{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9991836734693877,
  "eval_steps": 500,
  "global_step": 153,
  "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": -2.2019269466400146,
      "logits/rejected": -2.3159804344177246,
      "logps/chosen": -281.70477294921875,
      "logps/rejected": -326.98321533203125,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.07,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.3732211589813232,
      "logits/rejected": -2.3809614181518555,
      "logps/chosen": -289.0758361816406,
      "logps/rejected": -340.1326599121094,
      "loss": 0.6929,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.0009192235302180052,
      "rewards/margins": 0.0015418411931023002,
      "rewards/rejected": -0.0024610646069049835,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.989490450759331e-07,
      "logits/chosen": -2.348583698272705,
      "logits/rejected": -2.323251724243164,
      "logps/chosen": -267.167724609375,
      "logps/rejected": -304.1368103027344,
      "loss": 0.6912,
      "rewards/accuracies": 0.48124998807907104,
      "rewards/chosen": -0.021241169422864914,
      "rewards/margins": -0.005664441268891096,
      "rewards/rejected": -0.015576730482280254,
      "step": 20
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.872270441827174e-07,
      "logits/chosen": -2.158640146255493,
      "logits/rejected": -2.0465757846832275,
      "logps/chosen": -332.1878967285156,
      "logps/rejected": -344.91595458984375,
      "loss": 0.6799,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.16046644747257233,
      "rewards/margins": 0.043529022485017776,
      "rewards/rejected": -0.2039954662322998,
      "step": 30
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.6308512113530063e-07,
      "logits/chosen": -1.2571375370025635,
      "logits/rejected": -1.1778837442398071,
      "logps/chosen": -298.17205810546875,
      "logps/rejected": -393.14373779296875,
      "loss": 0.6628,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.3048003017902374,
      "rewards/margins": 0.12428674846887589,
      "rewards/rejected": -0.4290870726108551,
      "step": 40
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.277872161641681e-07,
      "logits/chosen": -0.9371835589408875,
      "logits/rejected": -0.858717143535614,
      "logps/chosen": -337.6984558105469,
      "logps/rejected": -373.6903381347656,
      "loss": 0.6419,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.6052837371826172,
      "rewards/margins": 0.15260179340839386,
      "rewards/rejected": -0.7578855156898499,
      "step": 50
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.8318133624280046e-07,
      "logits/chosen": -0.39854905009269714,
      "logits/rejected": -0.2388809621334076,
      "logps/chosen": -312.15020751953125,
      "logps/rejected": -398.1112976074219,
      "loss": 0.6197,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.593591570854187,
      "rewards/margins": 0.26548629999160767,
      "rewards/rejected": -0.8590778112411499,
      "step": 60
    },
    {
      "epoch": 0.46,
      "learning_rate": 3.316028034595861e-07,
      "logits/chosen": 0.051435090601444244,
      "logits/rejected": 0.5506113767623901,
      "logps/chosen": -363.5517883300781,
      "logps/rejected": -446.9190368652344,
      "loss": 0.6098,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.8179994821548462,
      "rewards/margins": 0.3051639199256897,
      "rewards/rejected": -1.1231633424758911,
      "step": 70
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.7575199021178855e-07,
      "logits/chosen": 0.08734619617462158,
      "logits/rejected": 0.42371615767478943,
      "logps/chosen": -360.5260009765625,
      "logps/rejected": -410.8260803222656,
      "loss": 0.626,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.8980604410171509,
      "rewards/margins": 0.37357035279273987,
      "rewards/rejected": -1.271630883216858,
      "step": 80
    },
    {
      "epoch": 0.59,
      "learning_rate": 2.1855294234408068e-07,
      "logits/chosen": 0.1543463170528412,
      "logits/rejected": 0.2287258356809616,
      "logps/chosen": -401.79833984375,
      "logps/rejected": -489.0059509277344,
      "loss": 0.6211,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.9424636960029602,
      "rewards/margins": 0.29775407910346985,
      "rewards/rejected": -1.2402178049087524,
      "step": 90
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.6300029195778453e-07,
      "logits/chosen": 0.2451750487089157,
      "logits/rejected": 0.6907630562782288,
      "logps/chosen": -392.75775146484375,
      "logps/rejected": -463.001708984375,
      "loss": 0.6186,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.1677660942077637,
      "rewards/margins": 0.35663342475891113,
      "rewards/rejected": -1.5243995189666748,
      "step": 100
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.1200247470632392e-07,
      "logits/chosen": 0.6963290572166443,
      "logits/rejected": 0.9720438122749329,
      "logps/chosen": -385.89471435546875,
      "logps/rejected": -471.2980041503906,
      "loss": 0.6021,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -1.2825762033462524,
      "rewards/margins": 0.39920058846473694,
      "rewards/rejected": -1.681776762008667,
      "step": 110
    },
    {
      "epoch": 0.78,
      "learning_rate": 6.822945986946385e-08,
      "logits/chosen": 0.07222006469964981,
      "logits/rejected": 0.5084388256072998,
      "logps/chosen": -332.77178955078125,
      "logps/rejected": -433.77471923828125,
      "loss": 0.5982,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.933348536491394,
      "rewards/margins": 0.4566923975944519,
      "rewards/rejected": -1.3900408744812012,
      "step": 120
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.397296523427806e-08,
      "logits/chosen": 0.2585779130458832,
      "logits/rejected": 0.6913169622421265,
      "logps/chosen": -379.4100036621094,
      "logps/rejected": -430.3365173339844,
      "loss": 0.6269,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.0685771703720093,
      "rewards/margins": 0.2655676603317261,
      "rewards/rejected": -1.3341448307037354,
      "step": 130
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.1026475173977978e-08,
      "logits/chosen": 0.14197850227355957,
      "logits/rejected": 0.22905144095420837,
      "logps/chosen": -339.82000732421875,
      "logps/rejected": -435.9139709472656,
      "loss": 0.6132,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.928871750831604,
      "rewards/margins": 0.3767978549003601,
      "rewards/rejected": -1.3056696653366089,
      "step": 140
    },
    {
      "epoch": 0.98,
      "learning_rate": 5.913435276374834e-10,
      "logits/chosen": 0.19138602912425995,
      "logits/rejected": 0.4097045361995697,
      "logps/chosen": -374.04345703125,
      "logps/rejected": -433.09942626953125,
      "loss": 0.6183,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.9743102788925171,
      "rewards/margins": 0.2956470847129822,
      "rewards/rejected": -1.2699575424194336,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 153,
      "total_flos": 0.0,
      "train_loss": 0.6341538374720056,
      "train_runtime": 9045.5837,
      "train_samples_per_second": 2.167,
      "train_steps_per_second": 0.017
    }
  ],
  "logging_steps": 10,
  "max_steps": 153,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}