{
  "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": 16.2200460330357,
      "learning_rate": 2.6315789473684208e-08,
      "logits/chosen": -1.5234375,
      "logits/rejected": -1.4375,
      "logps/chosen": -166.0,
      "logps/rejected": -129.0,
      "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": 16.23804389503029,
      "learning_rate": 2.631578947368421e-07,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.53125,
      "logps/chosen": -144.0,
      "logps/rejected": -120.0,
      "loss": 0.6946,
      "rewards/accuracies": 0.2361111044883728,
      "rewards/chosen": -0.000339508056640625,
      "rewards/margins": -0.000701904296875,
      "rewards/rejected": 0.0003509521484375,
      "step": 10
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 15.80676425202867,
      "learning_rate": 4.970588235294118e-07,
      "logits/chosen": -1.609375,
      "logits/rejected": -1.5234375,
      "logps/chosen": -162.0,
      "logps/rejected": -133.0,
      "loss": 0.6844,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.026611328125,
      "rewards/margins": 0.0341796875,
      "rewards/rejected": -0.060791015625,
      "step": 20
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 13.971741456622834,
      "learning_rate": 4.676470588235294e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.546875,
      "logps/chosen": -151.0,
      "logps/rejected": -133.0,
      "loss": 0.6601,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.10693359375,
      "rewards/margins": 0.07177734375,
      "rewards/rejected": -0.1787109375,
      "step": 30
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 14.439143279421419,
      "learning_rate": 4.38235294117647e-07,
      "logits/chosen": -1.625,
      "logits/rejected": -1.578125,
      "logps/chosen": -165.0,
      "logps/rejected": -145.0,
      "loss": 0.6307,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.158203125,
      "rewards/margins": 0.14453125,
      "rewards/rejected": -0.302734375,
      "step": 40
    },
    {
      "epoch": 0.7936507936507936,
      "grad_norm": 12.933116251658811,
      "learning_rate": 4.0882352941176465e-07,
      "logits/chosen": -1.5703125,
      "logits/rejected": -1.5078125,
      "logps/chosen": -155.0,
      "logps/rejected": -134.0,
      "loss": 0.6348,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.298828125,
      "rewards/margins": 0.203125,
      "rewards/rejected": -0.50390625,
      "step": 50
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 13.17457815976231,
      "learning_rate": 3.7941176470588235e-07,
      "logits/chosen": -1.5390625,
      "logits/rejected": -1.53125,
      "logps/chosen": -157.0,
      "logps/rejected": -148.0,
      "loss": 0.607,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.390625,
      "rewards/margins": 0.1884765625,
      "rewards/rejected": -0.578125,
      "step": 60
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.609375,
      "eval_logits/rejected": -1.453125,
      "eval_logps/chosen": -161.0,
      "eval_logps/rejected": -117.5,
      "eval_loss": 0.6323873400688171,
      "eval_rewards/accuracies": 0.7250000238418579,
      "eval_rewards/chosen": -0.5625,
      "eval_rewards/margins": 0.2578125,
      "eval_rewards/rejected": -0.8203125,
      "eval_runtime": 8.7111,
      "eval_samples_per_second": 16.875,
      "eval_steps_per_second": 0.574,
      "step": 63
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 10.940163753846868,
      "learning_rate": 3.5e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.5234375,
      "logps/chosen": -151.0,
      "logps/rejected": -126.5,
      "loss": 0.5062,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.4453125,
      "rewards/margins": 0.46875,
      "rewards/rejected": -0.9140625,
      "step": 70
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 10.448217147222541,
      "learning_rate": 3.205882352941177e-07,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.4609375,
      "logps/chosen": -142.0,
      "logps/rejected": -123.5,
      "loss": 0.4673,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.5625,
      "rewards/margins": 0.57421875,
      "rewards/rejected": -1.1328125,
      "step": 80
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 10.288072810717312,
      "learning_rate": 2.911764705882353e-07,
      "logits/chosen": -1.578125,
      "logits/rejected": -1.53125,
      "logps/chosen": -157.0,
      "logps/rejected": -147.0,
      "loss": 0.4288,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.53515625,
      "rewards/margins": 0.6953125,
      "rewards/rejected": -1.2265625,
      "step": 90
    },
    {
      "epoch": 1.5873015873015874,
      "grad_norm": 10.019896435666249,
      "learning_rate": 2.6176470588235295e-07,
      "logits/chosen": -1.609375,
      "logits/rejected": -1.6015625,
      "logps/chosen": -162.0,
      "logps/rejected": -155.0,
      "loss": 0.4419,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.640625,
      "rewards/margins": 0.65625,
      "rewards/rejected": -1.296875,
      "step": 100
    },
    {
      "epoch": 1.746031746031746,
      "grad_norm": 12.080281395370564,
      "learning_rate": 2.323529411764706e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.546875,
      "logps/chosen": -156.0,
      "logps/rejected": -148.0,
      "loss": 0.4641,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.53515625,
      "rewards/margins": 0.79296875,
      "rewards/rejected": -1.328125,
      "step": 110
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 11.329269122728626,
      "learning_rate": 2.0294117647058823e-07,
      "logits/chosen": -1.4921875,
      "logits/rejected": -1.5078125,
      "logps/chosen": -140.0,
      "logps/rejected": -145.0,
      "loss": 0.4385,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.5390625,
      "rewards/margins": 0.63671875,
      "rewards/rejected": -1.171875,
      "step": 120
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.6015625,
      "eval_logits/rejected": -1.453125,
      "eval_logps/chosen": -161.0,
      "eval_logps/rejected": -119.5,
      "eval_loss": 0.6192867755889893,
      "eval_rewards/accuracies": 0.7250000238418579,
      "eval_rewards/chosen": -0.625,
      "eval_rewards/margins": 0.3515625,
      "eval_rewards/rejected": -0.9765625,
      "eval_runtime": 10.8318,
      "eval_samples_per_second": 13.571,
      "eval_steps_per_second": 0.462,
      "step": 126
    },
    {
      "epoch": 2.0634920634920633,
      "grad_norm": 9.801925381761396,
      "learning_rate": 1.7352941176470587e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.4921875,
      "logps/chosen": -159.0,
      "logps/rejected": -144.0,
      "loss": 0.404,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.263671875,
      "rewards/margins": 0.90234375,
      "rewards/rejected": -1.1640625,
      "step": 130
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 10.394699460580199,
      "learning_rate": 1.441176470588235e-07,
      "logits/chosen": -1.5859375,
      "logits/rejected": -1.53125,
      "logps/chosen": -152.0,
      "logps/rejected": -146.0,
      "loss": 0.3731,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.271484375,
      "rewards/margins": 0.96875,
      "rewards/rejected": -1.2421875,
      "step": 140
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 7.8219119833188975,
      "learning_rate": 1.1470588235294116e-07,
      "logits/chosen": -1.5390625,
      "logits/rejected": -1.453125,
      "logps/chosen": -162.0,
      "logps/rejected": -132.0,
      "loss": 0.3477,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.376953125,
      "rewards/margins": 0.90625,
      "rewards/rejected": -1.28125,
      "step": 150
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 8.679287106154373,
      "learning_rate": 8.529411764705883e-08,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.4921875,
      "logps/chosen": -163.0,
      "logps/rejected": -153.0,
      "loss": 0.3479,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.3046875,
      "rewards/margins": 1.125,
      "rewards/rejected": -1.4296875,
      "step": 160
    },
    {
      "epoch": 2.6984126984126986,
      "grad_norm": 8.351083925741897,
      "learning_rate": 5.5882352941176474e-08,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.53125,
      "logps/chosen": -146.0,
      "logps/rejected": -157.0,
      "loss": 0.3607,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.4609375,
      "rewards/margins": 0.9375,
      "rewards/rejected": -1.3984375,
      "step": 170
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 7.683812751753157,
      "learning_rate": 2.6470588235294116e-08,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.546875,
      "logps/chosen": -154.0,
      "logps/rejected": -157.0,
      "loss": 0.3708,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.408203125,
      "rewards/margins": 0.96875,
      "rewards/rejected": -1.375,
      "step": 180
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -1.6015625,
      "eval_logits/rejected": -1.453125,
      "eval_logps/chosen": -161.0,
      "eval_logps/rejected": -120.5,
      "eval_loss": 0.6126700639724731,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": -0.67578125,
      "eval_rewards/margins": 0.42578125,
      "eval_rewards/rejected": -1.1015625,
      "eval_runtime": 13.2152,
      "eval_samples_per_second": 11.124,
      "eval_steps_per_second": 0.378,
      "step": 189
    },
    {
      "epoch": 3.0,
      "step": 189,
      "total_flos": 0.0,
      "train_loss": 0.48406885540674605,
      "train_runtime": 1618.971,
      "train_samples_per_second": 3.706,
      "train_steps_per_second": 0.117
    }
  ],
  "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
}