{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 50,
  "global_step": 436,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022935779816513763,
      "grad_norm": 4.5521214161319365,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.658930778503418,
      "logits/rejected": -2.6127209663391113,
      "logps/chosen": -310.2911682128906,
      "logps/rejected": -241.6273651123047,
      "loss": 0.6931,
      "rewards/accuracies": 0.3812499940395355,
      "rewards/chosen": -0.0002677589072845876,
      "rewards/margins": -0.0003533684357535094,
      "rewards/rejected": 8.560952846892178e-05,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 5.31294051195239,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.691143751144409,
      "logits/rejected": -2.6153030395507812,
      "logps/chosen": -293.57037353515625,
      "logps/rejected": -265.63885498046875,
      "loss": 0.6926,
      "rewards/accuracies": 0.5562499761581421,
      "rewards/chosen": 0.00123660231474787,
      "rewards/margins": 0.00207112031057477,
      "rewards/rejected": -0.0008345181122422218,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 4.285586645859575,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6962637901306152,
      "logits/rejected": -2.628978729248047,
      "logps/chosen": -277.83209228515625,
      "logps/rejected": -297.1653137207031,
      "loss": 0.6901,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": 0.004098366480320692,
      "rewards/margins": 0.00956634059548378,
      "rewards/rejected": -0.0054679736495018005,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 5.111796420680735,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.6144208908081055,
      "logits/rejected": -2.543553352355957,
      "logps/chosen": -284.1578674316406,
      "logps/rejected": -259.81475830078125,
      "loss": 0.6826,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.03460276871919632,
      "rewards/margins": 0.0436362698674202,
      "rewards/rejected": -0.009033503010869026,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 4.717770708541697,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.591963291168213,
      "logits/rejected": -2.5082640647888184,
      "logps/chosen": -285.80047607421875,
      "logps/rejected": -247.5881805419922,
      "loss": 0.674,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": -0.011667141690850258,
      "rewards/margins": 0.055136788636446,
      "rewards/rejected": -0.06680391728878021,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.6023333072662354,
      "eval_logits/rejected": -2.5038623809814453,
      "eval_logps/chosen": -288.564208984375,
      "eval_logps/rejected": -260.0650329589844,
      "eval_loss": 0.6643866300582886,
      "eval_rewards/accuracies": 0.6724137663841248,
      "eval_rewards/chosen": -0.04558267444372177,
      "eval_rewards/margins": 0.09859247505664825,
      "eval_rewards/rejected": -0.14417517185211182,
      "eval_runtime": 91.2705,
      "eval_samples_per_second": 19.919,
      "eval_steps_per_second": 0.318,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 6.057042762669634,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.584840774536133,
      "logits/rejected": -2.5118155479431152,
      "logps/chosen": -291.8733215332031,
      "logps/rejected": -281.57476806640625,
      "loss": 0.6519,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.11285074800252914,
      "rewards/margins": 0.18535511195659637,
      "rewards/rejected": -0.2982058525085449,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 12.297420969455343,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.4776272773742676,
      "logits/rejected": -2.4097015857696533,
      "logps/chosen": -299.2547912597656,
      "logps/rejected": -271.7998046875,
      "loss": 0.6509,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.30904772877693176,
      "rewards/margins": 0.21535193920135498,
      "rewards/rejected": -0.5243996381759644,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 8.87671354702832,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.565410852432251,
      "logits/rejected": -2.4615464210510254,
      "logps/chosen": -310.5340270996094,
      "logps/rejected": -324.5120544433594,
      "loss": 0.637,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.40506935119628906,
      "rewards/margins": 0.2631128430366516,
      "rewards/rejected": -0.6681821942329407,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 13.125498153678837,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -1.73556649684906,
      "logits/rejected": -1.575148105621338,
      "logps/chosen": -354.5548400878906,
      "logps/rejected": -374.177978515625,
      "loss": 0.6117,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.5037232041358948,
      "rewards/margins": 0.48656755685806274,
      "rewards/rejected": -0.9902908205986023,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 13.953968120873695,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -0.5573834180831909,
      "logits/rejected": -0.19994211196899414,
      "logps/chosen": -354.9537048339844,
      "logps/rejected": -379.1808166503906,
      "loss": 0.5874,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.7568774819374084,
      "rewards/margins": 0.5189547538757324,
      "rewards/rejected": -1.2758322954177856,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -0.22263701260089874,
      "eval_logits/rejected": 0.3007841110229492,
      "eval_logps/chosen": -382.2076416015625,
      "eval_logps/rejected": -402.1482238769531,
      "eval_loss": 0.5920498371124268,
      "eval_rewards/accuracies": 0.681034505367279,
      "eval_rewards/chosen": -0.9820166230201721,
      "eval_rewards/margins": 0.5829907059669495,
      "eval_rewards/rejected": -1.5650073289871216,
      "eval_runtime": 91.5375,
      "eval_samples_per_second": 19.861,
      "eval_steps_per_second": 0.317,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 23.915856678856265,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -0.40947216749191284,
      "logits/rejected": 0.038003671914339066,
      "logps/chosen": -379.87451171875,
      "logps/rejected": -438.7657165527344,
      "loss": 0.5763,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.826363742351532,
      "rewards/margins": 0.7436057329177856,
      "rewards/rejected": -1.5699694156646729,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 14.304332284980681,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": -0.18220779299736023,
      "logits/rejected": 0.7561357021331787,
      "logps/chosen": -424.419677734375,
      "logps/rejected": -439.09320068359375,
      "loss": 0.5937,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -1.1400790214538574,
      "rewards/margins": 0.6504500508308411,
      "rewards/rejected": -1.7905290126800537,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 15.490141656078896,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": 0.37701982259750366,
      "logits/rejected": 1.372661828994751,
      "logps/chosen": -391.7457580566406,
      "logps/rejected": -451.080810546875,
      "loss": 0.5829,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.0409055948257446,
      "rewards/margins": 0.9019233584403992,
      "rewards/rejected": -1.942828893661499,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 15.271366347742205,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": -0.13376641273498535,
      "logits/rejected": 1.0010212659835815,
      "logps/chosen": -383.65301513671875,
      "logps/rejected": -419.6103515625,
      "loss": 0.5623,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.9957970380783081,
      "rewards/margins": 0.7610489130020142,
      "rewards/rejected": -1.7568460702896118,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 22.33400136982847,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": 0.09550192952156067,
      "logits/rejected": 1.2170594930648804,
      "logps/chosen": -387.51849365234375,
      "logps/rejected": -434.84136962890625,
      "loss": 0.5612,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.9680271148681641,
      "rewards/margins": 0.772575318813324,
      "rewards/rejected": -1.7406024932861328,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": 1.4310089349746704,
      "eval_logits/rejected": 2.314030885696411,
      "eval_logps/chosen": -430.7732238769531,
      "eval_logps/rejected": -482.2997741699219,
      "eval_loss": 0.5694876909255981,
      "eval_rewards/accuracies": 0.6896551847457886,
      "eval_rewards/chosen": -1.4676721096038818,
      "eval_rewards/margins": 0.8988507390022278,
      "eval_rewards/rejected": -2.366523027420044,
      "eval_runtime": 91.6244,
      "eval_samples_per_second": 19.842,
      "eval_steps_per_second": 0.317,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 23.657529687976723,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": 1.280305027961731,
      "logits/rejected": 2.0924124717712402,
      "logps/chosen": -457.1259765625,
      "logps/rejected": -502.7013244628906,
      "loss": 0.5762,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -1.6072089672088623,
      "rewards/margins": 0.868707001209259,
      "rewards/rejected": -2.4759163856506348,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 20.186960313509775,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": -0.41927653551101685,
      "logits/rejected": 0.5089612007141113,
      "logps/chosen": -360.0221252441406,
      "logps/rejected": -393.9572448730469,
      "loss": 0.5498,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.9409689903259277,
      "rewards/margins": 0.7842710614204407,
      "rewards/rejected": -1.7252399921417236,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 19.134949732610515,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": -0.13538230955600739,
      "logits/rejected": 0.8745505213737488,
      "logps/chosen": -365.5416259765625,
      "logps/rejected": -449.64080810546875,
      "loss": 0.5571,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -1.0540037155151367,
      "rewards/margins": 1.0069358348846436,
      "rewards/rejected": -2.060939311981201,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 18.989463695568457,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": 0.555556058883667,
      "logits/rejected": 1.7799046039581299,
      "logps/chosen": -379.05133056640625,
      "logps/rejected": -495.5801696777344,
      "loss": 0.5566,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -1.2038367986679077,
      "rewards/margins": 1.1251088380813599,
      "rewards/rejected": -2.3289456367492676,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 20.721540306309656,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": 0.1556408703327179,
      "logits/rejected": 1.0401029586791992,
      "logps/chosen": -403.26068115234375,
      "logps/rejected": -452.82684326171875,
      "loss": 0.5427,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -1.3739937543869019,
      "rewards/margins": 0.7915691137313843,
      "rewards/rejected": -2.1655631065368652,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": -0.3630250096321106,
      "eval_logits/rejected": 0.9222813248634338,
      "eval_logps/chosen": -418.6947021484375,
      "eval_logps/rejected": -471.8922119140625,
      "eval_loss": 0.5523015260696411,
      "eval_rewards/accuracies": 0.7241379022598267,
      "eval_rewards/chosen": -1.346887469291687,
      "eval_rewards/margins": 0.9155599474906921,
      "eval_rewards/rejected": -2.2624475955963135,
      "eval_runtime": 92.3698,
      "eval_samples_per_second": 19.682,
      "eval_steps_per_second": 0.314,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 17.428734986299386,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": 0.16718950867652893,
      "logits/rejected": 1.1019595861434937,
      "logps/chosen": -430.6678771972656,
      "logps/rejected": -517.311767578125,
      "loss": 0.5458,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.5387192964553833,
      "rewards/margins": 0.787868857383728,
      "rewards/rejected": -2.3265881538391113,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 18.835466832309084,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": -0.19043000042438507,
      "logits/rejected": 1.0126752853393555,
      "logps/chosen": -419.5379943847656,
      "logps/rejected": -484.6988830566406,
      "loss": 0.5528,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.511856198310852,
      "rewards/margins": 0.7948068976402283,
      "rewards/rejected": -2.3066630363464355,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 23.197020188064432,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": -0.30360710620880127,
      "logits/rejected": 1.0719916820526123,
      "logps/chosen": -386.96087646484375,
      "logps/rejected": -477.74029541015625,
      "loss": 0.5482,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -1.1279244422912598,
      "rewards/margins": 1.1136093139648438,
      "rewards/rejected": -2.2415337562561035,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 23.485163773762974,
      "learning_rate": 2.5e-07,
      "logits/chosen": 0.9384336471557617,
      "logits/rejected": 1.5441758632659912,
      "logps/chosen": -413.4400329589844,
      "logps/rejected": -505.8965759277344,
      "loss": 0.5778,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.4366745948791504,
      "rewards/margins": 0.9647194147109985,
      "rewards/rejected": -2.4013938903808594,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 18.02833446289089,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": 0.6526413559913635,
      "logits/rejected": 1.6761877536773682,
      "logps/chosen": -408.7395324707031,
      "logps/rejected": -484.34417724609375,
      "loss": 0.5474,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.438388705253601,
      "rewards/margins": 0.9599205851554871,
      "rewards/rejected": -2.3983092308044434,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": -0.44025513529777527,
      "eval_logits/rejected": 0.9070881605148315,
      "eval_logps/chosen": -393.58612060546875,
      "eval_logps/rejected": -449.3501281738281,
      "eval_loss": 0.5429869890213013,
      "eval_rewards/accuracies": 0.6896551847457886,
      "eval_rewards/chosen": -1.0958014726638794,
      "eval_rewards/margins": 0.9412251114845276,
      "eval_rewards/rejected": -2.0370266437530518,
      "eval_runtime": 91.0397,
      "eval_samples_per_second": 19.969,
      "eval_steps_per_second": 0.319,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 21.769039546192015,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": 0.1161709800362587,
      "logits/rejected": 1.337697982788086,
      "logps/chosen": -419.63995361328125,
      "logps/rejected": -492.03363037109375,
      "loss": 0.5494,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -1.1713824272155762,
      "rewards/margins": 1.1571643352508545,
      "rewards/rejected": -2.3285470008850098,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 19.115539205351443,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": 1.2381702661514282,
      "logits/rejected": 2.2026009559631348,
      "logps/chosen": -462.31072998046875,
      "logps/rejected": -520.6209716796875,
      "loss": 0.5714,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -1.8605678081512451,
      "rewards/margins": 0.8712261319160461,
      "rewards/rejected": -2.7317938804626465,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 20.348758953970556,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": 1.1374260187149048,
      "logits/rejected": 2.329713821411133,
      "logps/chosen": -465.6629333496094,
      "logps/rejected": -563.9349365234375,
      "loss": 0.543,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.8833774328231812,
      "rewards/margins": 1.0653350353240967,
      "rewards/rejected": -2.948712110519409,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 18.788678367954812,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": -0.17822352051734924,
      "logits/rejected": 0.844377338886261,
      "logps/chosen": -437.96435546875,
      "logps/rejected": -507.1295471191406,
      "loss": 0.542,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.4787204265594482,
      "rewards/margins": 0.8510695695877075,
      "rewards/rejected": -2.329789876937866,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 22.907322461677648,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": 0.011055344715714455,
      "logits/rejected": 1.3963500261306763,
      "logps/chosen": -432.23321533203125,
      "logps/rejected": -509.43939208984375,
      "loss": 0.5556,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.4701194763183594,
      "rewards/margins": 1.124694585800171,
      "rewards/rejected": -2.5948143005371094,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": -0.1993160843849182,
      "eval_logits/rejected": 1.1949646472930908,
      "eval_logps/chosen": -423.5918884277344,
      "eval_logps/rejected": -484.2666320800781,
      "eval_loss": 0.5403606295585632,
      "eval_rewards/accuracies": 0.7198275923728943,
      "eval_rewards/chosen": -1.3958592414855957,
      "eval_rewards/margins": 0.9903314709663391,
      "eval_rewards/rejected": -2.38619065284729,
      "eval_runtime": 90.6243,
      "eval_samples_per_second": 20.061,
      "eval_steps_per_second": 0.32,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 27.111269521337316,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": 0.3549458086490631,
      "logits/rejected": 1.506676435470581,
      "logps/chosen": -407.77020263671875,
      "logps/rejected": -494.737548828125,
      "loss": 0.5603,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -1.536117434501648,
      "rewards/margins": 0.9887636303901672,
      "rewards/rejected": -2.524880886077881,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 20.93281634880908,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": -0.10006646066904068,
      "logits/rejected": 1.6550134420394897,
      "logps/chosen": -442.34283447265625,
      "logps/rejected": -508.07061767578125,
      "loss": 0.5395,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.3644827604293823,
      "rewards/margins": 1.2163000106811523,
      "rewards/rejected": -2.580782651901245,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 22.09617953181825,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": 0.12130077183246613,
      "logits/rejected": 1.2904590368270874,
      "logps/chosen": -437.9993591308594,
      "logps/rejected": -525.1572875976562,
      "loss": 0.5394,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -1.592714548110962,
      "rewards/margins": 0.9397724270820618,
      "rewards/rejected": -2.5324866771698,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 21.35703312189058,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": 0.1835191398859024,
      "logits/rejected": 1.8147304058074951,
      "logps/chosen": -403.4627380371094,
      "logps/rejected": -524.2403564453125,
      "loss": 0.5415,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -1.4125099182128906,
      "rewards/margins": 1.2409336566925049,
      "rewards/rejected": -2.6534433364868164,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 21.04805960149964,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": 0.23725076019763947,
      "logits/rejected": 2.2347147464752197,
      "logps/chosen": -436.37939453125,
      "logps/rejected": -537.32373046875,
      "loss": 0.5373,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -1.4519966840744019,
      "rewards/margins": 1.4744818210601807,
      "rewards/rejected": -2.926478624343872,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": 0.23955857753753662,
      "eval_logits/rejected": 1.7159334421157837,
      "eval_logps/chosen": -439.8386535644531,
      "eval_logps/rejected": -505.62298583984375,
      "eval_loss": 0.5416247844696045,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -1.5583269596099854,
      "eval_rewards/margins": 1.041427493095398,
      "eval_rewards/rejected": -2.599754810333252,
      "eval_runtime": 92.2289,
      "eval_samples_per_second": 19.712,
      "eval_steps_per_second": 0.314,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 28.15381228175136,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": 0.37452951073646545,
      "logits/rejected": 1.507530689239502,
      "logps/chosen": -437.071044921875,
      "logps/rejected": -533.8405151367188,
      "loss": 0.535,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.5732779502868652,
      "rewards/margins": 1.106673002243042,
      "rewards/rejected": -2.679950714111328,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 17.904935938367142,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": -0.39112910628318787,
      "logits/rejected": 1.349570631980896,
      "logps/chosen": -471.1695861816406,
      "logps/rejected": -532.53271484375,
      "loss": 0.5285,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -1.3933590650558472,
      "rewards/margins": 1.198838472366333,
      "rewards/rejected": -2.5921976566314697,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 21.395515100278608,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": -0.16580775380134583,
      "logits/rejected": 1.5711965560913086,
      "logps/chosen": -442.42633056640625,
      "logps/rejected": -471.59808349609375,
      "loss": 0.5345,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -1.4915008544921875,
      "rewards/margins": 0.9731476902961731,
      "rewards/rejected": -2.464648723602295,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 18.765474571368344,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": -0.5060194730758667,
      "logits/rejected": 1.3254286050796509,
      "logps/chosen": -439.9669494628906,
      "logps/rejected": -542.6121826171875,
      "loss": 0.5238,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -1.3199481964111328,
      "rewards/margins": 1.575811743736267,
      "rewards/rejected": -2.8957600593566895,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 22.250966254695054,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": 0.07749566435813904,
      "logits/rejected": 1.2516423463821411,
      "logps/chosen": -449.790283203125,
      "logps/rejected": -540.7836303710938,
      "loss": 0.5405,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -1.5793349742889404,
      "rewards/margins": 0.9962536692619324,
      "rewards/rejected": -2.5755887031555176,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": -0.2526824176311493,
      "eval_logits/rejected": 1.2749311923980713,
      "eval_logps/chosen": -423.50390625,
      "eval_logps/rejected": -490.2825927734375,
      "eval_loss": 0.5382627844810486,
      "eval_rewards/accuracies": 0.7241379022598267,
      "eval_rewards/chosen": -1.3949792385101318,
      "eval_rewards/margins": 1.0513713359832764,
      "eval_rewards/rejected": -2.446350574493408,
      "eval_runtime": 90.812,
      "eval_samples_per_second": 20.019,
      "eval_steps_per_second": 0.319,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 20.214960371040412,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": 0.10587388277053833,
      "logits/rejected": 1.2455976009368896,
      "logps/chosen": -443.05621337890625,
      "logps/rejected": -499.88531494140625,
      "loss": 0.551,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.5565433502197266,
      "rewards/margins": 0.8829916715621948,
      "rewards/rejected": -2.4395346641540527,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 24.19047240295495,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": -0.0691867247223854,
      "logits/rejected": 1.6679766178131104,
      "logps/chosen": -455.8050231933594,
      "logps/rejected": -492.05401611328125,
      "loss": 0.5468,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -1.5099737644195557,
      "rewards/margins": 1.0747706890106201,
      "rewards/rejected": -2.584744691848755,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 16.600958628723166,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": 0.0887439101934433,
      "logits/rejected": 1.3473814725875854,
      "logps/chosen": -406.8893737792969,
      "logps/rejected": -502.24786376953125,
      "loss": 0.5367,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.4497406482696533,
      "rewards/margins": 1.044012188911438,
      "rewards/rejected": -2.4937527179718018,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.5759701597581216,
      "train_runtime": 11541.3114,
      "train_samples_per_second": 4.831,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 10,
  "max_steps": 436,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "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
}