{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.8666666666666667,
  "global_step": 14,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "logps_train/chosen": -100.49485778808594,
      "logps_train/ref_chosen": -100.5,
      "logps_train/ref_rejected": -105.0,
      "logps_train/rejected": -104.80752563476562,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": -0.020498279482126236,
      "rewards_train/margins": -0.021032828837633133,
      "rewards_train/rejected": 0.000534549355506897,
      "step": 0
    },
    {
      "epoch": 0,
      "logps_train/chosen": -89.90950012207031,
      "logps_train/ref_chosen": -90.0,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -100.86872863769531,
      "rewards_train/accuracies": 0.5,
      "rewards_train/chosen": 0.0191262885928154,
      "rewards_train/margins": 0.0020734891295433044,
      "rewards_train/rejected": 0.017052799463272095,
      "step": 0
    },
    {
      "epoch": 0.13,
      "learning_rate": 5e-05,
      "loss": 0.7043,
      "step": 1
    },
    {
      "epoch": 0.13,
      "logps_train/chosen": -104.38166809082031,
      "logps_train/ref_chosen": -104.5,
      "logps_train/ref_rejected": -98.0,
      "logps_train/rejected": -97.89907836914062,
      "rewards_train/accuracies": 0.515625,
      "rewards_train/chosen": 0.03059956058859825,
      "rewards_train/margins": 0.004916800186038017,
      "rewards_train/rejected": 0.025682760402560234,
      "step": 1
    },
    {
      "epoch": 0.13,
      "logps_train/chosen": -94.88069152832031,
      "logps_train/ref_chosen": -95.0,
      "logps_train/ref_rejected": -95.0,
      "logps_train/rejected": -95.29293823242188,
      "rewards_train/accuracies": 0.546875,
      "rewards_train/chosen": -0.009438544511795044,
      "rewards_train/margins": -0.020688317716121674,
      "rewards_train/rejected": 0.01124977320432663,
      "step": 1
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.92735454356513e-05,
      "loss": 0.7061,
      "step": 2
    },
    {
      "epoch": 0.27,
      "logps_train/chosen": -82.24234771728516,
      "logps_train/ref_chosen": -82.5,
      "logps_train/ref_rejected": -89.0,
      "logps_train/rejected": -88.62696075439453,
      "rewards_train/accuracies": 0.421875,
      "rewards_train/chosen": 0.15220294892787933,
      "rewards_train/margins": -0.038771942257881165,
      "rewards_train/rejected": 0.1909748911857605,
      "step": 2
    },
    {
      "epoch": 0.27,
      "logps_train/chosen": -99.8189926147461,
      "logps_train/ref_chosen": -100.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -105.7823257446289,
      "rewards_train/accuracies": 0.484375,
      "rewards_train/chosen": 0.17058409750461578,
      "rewards_train/margins": 0.055155299603939056,
      "rewards_train/rejected": 0.11542879790067673,
      "step": 2
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.713640064133025e-05,
      "loss": 0.7021,
      "step": 3
    },
    {
      "epoch": 0.4,
      "logps_train/chosen": -85.39012145996094,
      "logps_train/ref_chosen": -86.0,
      "logps_train/ref_rejected": -90.0,
      "logps_train/rejected": -89.733154296875,
      "rewards_train/accuracies": 0.609375,
      "rewards_train/chosen": 0.22778785228729248,
      "rewards_train/margins": 0.07618004083633423,
      "rewards_train/rejected": 0.15160781145095825,
      "step": 3
    },
    {
      "epoch": 0.4,
      "logps_train/chosen": -91.78669738769531,
      "logps_train/ref_chosen": -92.0,
      "logps_train/ref_rejected": -77.5,
      "logps_train/rejected": -77.08824157714844,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": 0.2614349126815796,
      "rewards_train/margins": 0.015637695789337158,
      "rewards_train/rejected": 0.24579721689224243,
      "step": 3
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.371276870427753e-05,
      "loss": 0.701,
      "step": 4
    },
    {
      "epoch": 0.53,
      "logps_train/chosen": -88.01852416992188,
      "logps_train/ref_chosen": -88.5,
      "logps_train/ref_rejected": -91.5,
      "logps_train/rejected": -91.21440124511719,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.29049867391586304,
      "rewards_train/margins": 0.13419204950332642,
      "rewards_train/rejected": 0.15630662441253662,
      "step": 4
    },
    {
      "epoch": 0.53,
      "logps_train/chosen": -103.98135375976562,
      "logps_train/ref_chosen": -105.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -103.89208984375,
      "rewards_train/accuracies": 0.640625,
      "rewards_train/chosen": 0.4727012515068054,
      "rewards_train/margins": 0.21867156028747559,
      "rewards_train/rejected": 0.25402969121932983,
      "step": 4
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.920161866827889e-05,
      "loss": 0.6539,
      "step": 5
    },
    {
      "epoch": 0.67,
      "logps_train/chosen": -108.25979614257812,
      "logps_train/ref_chosen": -109.0,
      "logps_train/ref_rejected": -99.5,
      "logps_train/rejected": -98.85391235351562,
      "rewards_train/accuracies": 0.609375,
      "rewards_train/chosen": 0.5204911231994629,
      "rewards_train/margins": 0.25702032446861267,
      "rewards_train/rejected": 0.2634707987308502,
      "step": 5
    },
    {
      "epoch": 0.67,
      "logps_train/chosen": -81.6476821899414,
      "logps_train/ref_chosen": -82.5,
      "logps_train/ref_rejected": -80.0,
      "logps_train/rejected": -78.95549011230469,
      "rewards_train/accuracies": 0.484375,
      "rewards_train/chosen": 0.4765724837779999,
      "rewards_train/margins": -0.0074152350425720215,
      "rewards_train/rejected": 0.4839877188205719,
      "step": 5
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.386512217606339e-05,
      "loss": 0.7324,
      "step": 6
    },
    {
      "epoch": 0.8,
      "logps_train/chosen": -91.17733001708984,
      "logps_train/ref_chosen": -92.0,
      "logps_train/ref_rejected": -89.5,
      "logps_train/rejected": -88.92269134521484,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": 0.518998384475708,
      "rewards_train/margins": 0.1605202555656433,
      "rewards_train/rejected": 0.3584781289100647,
      "step": 6
    },
    {
      "epoch": 0.8,
      "logps_train/chosen": -87.36846923828125,
      "logps_train/ref_chosen": -88.5,
      "logps_train/ref_rejected": -94.0,
      "logps_train/rejected": -93.19339752197266,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.4690890610218048,
      "rewards_train/margins": 0.18302035331726074,
      "rewards_train/rejected": 0.28606870770454407,
      "step": 6
    },
    {
      "epoch": 0.93,
      "learning_rate": 2.8013417006383076e-05,
      "loss": 0.6995,
      "step": 7
    },
    {
      "epoch": 0.93,
      "logps_train/chosen": -98.05281066894531,
      "logps_train/ref_chosen": -107.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -112.13198852539062,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 4.626056671142578,
      "rewards_train/margins": 8.7623610496521,
      "rewards_train/rejected": -4.1363043785095215,
      "step": 7
    },
    {
      "epoch": 0.93,
      "logps_train/chosen": -88.66700744628906,
      "logps_train/ref_chosen": -96.0,
      "logps_train/ref_rejected": -90.5,
      "logps_train/rejected": -95.32691955566406,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 3.6146154403686523,
      "rewards_train/margins": 6.129823684692383,
      "rewards_train/rejected": -2.5152082443237305,
      "step": 7
    },
    {
      "epoch": 1.07,
      "learning_rate": 2.1986582993616926e-05,
      "loss": 0.1626,
      "step": 8
    },
    {
      "epoch": 1.07,
      "logps_train/chosen": -94.38668060302734,
      "logps_train/ref_chosen": -104.0,
      "logps_train/ref_rejected": -91.0,
      "logps_train/rejected": -98.58480834960938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 4.836201190948486,
      "rewards_train/margins": 8.564030647277832,
      "rewards_train/rejected": -3.7278294563293457,
      "step": 8
    },
    {
      "epoch": 1.07,
      "logps_train/chosen": -82.82890319824219,
      "logps_train/ref_chosen": -90.0,
      "logps_train/ref_rejected": -91.0,
      "logps_train/rejected": -97.17626953125,
      "rewards_train/accuracies": 0.984375,
      "rewards_train/chosen": 3.66611385345459,
      "rewards_train/margins": 6.670999765396118,
      "rewards_train/rejected": -3.0048859119415283,
      "step": 8
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.613487782393661e-05,
      "loss": 0.0834,
      "step": 9
    },
    {
      "epoch": 1.2,
      "logps_train/chosen": -91.4707260131836,
      "logps_train/ref_chosen": -100.0,
      "logps_train/ref_rejected": -91.5,
      "logps_train/rejected": -99.23442077636719,
      "rewards_train/accuracies": 0.984375,
      "rewards_train/chosen": 4.370593070983887,
      "rewards_train/margins": 8.266491174697876,
      "rewards_train/rejected": -3.8958981037139893,
      "step": 9
    },
    {
      "epoch": 1.2,
      "logps_train/chosen": -89.33580780029297,
      "logps_train/ref_chosen": -98.0,
      "logps_train/ref_rejected": -95.5,
      "logps_train/rejected": -101.87576293945312,
      "rewards_train/accuracies": 0.984375,
      "rewards_train/chosen": 4.202215671539307,
      "rewards_train/margins": 7.457237482070923,
      "rewards_train/rejected": -3.255021810531616,
      "step": 9
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.0798381331721109e-05,
      "loss": 0.0654,
      "step": 10
    },
    {
      "epoch": 1.33,
      "logps_train/chosen": -92.18928527832031,
      "logps_train/ref_chosen": -99.5,
      "logps_train/ref_rejected": -97.0,
      "logps_train/rejected": -102.6868896484375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 3.7073636054992676,
      "rewards_train/margins": 6.655301570892334,
      "rewards_train/rejected": -2.9479379653930664,
      "step": 10
    },
    {
      "epoch": 1.33,
      "logps_train/chosen": -83.0275650024414,
      "logps_train/ref_chosen": -91.0,
      "logps_train/ref_rejected": -97.5,
      "logps_train/rejected": -104.69689178466797,
      "rewards_train/accuracies": 0.984375,
      "rewards_train/chosen": 4.168407440185547,
      "rewards_train/margins": 7.665899753570557,
      "rewards_train/rejected": -3.4974923133850098,
      "step": 10
    },
    {
      "epoch": 1.47,
      "learning_rate": 6.28723129572247e-06,
      "loss": 0.0669,
      "step": 11
    },
    {
      "epoch": 1.47,
      "logps_train/chosen": -87.6376724243164,
      "logps_train/ref_chosen": -97.0,
      "logps_train/ref_rejected": -99.0,
      "logps_train/rejected": -106.98680877685547,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 4.543467044830322,
      "rewards_train/margins": 8.686774730682373,
      "rewards_train/rejected": -4.143307685852051,
      "step": 11
    },
    {
      "epoch": 1.47,
      "logps_train/chosen": -73.75891876220703,
      "logps_train/ref_chosen": -81.0,
      "logps_train/ref_rejected": -90.0,
      "logps_train/rejected": -95.56880950927734,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 3.678358554840088,
      "rewards_train/margins": 6.431073427200317,
      "rewards_train/rejected": -2.7527148723602295,
      "step": 11
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.8635993586697553e-06,
      "loss": 0.0643,
      "step": 12
    },
    {
      "epoch": 1.6,
      "logps_train/chosen": -81.66705322265625,
      "logps_train/ref_chosen": -90.0,
      "logps_train/ref_rejected": -96.0,
      "logps_train/rejected": -102.6563720703125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 4.149624824523926,
      "rewards_train/margins": 7.561353445053101,
      "rewards_train/rejected": -3.411728620529175,
      "step": 12
    },
    {
      "epoch": 1.6,
      "logps_train/chosen": -75.94863891601562,
      "logps_train/ref_chosen": -83.0,
      "logps_train/ref_rejected": -85.0,
      "logps_train/rejected": -91.42919158935547,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 3.591111183166504,
      "rewards_train/margins": 6.727336406707764,
      "rewards_train/rejected": -3.1362252235412598,
      "step": 12
    },
    {
      "epoch": 1.73,
      "learning_rate": 7.264545643486997e-07,
      "loss": 0.08,
      "step": 13
    },
    {
      "epoch": 1.73,
      "logps_train/chosen": -94.35868835449219,
      "logps_train/ref_chosen": -105.0,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -109.37545013427734,
      "rewards_train/accuracies": 0.984375,
      "rewards_train/chosen": 5.367044448852539,
      "rewards_train/margins": 9.60921335220337,
      "rewards_train/rejected": -4.24216890335083,
      "step": 13
    },
    {
      "epoch": 1.73,
      "logps_train/chosen": -81.82484436035156,
      "logps_train/ref_chosen": -91.0,
      "logps_train/ref_rejected": -85.0,
      "logps_train/rejected": -91.75001525878906,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 4.351003646850586,
      "rewards_train/margins": 7.681576251983643,
      "rewards_train/rejected": -3.3305726051330566,
      "step": 13
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0,
      "loss": 0.0427,
      "step": 14
    },
    {
      "epoch": 1.87,
      "step": 14,
      "total_flos": 0.0,
      "train_loss": 0.39032512956431936,
      "train_runtime": 164.095,
      "train_samples_per_second": 11.079,
      "train_steps_per_second": 0.085
    }
  ],
  "max_steps": 14,
  "num_train_epochs": 2,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}