{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 80,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "debug/policy_chosen_logits": 20.44719123840332,
      "debug/policy_chosen_logps": -408.70501708984375,
      "debug/policy_rejected_logits": 27.257293701171875,
      "debug/policy_rejected_logps": -446.4876708984375,
      "debug/reference_chosen_logps": -408.70501708984375,
      "debug/reference_rejected_logps": -446.4876708984375,
      "epoch": 0.0125,
      "grad_norm": 7.8166135524175155,
      "learning_rate": 1e-06,
      "logits/chosen": 20.44719123840332,
      "logits/rejected": 27.257293701171875,
      "logps/chosen": -408.70501708984375,
      "logps/rejected": -446.4876708984375,
      "loss": 0.5,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "debug/policy_chosen_logits": 19.666156768798828,
      "debug/policy_chosen_logps": -409.9366760253906,
      "debug/policy_rejected_logits": 24.532733917236328,
      "debug/policy_rejected_logps": -436.7401428222656,
      "debug/reference_chosen_logps": -409.74560546875,
      "debug/reference_rejected_logps": -435.607421875,
      "epoch": 0.025,
      "grad_norm": 7.95235642646064,
      "learning_rate": 1e-06,
      "logits/chosen": 19.666156768798828,
      "logits/rejected": 24.532733917236328,
      "logps/chosen": -409.9366760253906,
      "logps/rejected": -436.7401428222656,
      "loss": 0.4941,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.0019105530809611082,
      "rewards/margins": 0.009416657499969006,
      "rewards/rejected": -0.011327208951115608,
      "step": 2
    },
    {
      "debug/policy_chosen_logits": 20.438295364379883,
      "debug/policy_chosen_logps": -425.41778564453125,
      "debug/policy_rejected_logits": 21.396846771240234,
      "debug/policy_rejected_logps": -436.84344482421875,
      "debug/reference_chosen_logps": -423.95123291015625,
      "debug/reference_rejected_logps": -434.63232421875,
      "epoch": 0.0375,
      "grad_norm": 9.487672784484882,
      "learning_rate": 1e-06,
      "logits/chosen": 20.438295364379883,
      "logits/rejected": 21.396846771240234,
      "logps/chosen": -425.41778564453125,
      "logps/rejected": -436.84344482421875,
      "loss": 0.494,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.014665374532341957,
      "rewards/margins": 0.007445754483342171,
      "rewards/rejected": -0.022111129015684128,
      "step": 3
    },
    {
      "debug/policy_chosen_logits": 18.89822769165039,
      "debug/policy_chosen_logps": -431.1507568359375,
      "debug/policy_rejected_logits": 22.7398624420166,
      "debug/policy_rejected_logps": -449.3642578125,
      "debug/reference_chosen_logps": -431.5155944824219,
      "debug/reference_rejected_logps": -447.38824462890625,
      "epoch": 0.05,
      "grad_norm": 8.935635898408727,
      "learning_rate": 1e-06,
      "logits/chosen": 18.89822769165039,
      "logits/rejected": 22.7398624420166,
      "logps/chosen": -431.1507568359375,
      "logps/rejected": -449.3642578125,
      "loss": 0.4922,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.0036485292948782444,
      "rewards/margins": 0.023408813402056694,
      "rewards/rejected": -0.019760284572839737,
      "step": 4
    },
    {
      "debug/policy_chosen_logits": 19.336538314819336,
      "debug/policy_chosen_logps": -420.7198486328125,
      "debug/policy_rejected_logits": 28.9335880279541,
      "debug/policy_rejected_logps": -456.2585144042969,
      "debug/reference_chosen_logps": -420.73626708984375,
      "debug/reference_rejected_logps": -454.71124267578125,
      "epoch": 0.0625,
      "grad_norm": 9.253028260363507,
      "learning_rate": 1e-06,
      "logits/chosen": 19.336538314819336,
      "logits/rejected": 28.9335880279541,
      "logps/chosen": -420.7198486328125,
      "logps/rejected": -456.2585144042969,
      "loss": 0.4849,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.00016406993381679058,
      "rewards/margins": 0.015636596828699112,
      "rewards/rejected": -0.015472527593374252,
      "step": 5
    },
    {
      "debug/policy_chosen_logits": 22.817670822143555,
      "debug/policy_chosen_logps": -433.24945068359375,
      "debug/policy_rejected_logits": 23.68866729736328,
      "debug/policy_rejected_logps": -431.10113525390625,
      "debug/reference_chosen_logps": -432.390869140625,
      "debug/reference_rejected_logps": -430.1065368652344,
      "epoch": 0.075,
      "grad_norm": 8.448473864866386,
      "learning_rate": 1e-06,
      "logits/chosen": 22.817670822143555,
      "logits/rejected": 23.68866729736328,
      "logps/chosen": -433.24945068359375,
      "logps/rejected": -431.10113525390625,
      "loss": 0.4789,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.008586043491959572,
      "rewards/margins": 0.0013600160600617528,
      "rewards/rejected": -0.009946059435606003,
      "step": 6
    },
    {
      "debug/policy_chosen_logits": 20.37796401977539,
      "debug/policy_chosen_logps": -423.56182861328125,
      "debug/policy_rejected_logits": 27.41672134399414,
      "debug/policy_rejected_logps": -450.90185546875,
      "debug/reference_chosen_logps": -423.9954833984375,
      "debug/reference_rejected_logps": -446.689208984375,
      "epoch": 0.0875,
      "grad_norm": 8.300945379199838,
      "learning_rate": 1e-06,
      "logits/chosen": 20.37796401977539,
      "logits/rejected": 27.41672134399414,
      "logps/chosen": -423.56182861328125,
      "logps/rejected": -450.90185546875,
      "loss": 0.4751,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.004336662590503693,
      "rewards/margins": 0.0464630126953125,
      "rewards/rejected": -0.04212634637951851,
      "step": 7
    },
    {
      "debug/policy_chosen_logits": 18.481666564941406,
      "debug/policy_chosen_logps": -427.36419677734375,
      "debug/policy_rejected_logits": 29.171558380126953,
      "debug/policy_rejected_logps": -466.4520263671875,
      "debug/reference_chosen_logps": -429.41650390625,
      "debug/reference_rejected_logps": -462.8001708984375,
      "epoch": 0.1,
      "grad_norm": 11.20281145898035,
      "learning_rate": 1e-06,
      "logits/chosen": 18.481666564941406,
      "logits/rejected": 29.171558380126953,
      "logps/chosen": -427.36419677734375,
      "logps/rejected": -466.4520263671875,
      "loss": 0.4584,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.02052280306816101,
      "rewards/margins": 0.05704108998179436,
      "rewards/rejected": -0.03651828691363335,
      "step": 8
    },
    {
      "debug/policy_chosen_logits": 22.923797607421875,
      "debug/policy_chosen_logps": -439.3218688964844,
      "debug/policy_rejected_logits": 27.558902740478516,
      "debug/policy_rejected_logps": -443.62939453125,
      "debug/reference_chosen_logps": -441.331298828125,
      "debug/reference_rejected_logps": -439.750244140625,
      "epoch": 0.1125,
      "grad_norm": 8.404029783243205,
      "learning_rate": 1e-06,
      "logits/chosen": 22.923797607421875,
      "logits/rejected": 27.558902740478516,
      "logps/chosen": -439.3218688964844,
      "logps/rejected": -443.62939453125,
      "loss": 0.4508,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.020094260573387146,
      "rewards/margins": 0.05888580530881882,
      "rewards/rejected": -0.03879154101014137,
      "step": 9
    },
    {
      "debug/policy_chosen_logits": 18.430870056152344,
      "debug/policy_chosen_logps": -405.11029052734375,
      "debug/policy_rejected_logits": 24.46626091003418,
      "debug/policy_rejected_logps": -480.96319580078125,
      "debug/reference_chosen_logps": -408.2779235839844,
      "debug/reference_rejected_logps": -473.3382263183594,
      "epoch": 0.125,
      "grad_norm": 7.719942039636563,
      "learning_rate": 1e-06,
      "logits/chosen": 18.430870056152344,
      "logits/rejected": 24.46626091003418,
      "logps/chosen": -405.11029052734375,
      "logps/rejected": -480.96319580078125,
      "loss": 0.421,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.03167613595724106,
      "rewards/margins": 0.10792595148086548,
      "rewards/rejected": -0.07624980807304382,
      "step": 10
    },
    {
      "debug/policy_chosen_logits": 18.422691345214844,
      "debug/policy_chosen_logps": -430.35479736328125,
      "debug/policy_rejected_logits": 19.293363571166992,
      "debug/policy_rejected_logps": -440.55755615234375,
      "debug/reference_chosen_logps": -431.34063720703125,
      "debug/reference_rejected_logps": -434.83477783203125,
      "epoch": 0.1375,
      "grad_norm": 7.022895536807182,
      "learning_rate": 1e-06,
      "logits/chosen": 18.422691345214844,
      "logits/rejected": 19.293363571166992,
      "logps/chosen": -430.35479736328125,
      "logps/rejected": -440.55755615234375,
      "loss": 0.4078,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.009858207777142525,
      "rewards/margins": 0.06708598881959915,
      "rewards/rejected": -0.05722777917981148,
      "step": 11
    },
    {
      "debug/policy_chosen_logits": 19.392061233520508,
      "debug/policy_chosen_logps": -428.3876037597656,
      "debug/policy_rejected_logits": 27.25312042236328,
      "debug/policy_rejected_logps": -461.404296875,
      "debug/reference_chosen_logps": -431.02984619140625,
      "debug/reference_rejected_logps": -451.54437255859375,
      "epoch": 0.15,
      "grad_norm": 6.6130326271590585,
      "learning_rate": 1e-06,
      "logits/chosen": 19.392061233520508,
      "logits/rejected": 27.25312042236328,
      "logps/chosen": -428.3876037597656,
      "logps/rejected": -461.404296875,
      "loss": 0.416,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.026422196999192238,
      "rewards/margins": 0.12502101063728333,
      "rewards/rejected": -0.09859882295131683,
      "step": 12
    },
    {
      "debug/policy_chosen_logits": 20.876129150390625,
      "debug/policy_chosen_logps": -419.8938293457031,
      "debug/policy_rejected_logits": 24.144548416137695,
      "debug/policy_rejected_logps": -437.48370361328125,
      "debug/reference_chosen_logps": -426.415771484375,
      "debug/reference_rejected_logps": -429.47802734375,
      "epoch": 0.1625,
      "grad_norm": 7.606735140301725,
      "learning_rate": 1e-06,
      "logits/chosen": 20.876129150390625,
      "logits/rejected": 24.144548416137695,
      "logps/chosen": -419.8938293457031,
      "logps/rejected": -437.48370361328125,
      "loss": 0.3942,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.0652196854352951,
      "rewards/margins": 0.1452765166759491,
      "rewards/rejected": -0.08005683869123459,
      "step": 13
    },
    {
      "debug/policy_chosen_logits": 17.269620895385742,
      "debug/policy_chosen_logps": -398.87115478515625,
      "debug/policy_rejected_logits": 25.764076232910156,
      "debug/policy_rejected_logps": -460.26678466796875,
      "debug/reference_chosen_logps": -408.37481689453125,
      "debug/reference_rejected_logps": -455.5852966308594,
      "epoch": 0.175,
      "grad_norm": 5.932839484754313,
      "learning_rate": 1e-06,
      "logits/chosen": 17.269620895385742,
      "logits/rejected": 25.764076232910156,
      "logps/chosen": -398.87115478515625,
      "logps/rejected": -460.26678466796875,
      "loss": 0.4258,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.095036581158638,
      "rewards/margins": 0.1418514996767044,
      "rewards/rejected": -0.046814918518066406,
      "step": 14
    },
    {
      "debug/policy_chosen_logits": 18.60376739501953,
      "debug/policy_chosen_logps": -411.95623779296875,
      "debug/policy_rejected_logits": 23.55546760559082,
      "debug/policy_rejected_logps": -423.6629638671875,
      "debug/reference_chosen_logps": -427.5350341796875,
      "debug/reference_rejected_logps": -418.62518310546875,
      "epoch": 0.1875,
      "grad_norm": 5.826719160562423,
      "learning_rate": 1e-06,
      "logits/chosen": 18.60376739501953,
      "logits/rejected": 23.55546760559082,
      "logps/chosen": -411.95623779296875,
      "logps/rejected": -423.6629638671875,
      "loss": 0.4082,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.15578752756118774,
      "rewards/margins": 0.20616531372070312,
      "rewards/rejected": -0.05037776753306389,
      "step": 15
    },
    {
      "debug/policy_chosen_logits": 19.708065032958984,
      "debug/policy_chosen_logps": -421.86419677734375,
      "debug/policy_rejected_logits": 23.621994018554688,
      "debug/policy_rejected_logps": -431.23486328125,
      "debug/reference_chosen_logps": -426.4266662597656,
      "debug/reference_rejected_logps": -436.3722229003906,
      "epoch": 0.2,
      "grad_norm": 6.112736963468344,
      "learning_rate": 1e-06,
      "logits/chosen": 19.708065032958984,
      "logits/rejected": 23.621994018554688,
      "logps/chosen": -421.86419677734375,
      "logps/rejected": -431.23486328125,
      "loss": 0.3901,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.04562458395957947,
      "rewards/margins": -0.005748823285102844,
      "rewards/rejected": 0.051373403519392014,
      "step": 16
    },
    {
      "debug/policy_chosen_logits": 20.457584381103516,
      "debug/policy_chosen_logps": -421.99639892578125,
      "debug/policy_rejected_logits": 27.524091720581055,
      "debug/policy_rejected_logps": -444.53289794921875,
      "debug/reference_chosen_logps": -430.9820556640625,
      "debug/reference_rejected_logps": -434.97113037109375,
      "epoch": 0.2125,
      "grad_norm": 5.201415641144627,
      "learning_rate": 1e-06,
      "logits/chosen": 20.457584381103516,
      "logits/rejected": 27.524091720581055,
      "logps/chosen": -421.99639892578125,
      "logps/rejected": -444.53289794921875,
      "loss": 0.3627,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.08985641598701477,
      "rewards/margins": 0.18547385931015015,
      "rewards/rejected": -0.09561744332313538,
      "step": 17
    },
    {
      "debug/policy_chosen_logits": 19.323848724365234,
      "debug/policy_chosen_logps": -410.0373840332031,
      "debug/policy_rejected_logits": 30.60475730895996,
      "debug/policy_rejected_logps": -429.958740234375,
      "debug/reference_chosen_logps": -416.14520263671875,
      "debug/reference_rejected_logps": -423.42071533203125,
      "epoch": 0.225,
      "grad_norm": 5.252127969055532,
      "learning_rate": 1e-06,
      "logits/chosen": 19.323848724365234,
      "logits/rejected": 30.60475730895996,
      "logps/chosen": -410.0373840332031,
      "logps/rejected": -429.958740234375,
      "loss": 0.3815,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.061078451573848724,
      "rewards/margins": 0.1264585554599762,
      "rewards/rejected": -0.06538009643554688,
      "step": 18
    },
    {
      "debug/policy_chosen_logits": 16.97679328918457,
      "debug/policy_chosen_logps": -392.15478515625,
      "debug/policy_rejected_logits": 27.920000076293945,
      "debug/policy_rejected_logps": -447.72747802734375,
      "debug/reference_chosen_logps": -410.5232238769531,
      "debug/reference_rejected_logps": -432.1972961425781,
      "epoch": 0.2375,
      "grad_norm": 4.748490487188559,
      "learning_rate": 1e-06,
      "logits/chosen": 16.97679328918457,
      "logits/rejected": 27.920000076293945,
      "logps/chosen": -392.15478515625,
      "logps/rejected": -447.72747802734375,
      "loss": 0.338,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.1836845427751541,
      "rewards/margins": 0.33898642659187317,
      "rewards/rejected": -0.15530189871788025,
      "step": 19
    },
    {
      "debug/policy_chosen_logits": 21.503040313720703,
      "debug/policy_chosen_logps": -425.363037109375,
      "debug/policy_rejected_logits": 31.72104263305664,
      "debug/policy_rejected_logps": -472.585693359375,
      "debug/reference_chosen_logps": -439.15594482421875,
      "debug/reference_rejected_logps": -448.0031433105469,
      "epoch": 0.25,
      "grad_norm": 4.598515072265396,
      "learning_rate": 1e-06,
      "logits/chosen": 21.503040313720703,
      "logits/rejected": 31.72104263305664,
      "logps/chosen": -425.363037109375,
      "logps/rejected": -472.585693359375,
      "loss": 0.2996,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.1379288136959076,
      "rewards/margins": 0.3837544322013855,
      "rewards/rejected": -0.2458256185054779,
      "step": 20
    },
    {
      "debug/policy_chosen_logits": 19.64223289489746,
      "debug/policy_chosen_logps": -411.8244323730469,
      "debug/policy_rejected_logits": 25.16058349609375,
      "debug/policy_rejected_logps": -479.46990966796875,
      "debug/reference_chosen_logps": -419.84149169921875,
      "debug/reference_rejected_logps": -448.9364013671875,
      "epoch": 0.2625,
      "grad_norm": 4.506314070543816,
      "learning_rate": 1e-06,
      "logits/chosen": 19.64223289489746,
      "logits/rejected": 25.16058349609375,
      "logps/chosen": -411.8244323730469,
      "logps/rejected": -479.46990966796875,
      "loss": 0.3326,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.08017054945230484,
      "rewards/margins": 0.38550513982772827,
      "rewards/rejected": -0.30533456802368164,
      "step": 21
    },
    {
      "debug/policy_chosen_logits": 20.455036163330078,
      "debug/policy_chosen_logps": -401.1058044433594,
      "debug/policy_rejected_logits": 27.924205780029297,
      "debug/policy_rejected_logps": -479.4624938964844,
      "debug/reference_chosen_logps": -415.33001708984375,
      "debug/reference_rejected_logps": -445.6070861816406,
      "epoch": 0.275,
      "grad_norm": 8.086589389208452,
      "learning_rate": 1e-06,
      "logits/chosen": 20.455036163330078,
      "logits/rejected": 27.924205780029297,
      "logps/chosen": -401.1058044433594,
      "logps/rejected": -479.4624938964844,
      "loss": 0.284,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.14224204421043396,
      "rewards/margins": 0.4807964563369751,
      "rewards/rejected": -0.3385544419288635,
      "step": 22
    },
    {
      "debug/policy_chosen_logits": 22.28139305114746,
      "debug/policy_chosen_logps": -428.8080139160156,
      "debug/policy_rejected_logits": 29.031978607177734,
      "debug/policy_rejected_logps": -496.9381103515625,
      "debug/reference_chosen_logps": -425.82586669921875,
      "debug/reference_rejected_logps": -460.11199951171875,
      "epoch": 0.2875,
      "grad_norm": 7.117765571472066,
      "learning_rate": 1e-06,
      "logits/chosen": 22.28139305114746,
      "logits/rejected": 29.031978607177734,
      "logps/chosen": -428.8080139160156,
      "logps/rejected": -496.9381103515625,
      "loss": 0.3124,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.02982150949537754,
      "rewards/margins": 0.33843994140625,
      "rewards/rejected": -0.3682614266872406,
      "step": 23
    },
    {
      "debug/policy_chosen_logits": 16.489967346191406,
      "debug/policy_chosen_logps": -386.43011474609375,
      "debug/policy_rejected_logits": 27.82581329345703,
      "debug/policy_rejected_logps": -452.87548828125,
      "debug/reference_chosen_logps": -407.1941223144531,
      "debug/reference_rejected_logps": -445.4601745605469,
      "epoch": 0.3,
      "grad_norm": 5.311518244442526,
      "learning_rate": 1e-06,
      "logits/chosen": 16.489967346191406,
      "logits/rejected": 27.82581329345703,
      "logps/chosen": -386.43011474609375,
      "logps/rejected": -452.87548828125,
      "loss": 0.3473,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.20764002203941345,
      "rewards/margins": 0.2817932963371277,
      "rewards/rejected": -0.07415324449539185,
      "step": 24
    },
    {
      "debug/policy_chosen_logits": 20.677587509155273,
      "debug/policy_chosen_logps": -401.9738464355469,
      "debug/policy_rejected_logits": 26.294002532958984,
      "debug/policy_rejected_logps": -462.0292663574219,
      "debug/reference_chosen_logps": -419.8895568847656,
      "debug/reference_rejected_logps": -468.2488708496094,
      "epoch": 0.3125,
      "grad_norm": 6.577603302459432,
      "learning_rate": 1e-06,
      "logits/chosen": 20.677587509155273,
      "logits/rejected": 26.294002532958984,
      "logps/chosen": -401.9738464355469,
      "logps/rejected": -462.0292663574219,
      "loss": 0.3873,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.1791572868824005,
      "rewards/margins": 0.11696115881204605,
      "rewards/rejected": 0.06219611316919327,
      "step": 25
    },
    {
      "debug/policy_chosen_logits": 21.27174186706543,
      "debug/policy_chosen_logps": -403.5057678222656,
      "debug/policy_rejected_logits": 28.13071060180664,
      "debug/policy_rejected_logps": -465.28216552734375,
      "debug/reference_chosen_logps": -426.16949462890625,
      "debug/reference_rejected_logps": -428.91204833984375,
      "epoch": 0.325,
      "grad_norm": 9.280568242676463,
      "learning_rate": 1e-06,
      "logits/chosen": 21.27174186706543,
      "logits/rejected": 28.13071060180664,
      "logps/chosen": -403.5057678222656,
      "logps/rejected": -465.28216552734375,
      "loss": 0.3524,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2266373336315155,
      "rewards/margins": 0.5903383493423462,
      "rewards/rejected": -0.3637010157108307,
      "step": 26
    },
    {
      "debug/policy_chosen_logits": 19.095767974853516,
      "debug/policy_chosen_logps": -382.9574890136719,
      "debug/policy_rejected_logits": 31.333406448364258,
      "debug/policy_rejected_logps": -468.0421447753906,
      "debug/reference_chosen_logps": -405.85858154296875,
      "debug/reference_rejected_logps": -444.3897705078125,
      "epoch": 0.3375,
      "grad_norm": 4.305140313967255,
      "learning_rate": 1e-06,
      "logits/chosen": 19.095767974853516,
      "logits/rejected": 31.333406448364258,
      "logps/chosen": -382.9574890136719,
      "logps/rejected": -468.0421447753906,
      "loss": 0.3094,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.2290109544992447,
      "rewards/margins": 0.46553459763526917,
      "rewards/rejected": -0.23652365803718567,
      "step": 27
    },
    {
      "debug/policy_chosen_logits": 20.944793701171875,
      "debug/policy_chosen_logps": -403.42010498046875,
      "debug/policy_rejected_logits": 25.954708099365234,
      "debug/policy_rejected_logps": -493.7495422363281,
      "debug/reference_chosen_logps": -419.398681640625,
      "debug/reference_rejected_logps": -464.70562744140625,
      "epoch": 0.35,
      "grad_norm": 5.091666773488247,
      "learning_rate": 1e-06,
      "logits/chosen": 20.944793701171875,
      "logits/rejected": 25.954708099365234,
      "logps/chosen": -403.42010498046875,
      "logps/rejected": -493.7495422363281,
      "loss": 0.3118,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.15978576242923737,
      "rewards/margins": 0.45022472739219666,
      "rewards/rejected": -0.29043900966644287,
      "step": 28
    },
    {
      "debug/policy_chosen_logits": 18.986360549926758,
      "debug/policy_chosen_logps": -396.65667724609375,
      "debug/policy_rejected_logits": 26.656171798706055,
      "debug/policy_rejected_logps": -464.8284912109375,
      "debug/reference_chosen_logps": -413.4096984863281,
      "debug/reference_rejected_logps": -439.12982177734375,
      "epoch": 0.3625,
      "grad_norm": 7.528653365476016,
      "learning_rate": 1e-06,
      "logits/chosen": 18.986360549926758,
      "logits/rejected": 26.656171798706055,
      "logps/chosen": -396.65667724609375,
      "logps/rejected": -464.8284912109375,
      "loss": 0.3019,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.16753005981445312,
      "rewards/margins": 0.4245164692401886,
      "rewards/rejected": -0.2569863796234131,
      "step": 29
    },
    {
      "debug/policy_chosen_logits": 19.743337631225586,
      "debug/policy_chosen_logps": -390.146484375,
      "debug/policy_rejected_logits": 31.34699058532715,
      "debug/policy_rejected_logps": -483.0184631347656,
      "debug/reference_chosen_logps": -425.2975158691406,
      "debug/reference_rejected_logps": -459.4593811035156,
      "epoch": 0.375,
      "grad_norm": 6.576815284840866,
      "learning_rate": 1e-06,
      "logits/chosen": 19.743337631225586,
      "logits/rejected": 31.34699058532715,
      "logps/chosen": -390.146484375,
      "logps/rejected": -483.0184631347656,
      "loss": 0.3144,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.35151058435440063,
      "rewards/margins": 0.5871013402938843,
      "rewards/rejected": -0.23559081554412842,
      "step": 30
    },
    {
      "debug/policy_chosen_logits": 23.170053482055664,
      "debug/policy_chosen_logps": -412.2349548339844,
      "debug/policy_rejected_logits": 29.11348533630371,
      "debug/policy_rejected_logps": -455.674072265625,
      "debug/reference_chosen_logps": -428.70257568359375,
      "debug/reference_rejected_logps": -441.6961975097656,
      "epoch": 0.3875,
      "grad_norm": 5.728033865834093,
      "learning_rate": 1e-06,
      "logits/chosen": 23.170053482055664,
      "logits/rejected": 29.11348533630371,
      "logps/chosen": -412.2349548339844,
      "logps/rejected": -455.674072265625,
      "loss": 0.3214,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.16467612981796265,
      "rewards/margins": 0.3044551610946655,
      "rewards/rejected": -0.1397790014743805,
      "step": 31
    },
    {
      "debug/policy_chosen_logits": 16.38572883605957,
      "debug/policy_chosen_logps": -390.042724609375,
      "debug/policy_rejected_logits": 22.589452743530273,
      "debug/policy_rejected_logps": -453.4793701171875,
      "debug/reference_chosen_logps": -415.5836181640625,
      "debug/reference_rejected_logps": -437.25,
      "epoch": 0.4,
      "grad_norm": 5.438026921069897,
      "learning_rate": 1e-06,
      "logits/chosen": 16.38572883605957,
      "logits/rejected": 22.589452743530273,
      "logps/chosen": -390.042724609375,
      "logps/rejected": -453.4793701171875,
      "loss": 0.3274,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.2554091513156891,
      "rewards/margins": 0.41770267486572266,
      "rewards/rejected": -0.16229349374771118,
      "step": 32
    },
    {
      "debug/policy_chosen_logits": 19.873872756958008,
      "debug/policy_chosen_logps": -417.18951416015625,
      "debug/policy_rejected_logits": 26.914785385131836,
      "debug/policy_rejected_logps": -478.3274841308594,
      "debug/reference_chosen_logps": -428.84234619140625,
      "debug/reference_rejected_logps": -451.3583068847656,
      "epoch": 0.4125,
      "grad_norm": 5.039379106966324,
      "learning_rate": 1e-06,
      "logits/chosen": 19.873872756958008,
      "logits/rejected": 26.914785385131836,
      "logps/chosen": -417.18951416015625,
      "logps/rejected": -478.3274841308594,
      "loss": 0.2881,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.11652831733226776,
      "rewards/margins": 0.38622018694877625,
      "rewards/rejected": -0.2696918845176697,
      "step": 33
    },
    {
      "debug/policy_chosen_logits": 21.201560974121094,
      "debug/policy_chosen_logps": -414.26641845703125,
      "debug/policy_rejected_logits": 23.623708724975586,
      "debug/policy_rejected_logps": -460.96356201171875,
      "debug/reference_chosen_logps": -434.948974609375,
      "debug/reference_rejected_logps": -444.27191162109375,
      "epoch": 0.425,
      "grad_norm": 4.310717910219745,
      "learning_rate": 1e-06,
      "logits/chosen": 21.201560974121094,
      "logits/rejected": 23.623708724975586,
      "logps/chosen": -414.26641845703125,
      "logps/rejected": -460.96356201171875,
      "loss": 0.2987,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.20682545006275177,
      "rewards/margins": 0.3737417459487915,
      "rewards/rejected": -0.16691631078720093,
      "step": 34
    },
    {
      "debug/policy_chosen_logits": 21.479965209960938,
      "debug/policy_chosen_logps": -378.7373352050781,
      "debug/policy_rejected_logits": 28.288991928100586,
      "debug/policy_rejected_logps": -499.9162292480469,
      "debug/reference_chosen_logps": -399.58843994140625,
      "debug/reference_rejected_logps": -472.7664489746094,
      "epoch": 0.4375,
      "grad_norm": 4.1957614688185645,
      "learning_rate": 1e-06,
      "logits/chosen": 21.479965209960938,
      "logits/rejected": 28.288991928100586,
      "logps/chosen": -378.7373352050781,
      "logps/rejected": -499.9162292480469,
      "loss": 0.2693,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.20851120352745056,
      "rewards/margins": 0.48000913858413696,
      "rewards/rejected": -0.2714979648590088,
      "step": 35
    },
    {
      "debug/policy_chosen_logits": 20.284452438354492,
      "debug/policy_chosen_logps": -394.7576904296875,
      "debug/policy_rejected_logits": 24.526214599609375,
      "debug/policy_rejected_logps": -462.82427978515625,
      "debug/reference_chosen_logps": -414.8307800292969,
      "debug/reference_rejected_logps": -430.0937805175781,
      "epoch": 0.45,
      "grad_norm": 4.529884106044187,
      "learning_rate": 1e-06,
      "logits/chosen": 20.284452438354492,
      "logits/rejected": 24.526214599609375,
      "logps/chosen": -394.7576904296875,
      "logps/rejected": -462.82427978515625,
      "loss": 0.2508,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2007310390472412,
      "rewards/margins": 0.5280359387397766,
      "rewards/rejected": -0.3273049294948578,
      "step": 36
    },
    {
      "debug/policy_chosen_logits": 19.54199981689453,
      "debug/policy_chosen_logps": -400.54559326171875,
      "debug/policy_rejected_logits": 32.27294158935547,
      "debug/policy_rejected_logps": -503.24749755859375,
      "debug/reference_chosen_logps": -424.2836608886719,
      "debug/reference_rejected_logps": -475.5150146484375,
      "epoch": 0.4625,
      "grad_norm": 4.291241072360298,
      "learning_rate": 1e-06,
      "logits/chosen": 19.54199981689453,
      "logits/rejected": 32.27294158935547,
      "logps/chosen": -400.54559326171875,
      "logps/rejected": -503.24749755859375,
      "loss": 0.2755,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.23738029599189758,
      "rewards/margins": 0.51470547914505,
      "rewards/rejected": -0.27732518315315247,
      "step": 37
    },
    {
      "debug/policy_chosen_logits": 24.72857093811035,
      "debug/policy_chosen_logps": -426.8404541015625,
      "debug/policy_rejected_logits": 27.871118545532227,
      "debug/policy_rejected_logps": -462.07275390625,
      "debug/reference_chosen_logps": -438.54156494140625,
      "debug/reference_rejected_logps": -446.54925537109375,
      "epoch": 0.475,
      "grad_norm": 4.049213448224943,
      "learning_rate": 1e-06,
      "logits/chosen": 24.72857093811035,
      "logits/rejected": 27.871118545532227,
      "logps/chosen": -426.8404541015625,
      "logps/rejected": -462.07275390625,
      "loss": 0.3102,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.11701113730669022,
      "rewards/margins": 0.27224594354629517,
      "rewards/rejected": -0.15523478388786316,
      "step": 38
    },
    {
      "debug/policy_chosen_logits": 21.91419792175293,
      "debug/policy_chosen_logps": -408.05224609375,
      "debug/policy_rejected_logits": 21.785812377929688,
      "debug/policy_rejected_logps": -427.7870178222656,
      "debug/reference_chosen_logps": -421.9055480957031,
      "debug/reference_rejected_logps": -422.02252197265625,
      "epoch": 0.4875,
      "grad_norm": 5.051022630915263,
      "learning_rate": 1e-06,
      "logits/chosen": 21.91419792175293,
      "logits/rejected": 21.785812377929688,
      "logps/chosen": -408.05224609375,
      "logps/rejected": -427.7870178222656,
      "loss": 0.3337,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.13853324949741364,
      "rewards/margins": 0.19617804884910583,
      "rewards/rejected": -0.0576448068022728,
      "step": 39
    },
    {
      "debug/policy_chosen_logits": 21.50871467590332,
      "debug/policy_chosen_logps": -404.84783935546875,
      "debug/policy_rejected_logits": 26.14636993408203,
      "debug/policy_rejected_logps": -463.62982177734375,
      "debug/reference_chosen_logps": -417.2750244140625,
      "debug/reference_rejected_logps": -446.09521484375,
      "epoch": 0.5,
      "grad_norm": 4.15379359930677,
      "learning_rate": 1e-06,
      "logits/chosen": 21.50871467590332,
      "logits/rejected": 26.14636993408203,
      "logps/chosen": -404.84783935546875,
      "logps/rejected": -463.62982177734375,
      "loss": 0.3061,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.12427165359258652,
      "rewards/margins": 0.2996178865432739,
      "rewards/rejected": -0.1753462255001068,
      "step": 40
    },
    {
      "debug/policy_chosen_logits": 19.015485763549805,
      "debug/policy_chosen_logps": -403.9427185058594,
      "debug/policy_rejected_logits": 25.147703170776367,
      "debug/policy_rejected_logps": -482.85748291015625,
      "debug/reference_chosen_logps": -431.6396179199219,
      "debug/reference_rejected_logps": -454.42474365234375,
      "epoch": 0.5125,
      "grad_norm": 4.030629767467611,
      "learning_rate": 1e-06,
      "logits/chosen": 19.015485763549805,
      "logits/rejected": 25.147703170776367,
      "logps/chosen": -403.9427185058594,
      "logps/rejected": -482.85748291015625,
      "loss": 0.2803,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.27696919441223145,
      "rewards/margins": 0.5612967014312744,
      "rewards/rejected": -0.28432750701904297,
      "step": 41
    },
    {
      "debug/policy_chosen_logits": 20.53624153137207,
      "debug/policy_chosen_logps": -390.5385437011719,
      "debug/policy_rejected_logits": 29.287269592285156,
      "debug/policy_rejected_logps": -497.9327392578125,
      "debug/reference_chosen_logps": -414.4248962402344,
      "debug/reference_rejected_logps": -457.1654357910156,
      "epoch": 0.525,
      "grad_norm": 5.170853621888292,
      "learning_rate": 1e-06,
      "logits/chosen": 20.53624153137207,
      "logits/rejected": 29.287269592285156,
      "logps/chosen": -390.5385437011719,
      "logps/rejected": -497.9327392578125,
      "loss": 0.2504,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2388634830713272,
      "rewards/margins": 0.6465361714363098,
      "rewards/rejected": -0.4076727032661438,
      "step": 42
    },
    {
      "debug/policy_chosen_logits": 20.2578125,
      "debug/policy_chosen_logps": -405.1622314453125,
      "debug/policy_rejected_logits": 28.01288414001465,
      "debug/policy_rejected_logps": -465.55914306640625,
      "debug/reference_chosen_logps": -429.2195129394531,
      "debug/reference_rejected_logps": -439.44915771484375,
      "epoch": 0.5375,
      "grad_norm": 3.882500417686335,
      "learning_rate": 1e-06,
      "logits/chosen": 20.2578125,
      "logits/rejected": 28.01288414001465,
      "logps/chosen": -405.1622314453125,
      "logps/rejected": -465.55914306640625,
      "loss": 0.2292,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.2405730038881302,
      "rewards/margins": 0.5016728639602661,
      "rewards/rejected": -0.2610998749732971,
      "step": 43
    },
    {
      "debug/policy_chosen_logits": 17.483173370361328,
      "debug/policy_chosen_logps": -364.91351318359375,
      "debug/policy_rejected_logits": 29.26898765563965,
      "debug/policy_rejected_logps": -483.2657165527344,
      "debug/reference_chosen_logps": -391.24664306640625,
      "debug/reference_rejected_logps": -457.18780517578125,
      "epoch": 0.55,
      "grad_norm": 4.7605947943975675,
      "learning_rate": 1e-06,
      "logits/chosen": 17.483173370361328,
      "logits/rejected": 29.26898765563965,
      "logps/chosen": -364.91351318359375,
      "logps/rejected": -483.2657165527344,
      "loss": 0.285,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.26333099603652954,
      "rewards/margins": 0.5241098403930664,
      "rewards/rejected": -0.26077884435653687,
      "step": 44
    },
    {
      "debug/policy_chosen_logits": 19.23508071899414,
      "debug/policy_chosen_logps": -387.4403076171875,
      "debug/policy_rejected_logits": 27.274477005004883,
      "debug/policy_rejected_logps": -444.69635009765625,
      "debug/reference_chosen_logps": -412.97003173828125,
      "debug/reference_rejected_logps": -437.71649169921875,
      "epoch": 0.5625,
      "grad_norm": 7.535235898290158,
      "learning_rate": 1e-06,
      "logits/chosen": 19.23508071899414,
      "logits/rejected": 27.274477005004883,
      "logps/chosen": -387.4403076171875,
      "logps/rejected": -444.69635009765625,
      "loss": 0.3417,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.25529688596725464,
      "rewards/margins": 0.32509559392929077,
      "rewards/rejected": -0.06979873776435852,
      "step": 45
    },
    {
      "debug/policy_chosen_logits": 20.150699615478516,
      "debug/policy_chosen_logps": -383.12786865234375,
      "debug/policy_rejected_logits": 23.752517700195312,
      "debug/policy_rejected_logps": -429.2421875,
      "debug/reference_chosen_logps": -409.3837585449219,
      "debug/reference_rejected_logps": -413.81390380859375,
      "epoch": 0.575,
      "grad_norm": 6.267381407493171,
      "learning_rate": 1e-06,
      "logits/chosen": 20.150699615478516,
      "logits/rejected": 23.752517700195312,
      "logps/chosen": -383.12786865234375,
      "logps/rejected": -429.2421875,
      "loss": 0.3195,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.26255881786346436,
      "rewards/margins": 0.41684138774871826,
      "rewards/rejected": -0.1542825698852539,
      "step": 46
    },
    {
      "debug/policy_chosen_logits": 19.68763542175293,
      "debug/policy_chosen_logps": -385.29144287109375,
      "debug/policy_rejected_logits": 28.113876342773438,
      "debug/policy_rejected_logps": -451.77642822265625,
      "debug/reference_chosen_logps": -412.47412109375,
      "debug/reference_rejected_logps": -438.6710205078125,
      "epoch": 0.5875,
      "grad_norm": 5.633066278680009,
      "learning_rate": 1e-06,
      "logits/chosen": 19.68763542175293,
      "logits/rejected": 28.113876342773438,
      "logps/chosen": -385.29144287109375,
      "logps/rejected": -451.77642822265625,
      "loss": 0.2709,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.2718266546726227,
      "rewards/margins": 0.4028809666633606,
      "rewards/rejected": -0.13105429708957672,
      "step": 47
    },
    {
      "debug/policy_chosen_logits": 21.418119430541992,
      "debug/policy_chosen_logps": -418.92596435546875,
      "debug/policy_rejected_logits": 28.619443893432617,
      "debug/policy_rejected_logps": -447.74090576171875,
      "debug/reference_chosen_logps": -431.38714599609375,
      "debug/reference_rejected_logps": -444.2582092285156,
      "epoch": 0.6,
      "grad_norm": 4.297385322384447,
      "learning_rate": 1e-06,
      "logits/chosen": 21.418119430541992,
      "logits/rejected": 28.619443893432617,
      "logps/chosen": -418.92596435546875,
      "logps/rejected": -447.74090576171875,
      "loss": 0.302,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.12461188435554504,
      "rewards/margins": 0.15943893790245056,
      "rewards/rejected": -0.034827034920454025,
      "step": 48
    },
    {
      "debug/policy_chosen_logits": 21.347043991088867,
      "debug/policy_chosen_logps": -409.0511474609375,
      "debug/policy_rejected_logits": 29.213226318359375,
      "debug/policy_rejected_logps": -479.66387939453125,
      "debug/reference_chosen_logps": -424.4683837890625,
      "debug/reference_rejected_logps": -453.17791748046875,
      "epoch": 0.6125,
      "grad_norm": 3.600194804734596,
      "learning_rate": 1e-06,
      "logits/chosen": 21.347043991088867,
      "logits/rejected": 29.213226318359375,
      "logps/chosen": -409.0511474609375,
      "logps/rejected": -479.66387939453125,
      "loss": 0.3015,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.15417221188545227,
      "rewards/margins": 0.41903185844421387,
      "rewards/rejected": -0.2648596465587616,
      "step": 49
    },
    {
      "debug/policy_chosen_logits": 19.539743423461914,
      "debug/policy_chosen_logps": -406.68109130859375,
      "debug/policy_rejected_logits": 23.604896545410156,
      "debug/policy_rejected_logps": -447.7806701660156,
      "debug/reference_chosen_logps": -427.583740234375,
      "debug/reference_rejected_logps": -432.884765625,
      "epoch": 0.625,
      "grad_norm": 4.128916174651822,
      "learning_rate": 1e-06,
      "logits/chosen": 19.539743423461914,
      "logits/rejected": 23.604896545410156,
      "logps/chosen": -406.68109130859375,
      "logps/rejected": -447.7806701660156,
      "loss": 0.3165,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2090263068675995,
      "rewards/margins": 0.35798513889312744,
      "rewards/rejected": -0.14895884692668915,
      "step": 50
    },
    {
      "debug/policy_chosen_logits": 23.078472137451172,
      "debug/policy_chosen_logps": -413.09417724609375,
      "debug/policy_rejected_logits": 27.335500717163086,
      "debug/policy_rejected_logps": -472.38079833984375,
      "debug/reference_chosen_logps": -425.72039794921875,
      "debug/reference_rejected_logps": -457.5447082519531,
      "epoch": 0.6375,
      "grad_norm": 4.640148404581812,
      "learning_rate": 1e-06,
      "logits/chosen": 23.078472137451172,
      "logits/rejected": 27.335500717163086,
      "logps/chosen": -413.09417724609375,
      "logps/rejected": -472.38079833984375,
      "loss": 0.339,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.12626220285892487,
      "rewards/margins": 0.27462291717529297,
      "rewards/rejected": -0.1483606994152069,
      "step": 51
    },
    {
      "debug/policy_chosen_logits": 18.798938751220703,
      "debug/policy_chosen_logps": -392.9048767089844,
      "debug/policy_rejected_logits": 31.566543579101562,
      "debug/policy_rejected_logps": -530.4522094726562,
      "debug/reference_chosen_logps": -410.1727294921875,
      "debug/reference_rejected_logps": -481.0628967285156,
      "epoch": 0.65,
      "grad_norm": 5.586177810089015,
      "learning_rate": 1e-06,
      "logits/chosen": 18.798938751220703,
      "logits/rejected": 31.566543579101562,
      "logps/chosen": -392.9048767089844,
      "logps/rejected": -530.4522094726562,
      "loss": 0.2993,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.17267844080924988,
      "rewards/margins": 0.6665716171264648,
      "rewards/rejected": -0.49389317631721497,
      "step": 52
    },
    {
      "debug/policy_chosen_logits": 22.037538528442383,
      "debug/policy_chosen_logps": -415.8323669433594,
      "debug/policy_rejected_logits": 27.0804386138916,
      "debug/policy_rejected_logps": -468.87591552734375,
      "debug/reference_chosen_logps": -431.1224365234375,
      "debug/reference_rejected_logps": -426.37371826171875,
      "epoch": 0.6625,
      "grad_norm": 6.488572834551389,
      "learning_rate": 1e-06,
      "logits/chosen": 22.037538528442383,
      "logits/rejected": 27.0804386138916,
      "logps/chosen": -415.8323669433594,
      "logps/rejected": -468.87591552734375,
      "loss": 0.3265,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.1529008448123932,
      "rewards/margins": 0.5779224634170532,
      "rewards/rejected": -0.42502161860466003,
      "step": 53
    },
    {
      "debug/policy_chosen_logits": 18.0117130279541,
      "debug/policy_chosen_logps": -398.383056640625,
      "debug/policy_rejected_logits": 30.224939346313477,
      "debug/policy_rejected_logps": -452.33502197265625,
      "debug/reference_chosen_logps": -418.76708984375,
      "debug/reference_rejected_logps": -439.8731384277344,
      "epoch": 0.675,
      "grad_norm": 6.398923285824906,
      "learning_rate": 1e-06,
      "logits/chosen": 18.0117130279541,
      "logits/rejected": 30.224939346313477,
      "logps/chosen": -398.383056640625,
      "logps/rejected": -452.33502197265625,
      "loss": 0.2931,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.20384037494659424,
      "rewards/margins": 0.32845935225486755,
      "rewards/rejected": -0.12461898475885391,
      "step": 54
    },
    {
      "debug/policy_chosen_logits": 19.977191925048828,
      "debug/policy_chosen_logps": -387.9195556640625,
      "debug/policy_rejected_logits": 27.902311325073242,
      "debug/policy_rejected_logps": -465.3125,
      "debug/reference_chosen_logps": -413.6059265136719,
      "debug/reference_rejected_logps": -443.1114196777344,
      "epoch": 0.6875,
      "grad_norm": 3.685551814432106,
      "learning_rate": 1e-06,
      "logits/chosen": 19.977191925048828,
      "logits/rejected": 27.902311325073242,
      "logps/chosen": -387.9195556640625,
      "logps/rejected": -465.3125,
      "loss": 0.2927,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.25686389207839966,
      "rewards/margins": 0.4788748323917389,
      "rewards/rejected": -0.22201094031333923,
      "step": 55
    },
    {
      "debug/policy_chosen_logits": 19.4679012298584,
      "debug/policy_chosen_logps": -387.61016845703125,
      "debug/policy_rejected_logits": 24.729759216308594,
      "debug/policy_rejected_logps": -484.8683776855469,
      "debug/reference_chosen_logps": -407.5805969238281,
      "debug/reference_rejected_logps": -456.95703125,
      "epoch": 0.7,
      "grad_norm": 3.5458155693246467,
      "learning_rate": 1e-06,
      "logits/chosen": 19.4679012298584,
      "logits/rejected": 24.729759216308594,
      "logps/chosen": -387.61016845703125,
      "logps/rejected": -484.8683776855469,
      "loss": 0.2736,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.19970446825027466,
      "rewards/margins": 0.4788181483745575,
      "rewards/rejected": -0.27911368012428284,
      "step": 56
    },
    {
      "debug/policy_chosen_logits": 19.905282974243164,
      "debug/policy_chosen_logps": -404.56280517578125,
      "debug/policy_rejected_logits": 23.55043601989746,
      "debug/policy_rejected_logps": -455.38018798828125,
      "debug/reference_chosen_logps": -425.48101806640625,
      "debug/reference_rejected_logps": -444.97747802734375,
      "epoch": 0.7125,
      "grad_norm": 3.2623990350257848,
      "learning_rate": 1e-06,
      "logits/chosen": 19.905282974243164,
      "logits/rejected": 23.55043601989746,
      "logps/chosen": -404.56280517578125,
      "logps/rejected": -455.38018798828125,
      "loss": 0.2562,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.20918196439743042,
      "rewards/margins": 0.31320905685424805,
      "rewards/rejected": -0.10402709245681763,
      "step": 57
    },
    {
      "debug/policy_chosen_logits": 23.94160270690918,
      "debug/policy_chosen_logps": -399.81439208984375,
      "debug/policy_rejected_logits": 31.01986312866211,
      "debug/policy_rejected_logps": -480.4059753417969,
      "debug/reference_chosen_logps": -429.8614501953125,
      "debug/reference_rejected_logps": -448.6155700683594,
      "epoch": 0.725,
      "grad_norm": 5.094791792265245,
      "learning_rate": 1e-06,
      "logits/chosen": 23.94160270690918,
      "logits/rejected": 31.01986312866211,
      "logps/chosen": -399.81439208984375,
      "logps/rejected": -480.4059753417969,
      "loss": 0.3252,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.30047041177749634,
      "rewards/margins": 0.6183742880821228,
      "rewards/rejected": -0.31790387630462646,
      "step": 58
    },
    {
      "debug/policy_chosen_logits": 22.9466495513916,
      "debug/policy_chosen_logps": -385.3292541503906,
      "debug/policy_rejected_logits": 36.00619888305664,
      "debug/policy_rejected_logps": -493.22198486328125,
      "debug/reference_chosen_logps": -421.3563232421875,
      "debug/reference_rejected_logps": -477.6571350097656,
      "epoch": 0.7375,
      "grad_norm": 5.69752289722826,
      "learning_rate": 1e-06,
      "logits/chosen": 22.9466495513916,
      "logits/rejected": 36.00619888305664,
      "logps/chosen": -385.3292541503906,
      "logps/rejected": -493.22198486328125,
      "loss": 0.2982,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.360270619392395,
      "rewards/margins": 0.515919029712677,
      "rewards/rejected": -0.1556483805179596,
      "step": 59
    },
    {
      "debug/policy_chosen_logits": 23.001218795776367,
      "debug/policy_chosen_logps": -398.92132568359375,
      "debug/policy_rejected_logits": 30.06394386291504,
      "debug/policy_rejected_logps": -459.4337463378906,
      "debug/reference_chosen_logps": -427.70135498046875,
      "debug/reference_rejected_logps": -427.77716064453125,
      "epoch": 0.75,
      "grad_norm": 3.651921352682424,
      "learning_rate": 1e-06,
      "logits/chosen": 23.001218795776367,
      "logits/rejected": 30.06394386291504,
      "logps/chosen": -398.92132568359375,
      "logps/rejected": -459.4337463378906,
      "loss": 0.26,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2878003716468811,
      "rewards/margins": 0.6043664216995239,
      "rewards/rejected": -0.3165660500526428,
      "step": 60
    },
    {
      "debug/policy_chosen_logits": 19.97127914428711,
      "debug/policy_chosen_logps": -396.8033447265625,
      "debug/policy_rejected_logits": 27.122671127319336,
      "debug/policy_rejected_logps": -485.0689697265625,
      "debug/reference_chosen_logps": -416.9521484375,
      "debug/reference_rejected_logps": -456.4365539550781,
      "epoch": 0.7625,
      "grad_norm": 3.768202921022212,
      "learning_rate": 1e-06,
      "logits/chosen": 19.97127914428711,
      "logits/rejected": 27.122671127319336,
      "logps/chosen": -396.8033447265625,
      "logps/rejected": -485.0689697265625,
      "loss": 0.2754,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.20148837566375732,
      "rewards/margins": 0.487812340259552,
      "rewards/rejected": -0.28632399439811707,
      "step": 61
    },
    {
      "debug/policy_chosen_logits": 25.002408981323242,
      "debug/policy_chosen_logps": -440.4338073730469,
      "debug/policy_rejected_logits": 29.166627883911133,
      "debug/policy_rejected_logps": -472.5847473144531,
      "debug/reference_chosen_logps": -458.51873779296875,
      "debug/reference_rejected_logps": -445.6849365234375,
      "epoch": 0.775,
      "grad_norm": 3.8018379677299614,
      "learning_rate": 1e-06,
      "logits/chosen": 25.002408981323242,
      "logits/rejected": 29.166627883911133,
      "logps/chosen": -440.4338073730469,
      "logps/rejected": -472.5847473144531,
      "loss": 0.2766,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.18084929883480072,
      "rewards/margins": 0.4498472809791565,
      "rewards/rejected": -0.26899799704551697,
      "step": 62
    },
    {
      "debug/policy_chosen_logits": 19.132322311401367,
      "debug/policy_chosen_logps": -405.4966735839844,
      "debug/policy_rejected_logits": 28.601638793945312,
      "debug/policy_rejected_logps": -506.548828125,
      "debug/reference_chosen_logps": -430.78936767578125,
      "debug/reference_rejected_logps": -464.448486328125,
      "epoch": 0.7875,
      "grad_norm": 3.9013050377479486,
      "learning_rate": 1e-06,
      "logits/chosen": 19.132322311401367,
      "logits/rejected": 28.601638793945312,
      "logps/chosen": -405.4966735839844,
      "logps/rejected": -506.548828125,
      "loss": 0.2766,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2529268264770508,
      "rewards/margins": 0.6739299297332764,
      "rewards/rejected": -0.42100316286087036,
      "step": 63
    },
    {
      "debug/policy_chosen_logits": 19.04224395751953,
      "debug/policy_chosen_logps": -399.923095703125,
      "debug/policy_rejected_logits": 29.6719970703125,
      "debug/policy_rejected_logps": -487.8897399902344,
      "debug/reference_chosen_logps": -422.9737243652344,
      "debug/reference_rejected_logps": -463.2634582519531,
      "epoch": 0.8,
      "grad_norm": 4.331208338864034,
      "learning_rate": 1e-06,
      "logits/chosen": 19.04224395751953,
      "logits/rejected": 29.6719970703125,
      "logps/chosen": -399.923095703125,
      "logps/rejected": -487.8897399902344,
      "loss": 0.2082,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.2305060774087906,
      "rewards/margins": 0.4767689108848572,
      "rewards/rejected": -0.2462628036737442,
      "step": 64
    },
    {
      "debug/policy_chosen_logits": 21.41056251525879,
      "debug/policy_chosen_logps": -421.7527770996094,
      "debug/policy_rejected_logits": 26.84547233581543,
      "debug/policy_rejected_logps": -474.83660888671875,
      "debug/reference_chosen_logps": -435.3292236328125,
      "debug/reference_rejected_logps": -446.2524108886719,
      "epoch": 0.8125,
      "grad_norm": 4.552945704427676,
      "learning_rate": 1e-06,
      "logits/chosen": 21.41056251525879,
      "logits/rejected": 26.84547233581543,
      "logps/chosen": -421.7527770996094,
      "logps/rejected": -474.83660888671875,
      "loss": 0.3366,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.13576430082321167,
      "rewards/margins": 0.421606183052063,
      "rewards/rejected": -0.2858418822288513,
      "step": 65
    },
    {
      "debug/policy_chosen_logits": 20.119802474975586,
      "debug/policy_chosen_logps": -381.8564147949219,
      "debug/policy_rejected_logits": 28.411861419677734,
      "debug/policy_rejected_logps": -466.48577880859375,
      "debug/reference_chosen_logps": -406.800048828125,
      "debug/reference_rejected_logps": -452.9040222167969,
      "epoch": 0.825,
      "grad_norm": 3.389829713086373,
      "learning_rate": 1e-06,
      "logits/chosen": 20.119802474975586,
      "logits/rejected": 28.411861419677734,
      "logps/chosen": -381.8564147949219,
      "logps/rejected": -466.48577880859375,
      "loss": 0.2748,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.24943621456623077,
      "rewards/margins": 0.385253369808197,
      "rewards/rejected": -0.13581714034080505,
      "step": 66
    },
    {
      "debug/policy_chosen_logits": 18.72797203063965,
      "debug/policy_chosen_logps": -390.49090576171875,
      "debug/policy_rejected_logits": 26.687557220458984,
      "debug/policy_rejected_logps": -512.882568359375,
      "debug/reference_chosen_logps": -418.24884033203125,
      "debug/reference_rejected_logps": -474.900390625,
      "epoch": 0.8375,
      "grad_norm": 3.623515504007148,
      "learning_rate": 1e-06,
      "logits/chosen": 18.72797203063965,
      "logits/rejected": 26.687557220458984,
      "logps/chosen": -390.49090576171875,
      "logps/rejected": -512.882568359375,
      "loss": 0.2272,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2775792181491852,
      "rewards/margins": 0.6574012637138367,
      "rewards/rejected": -0.3798220455646515,
      "step": 67
    },
    {
      "debug/policy_chosen_logits": 23.56059455871582,
      "debug/policy_chosen_logps": -410.64044189453125,
      "debug/policy_rejected_logits": 34.364986419677734,
      "debug/policy_rejected_logps": -478.304931640625,
      "debug/reference_chosen_logps": -428.95367431640625,
      "debug/reference_rejected_logps": -446.1668395996094,
      "epoch": 0.85,
      "grad_norm": 3.8118009157857644,
      "learning_rate": 1e-06,
      "logits/chosen": 23.56059455871582,
      "logits/rejected": 34.364986419677734,
      "logps/chosen": -410.64044189453125,
      "logps/rejected": -478.304931640625,
      "loss": 0.2653,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.18313205242156982,
      "rewards/margins": 0.5045130848884583,
      "rewards/rejected": -0.3213810324668884,
      "step": 68
    },
    {
      "debug/policy_chosen_logits": 20.668880462646484,
      "debug/policy_chosen_logps": -408.55419921875,
      "debug/policy_rejected_logits": 27.681875228881836,
      "debug/policy_rejected_logps": -442.37799072265625,
      "debug/reference_chosen_logps": -434.24542236328125,
      "debug/reference_rejected_logps": -441.95709228515625,
      "epoch": 0.8625,
      "grad_norm": 3.7604869397831107,
      "learning_rate": 1e-06,
      "logits/chosen": 20.668880462646484,
      "logits/rejected": 27.681875228881836,
      "logps/chosen": -408.55419921875,
      "logps/rejected": -442.37799072265625,
      "loss": 0.2795,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.2569122612476349,
      "rewards/margins": 0.2611211836338043,
      "rewards/rejected": -0.004208952188491821,
      "step": 69
    },
    {
      "debug/policy_chosen_logits": 25.379365921020508,
      "debug/policy_chosen_logps": -395.3825988769531,
      "debug/policy_rejected_logits": 31.113773345947266,
      "debug/policy_rejected_logps": -473.5229797363281,
      "debug/reference_chosen_logps": -419.5213623046875,
      "debug/reference_rejected_logps": -455.77142333984375,
      "epoch": 0.875,
      "grad_norm": 4.527244920289224,
      "learning_rate": 1e-06,
      "logits/chosen": 25.379365921020508,
      "logits/rejected": 31.113773345947266,
      "logps/chosen": -395.3825988769531,
      "logps/rejected": -473.5229797363281,
      "loss": 0.2967,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.24138759076595306,
      "rewards/margins": 0.41890305280685425,
      "rewards/rejected": -0.17751547694206238,
      "step": 70
    },
    {
      "debug/policy_chosen_logits": 19.842777252197266,
      "debug/policy_chosen_logps": -397.84136962890625,
      "debug/policy_rejected_logits": 28.35538101196289,
      "debug/policy_rejected_logps": -484.2126770019531,
      "debug/reference_chosen_logps": -415.0436096191406,
      "debug/reference_rejected_logps": -459.60400390625,
      "epoch": 0.8875,
      "grad_norm": 5.4667643203198075,
      "learning_rate": 1e-06,
      "logits/chosen": 19.842777252197266,
      "logits/rejected": 28.35538101196289,
      "logps/chosen": -397.84136962890625,
      "logps/rejected": -484.2126770019531,
      "loss": 0.2524,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.17202238738536835,
      "rewards/margins": 0.41810929775238037,
      "rewards/rejected": -0.24608686566352844,
      "step": 71
    },
    {
      "debug/policy_chosen_logits": 23.183486938476562,
      "debug/policy_chosen_logps": -400.0027160644531,
      "debug/policy_rejected_logits": 27.23314094543457,
      "debug/policy_rejected_logps": -473.27593994140625,
      "debug/reference_chosen_logps": -427.227783203125,
      "debug/reference_rejected_logps": -448.28363037109375,
      "epoch": 0.9,
      "grad_norm": 3.5365215033749213,
      "learning_rate": 1e-06,
      "logits/chosen": 23.183486938476562,
      "logits/rejected": 27.23314094543457,
      "logps/chosen": -400.0027160644531,
      "logps/rejected": -473.27593994140625,
      "loss": 0.2537,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.272250771522522,
      "rewards/margins": 0.5221738815307617,
      "rewards/rejected": -0.24992308020591736,
      "step": 72
    },
    {
      "debug/policy_chosen_logits": 20.710277557373047,
      "debug/policy_chosen_logps": -398.7235107421875,
      "debug/policy_rejected_logits": 24.029035568237305,
      "debug/policy_rejected_logps": -449.9461364746094,
      "debug/reference_chosen_logps": -419.6963195800781,
      "debug/reference_rejected_logps": -447.973876953125,
      "epoch": 0.9125,
      "grad_norm": 3.7346432080690812,
      "learning_rate": 1e-06,
      "logits/chosen": 20.710277557373047,
      "logits/rejected": 24.029035568237305,
      "logps/chosen": -398.7235107421875,
      "logps/rejected": -449.9461364746094,
      "loss": 0.2855,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.20972807705402374,
      "rewards/margins": 0.22945068776607513,
      "rewards/rejected": -0.019722595810890198,
      "step": 73
    },
    {
      "debug/policy_chosen_logits": 22.077816009521484,
      "debug/policy_chosen_logps": -401.99786376953125,
      "debug/policy_rejected_logits": 30.93896484375,
      "debug/policy_rejected_logps": -486.0145568847656,
      "debug/reference_chosen_logps": -420.620361328125,
      "debug/reference_rejected_logps": -452.76373291015625,
      "epoch": 0.925,
      "grad_norm": 4.320209693666537,
      "learning_rate": 1e-06,
      "logits/chosen": 22.077816009521484,
      "logits/rejected": 30.93896484375,
      "logps/chosen": -401.99786376953125,
      "logps/rejected": -486.0145568847656,
      "loss": 0.2651,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.18622498214244843,
      "rewards/margins": 0.5187329649925232,
      "rewards/rejected": -0.33250799775123596,
      "step": 74
    },
    {
      "debug/policy_chosen_logits": 16.68404197692871,
      "debug/policy_chosen_logps": -377.2335510253906,
      "debug/policy_rejected_logits": 31.94536781311035,
      "debug/policy_rejected_logps": -488.6978454589844,
      "debug/reference_chosen_logps": -400.86431884765625,
      "debug/reference_rejected_logps": -439.6689758300781,
      "epoch": 0.9375,
      "grad_norm": 3.3831819183050054,
      "learning_rate": 1e-06,
      "logits/chosen": 16.68404197692871,
      "logits/rejected": 31.94536781311035,
      "logps/chosen": -377.2335510253906,
      "logps/rejected": -488.6978454589844,
      "loss": 0.2509,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.23630763590335846,
      "rewards/margins": 0.7265963554382324,
      "rewards/rejected": -0.49028879404067993,
      "step": 75
    },
    {
      "debug/policy_chosen_logits": 24.752717971801758,
      "debug/policy_chosen_logps": -414.84014892578125,
      "debug/policy_rejected_logits": 25.71200942993164,
      "debug/policy_rejected_logps": -501.4398498535156,
      "debug/reference_chosen_logps": -441.64886474609375,
      "debug/reference_rejected_logps": -463.5169372558594,
      "epoch": 0.95,
      "grad_norm": 4.517327070548167,
      "learning_rate": 1e-06,
      "logits/chosen": 24.752717971801758,
      "logits/rejected": 25.71200942993164,
      "logps/chosen": -414.84014892578125,
      "logps/rejected": -501.4398498535156,
      "loss": 0.2746,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.26808685064315796,
      "rewards/margins": 0.6473158597946167,
      "rewards/rejected": -0.37922903895378113,
      "step": 76
    },
    {
      "debug/policy_chosen_logits": 17.914918899536133,
      "debug/policy_chosen_logps": -399.0069274902344,
      "debug/policy_rejected_logits": 22.21956443786621,
      "debug/policy_rejected_logps": -473.9609375,
      "debug/reference_chosen_logps": -425.5627136230469,
      "debug/reference_rejected_logps": -450.96319580078125,
      "epoch": 0.9625,
      "grad_norm": 3.1357429365118166,
      "learning_rate": 1e-06,
      "logits/chosen": 17.914918899536133,
      "logits/rejected": 22.21956443786621,
      "logps/chosen": -399.0069274902344,
      "logps/rejected": -473.9609375,
      "loss": 0.1999,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.26555800437927246,
      "rewards/margins": 0.4955354630947113,
      "rewards/rejected": -0.22997744381427765,
      "step": 77
    },
    {
      "debug/policy_chosen_logits": 18.327871322631836,
      "debug/policy_chosen_logps": -370.01171875,
      "debug/policy_rejected_logits": 28.05528450012207,
      "debug/policy_rejected_logps": -427.2958984375,
      "debug/reference_chosen_logps": -400.8182373046875,
      "debug/reference_rejected_logps": -420.2340393066406,
      "epoch": 0.975,
      "grad_norm": 4.140874548230489,
      "learning_rate": 1e-06,
      "logits/chosen": 18.327871322631836,
      "logits/rejected": 28.05528450012207,
      "logps/chosen": -370.01171875,
      "logps/rejected": -427.2958984375,
      "loss": 0.2728,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3080649673938751,
      "rewards/margins": 0.37868380546569824,
      "rewards/rejected": -0.07061880826950073,
      "step": 78
    },
    {
      "debug/policy_chosen_logits": 22.553178787231445,
      "debug/policy_chosen_logps": -408.61163330078125,
      "debug/policy_rejected_logits": 23.832901000976562,
      "debug/policy_rejected_logps": -502.69970703125,
      "debug/reference_chosen_logps": -414.4495544433594,
      "debug/reference_rejected_logps": -477.4999694824219,
      "epoch": 0.9875,
      "grad_norm": 3.8978087275053612,
      "learning_rate": 1e-06,
      "logits/chosen": 22.553178787231445,
      "logits/rejected": 23.832901000976562,
      "logps/chosen": -408.61163330078125,
      "logps/rejected": -502.69970703125,
      "loss": 0.271,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.05837925150990486,
      "rewards/margins": 0.31037673354148865,
      "rewards/rejected": -0.2519974708557129,
      "step": 79
    },
    {
      "debug/policy_chosen_logits": 20.476898193359375,
      "debug/policy_chosen_logps": -401.72320556640625,
      "debug/policy_rejected_logits": 29.332805633544922,
      "debug/policy_rejected_logps": -478.8862609863281,
      "debug/reference_chosen_logps": -428.4986572265625,
      "debug/reference_rejected_logps": -465.5113220214844,
      "epoch": 1.0,
      "grad_norm": 4.733733434133006,
      "learning_rate": 1e-06,
      "logits/chosen": 20.476898193359375,
      "logits/rejected": 29.332805633544922,
      "logps/chosen": -401.72320556640625,
      "logps/rejected": -478.8862609863281,
      "loss": 0.2109,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.267754465341568,
      "rewards/margins": 0.40150386095046997,
      "rewards/rejected": -0.13374939560890198,
      "step": 80
    },
    {
      "epoch": 1.0,
      "step": 80,
      "total_flos": 0.0,
      "train_loss": 0.32403188310563563,
      "train_runtime": 674.2321,
      "train_samples_per_second": 7.592,
      "train_steps_per_second": 0.119
    }
  ],
  "logging_steps": 1,
  "max_steps": 80,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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
}