{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9984,
  "eval_steps": 500,
  "global_step": 156,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.125e-08,
      "logits/chosen": -2.9335875511169434,
      "logits/rejected": -2.708059310913086,
      "logps/chosen": -367.5562744140625,
      "logps/rejected": -316.3052062988281,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.7960729598999023,
      "logits/rejected": -2.7260401248931885,
      "logps/chosen": -274.3499755859375,
      "logps/rejected": -233.05935668945312,
      "loss": 0.6928,
      "rewards/accuracies": 0.4861111044883728,
      "rewards/chosen": 7.348848885158077e-05,
      "rewards/margins": 0.00036154486588202417,
      "rewards/rejected": -0.0002880564716178924,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.989935734988097e-07,
      "logits/chosen": -2.7711944580078125,
      "logits/rejected": -2.701967716217041,
      "logps/chosen": -303.2231140136719,
      "logps/rejected": -258.30401611328125,
      "loss": 0.6874,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.02738715335726738,
      "rewards/margins": 0.014244809746742249,
      "rewards/rejected": 0.013142342679202557,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.877641290737883e-07,
      "logits/chosen": -2.702401638031006,
      "logits/rejected": -2.6644959449768066,
      "logps/chosen": -298.130859375,
      "logps/rejected": -281.53778076171875,
      "loss": 0.6707,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.0661591961979866,
      "rewards/margins": 0.05363141745328903,
      "rewards/rejected": 0.012527775950729847,
      "step": 30
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.646121984004665e-07,
      "logits/chosen": -2.713608741760254,
      "logits/rejected": -2.580498218536377,
      "logps/chosen": -287.88385009765625,
      "logps/rejected": -264.4161376953125,
      "loss": 0.648,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.0008304916555061936,
      "rewards/margins": 0.12442084401845932,
      "rewards/rejected": -0.12359035015106201,
      "step": 40
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.3069871595684787e-07,
      "logits/chosen": -2.5537896156311035,
      "logits/rejected": -2.501384735107422,
      "logps/chosen": -269.0035705566406,
      "logps/rejected": -289.73883056640625,
      "loss": 0.6372,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.021835196763277054,
      "rewards/margins": 0.18140152096748352,
      "rewards/rejected": -0.20323672890663147,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.877242453630256e-07,
      "logits/chosen": -2.577971935272217,
      "logits/rejected": -2.5247650146484375,
      "logps/chosen": -304.6644592285156,
      "logps/rejected": -324.11370849609375,
      "loss": 0.608,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.12026672065258026,
      "rewards/margins": 0.289926141500473,
      "rewards/rejected": -0.4101928174495697,
      "step": 60
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.378437060203357e-07,
      "logits/chosen": -2.5257785320281982,
      "logits/rejected": -2.4621853828430176,
      "logps/chosen": -316.51153564453125,
      "logps/rejected": -312.25164794921875,
      "loss": 0.6138,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.1019580215215683,
      "rewards/margins": 0.32222574949264526,
      "rewards/rejected": -0.42418378591537476,
      "step": 70
    },
    {
      "epoch": 0.51,
      "learning_rate": 2.8355831645441387e-07,
      "logits/chosen": -2.629840135574341,
      "logits/rejected": -2.492844343185425,
      "logps/chosen": -348.1588439941406,
      "logps/rejected": -312.782470703125,
      "loss": 0.5935,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.14375950396060944,
      "rewards/margins": 0.3931245803833008,
      "rewards/rejected": -0.5368840098381042,
      "step": 80
    },
    {
      "epoch": 0.58,
      "learning_rate": 2.2759017277414164e-07,
      "logits/chosen": -2.368922710418701,
      "logits/rejected": -2.346982479095459,
      "logps/chosen": -337.67193603515625,
      "logps/rejected": -351.6241760253906,
      "loss": 0.5782,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.3990400731563568,
      "rewards/margins": 0.3560941815376282,
      "rewards/rejected": -0.7551342844963074,
      "step": 90
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.7274575140626315e-07,
      "logits/chosen": -2.164156675338745,
      "logits/rejected": -2.03346586227417,
      "logps/chosen": -310.364013671875,
      "logps/rejected": -301.4464111328125,
      "loss": 0.5887,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.2869732081890106,
      "rewards/margins": 0.44102102518081665,
      "rewards/rejected": -0.7279942035675049,
      "step": 100
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.2177518064852348e-07,
      "logits/chosen": -1.8856881856918335,
      "logits/rejected": -1.5883822441101074,
      "logps/chosen": -327.2856750488281,
      "logps/rejected": -338.85137939453125,
      "loss": 0.5656,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.3426055610179901,
      "rewards/margins": 0.4826560616493225,
      "rewards/rejected": -0.8252617120742798,
      "step": 110
    },
    {
      "epoch": 0.77,
      "learning_rate": 7.723433775328384e-08,
      "logits/chosen": -1.5672389268875122,
      "logits/rejected": -1.4351329803466797,
      "logps/chosen": -334.88446044921875,
      "logps/rejected": -357.7071838378906,
      "loss": 0.5718,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.46880921721458435,
      "rewards/margins": 0.29977938532829285,
      "rewards/rejected": -0.768588662147522,
      "step": 120
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.1356686569674335e-08,
      "logits/chosen": -1.5653654336929321,
      "logits/rejected": -1.2208584547042847,
      "logps/chosen": -330.20635986328125,
      "logps/rejected": -355.11126708984375,
      "loss": 0.5632,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.34367313981056213,
      "rewards/margins": 0.5312191247940063,
      "rewards/rejected": -0.8748922348022461,
      "step": 130
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.5941282340065697e-08,
      "logits/chosen": -1.5312021970748901,
      "logits/rejected": -1.239989995956421,
      "logps/chosen": -325.05908203125,
      "logps/rejected": -337.0903015136719,
      "loss": 0.5504,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.2957434356212616,
      "rewards/margins": 0.4055696427822113,
      "rewards/rejected": -0.7013131380081177,
      "step": 140
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.2625595580163247e-09,
      "logits/chosen": -1.4374409914016724,
      "logits/rejected": -1.128422498703003,
      "logps/chosen": -334.95135498046875,
      "logps/rejected": -365.4905700683594,
      "loss": 0.5687,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": -0.3784746527671814,
      "rewards/margins": 0.64166659116745,
      "rewards/rejected": -1.020141363143921,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 156,
      "total_flos": 0.0,
      "train_loss": 0.3960995582433847,
      "train_runtime": 6783.1442,
      "train_samples_per_second": 2.948,
      "train_steps_per_second": 0.023
    }
  ],
  "logging_steps": 10,
  "max_steps": 156,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}