/rejected": -2.4375,
      "step": 180
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -2.21875,
      "eval_logits/rejected": -1.4921875,
      "eval_logps/chosen": -69.5,
      "eval_logps/rejected": -102.5,
      "eval_loss": 0.6340625286102295,
      "eval_rewards/accuracies": 0.6607142686843872,
      "eval_rewards/chosen": -0.76953125,
      "eval_rewards/margins": 0.921875,
      "eval_rewards/rejected": -1.6953125,
      "eval_runtime": 21.8558,
      "eval_samples_per_second": 9.151,
      "eval_steps_per_second": 0.32,
      "step": 189
    },
    {
      "epoch": 3.0,
      "step": 189,
      "total_flos": 0.0,
      "train_loss": 0.354867279214203,
      "train_runtime": 2422.2159,
      "train_samples_per_second": 2.477,
      "train_steps_per_second": 0.078
    }
  ],
  "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
}