{
  "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": 5.353972534143438,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.6582446098327637,
      "logits/rejected": -2.612395763397217,
      "logps/chosen": -310.3081359863281,
      "logps/rejected": -241.6246337890625,
      "loss": 0.6932,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": -0.00043685571290552616,
      "rewards/margins": -0.0005496515659615397,
      "rewards/rejected": 0.0001127958094002679,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 6.431385284276218,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.690976142883301,
      "logits/rejected": -2.615501880645752,
      "logps/chosen": -293.55859375,
      "logps/rejected": -265.65789794921875,
      "loss": 0.6924,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.001354431384243071,
      "rewards/margins": 0.0023786118254065514,
      "rewards/rejected": -0.0010241802083328366,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 5.140938328988767,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6976418495178223,
      "logits/rejected": -2.6304168701171875,
      "logps/chosen": -277.8341064453125,
      "logps/rejected": -297.1772155761719,
      "loss": 0.6892,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.004078004974871874,
      "rewards/margins": 0.009664928540587425,
      "rewards/rejected": -0.005586923565715551,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 5.971632655809275,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.616170883178711,
      "logits/rejected": -2.5451369285583496,
      "logps/chosen": -283.9632568359375,
      "logps/rejected": -259.82861328125,
      "loss": 0.6798,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.03654901683330536,
      "rewards/margins": 0.045721281319856644,
      "rewards/rejected": -0.00917226541787386,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 5.916656852320022,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.5970985889434814,
      "logits/rejected": -2.5133914947509766,
      "logps/chosen": -285.24835205078125,
      "logps/rejected": -247.303466796875,
      "loss": 0.6688,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": -0.006145569030195475,
      "rewards/margins": 0.0578111931681633,
      "rewards/rejected": -0.06395676732063293,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.6066324710845947,
      "eval_logits/rejected": -2.506901979446411,
      "eval_logps/chosen": -286.6465759277344,
      "eval_logps/rejected": -258.62078857421875,
      "eval_loss": 0.6561177968978882,
      "eval_rewards/accuracies": 0.6767241358757019,
      "eval_rewards/chosen": -0.02640603668987751,
      "eval_rewards/margins": 0.10332722216844559,
      "eval_rewards/rejected": -0.12973324954509735,
      "eval_runtime": 91.0244,
      "eval_samples_per_second": 19.973,
      "eval_steps_per_second": 0.319,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 7.499634288772489,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.582371234893799,
      "logits/rejected": -2.5081627368927,
      "logps/chosen": -292.10491943359375,
      "logps/rejected": -282.31195068359375,
      "loss": 0.6423,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.11516664922237396,
      "rewards/margins": 0.19041100144386292,
      "rewards/rejected": -0.30557766556739807,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 17.30037068758165,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.4502875804901123,
      "logits/rejected": -2.3790054321289062,
      "logps/chosen": -298.32244873046875,
      "logps/rejected": -273.11859130859375,
      "loss": 0.6397,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.2997247576713562,
      "rewards/margins": 0.23786215484142303,
      "rewards/rejected": -0.5375869870185852,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 11.228813057299567,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.5100908279418945,
      "logits/rejected": -2.4027259349823,
      "logps/chosen": -306.67510986328125,
      "logps/rejected": -322.7925720214844,
      "loss": 0.6257,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.36647987365722656,
      "rewards/margins": 0.28450754284858704,
      "rewards/rejected": -0.6509873867034912,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 15.58920411413326,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -1.5505931377410889,
      "logits/rejected": -1.3694034814834595,
      "logps/chosen": -357.6716613769531,
      "logps/rejected": -372.05133056640625,
      "loss": 0.5967,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.5348917245864868,
      "rewards/margins": 0.4341323971748352,
      "rewards/rejected": -0.9690243005752563,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 13.006583087547677,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -0.7837198972702026,
      "logits/rejected": -0.35428792238235474,
      "logps/chosen": -326.9918518066406,
      "logps/rejected": -346.36737060546875,
      "loss": 0.5822,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.4772585332393646,
      "rewards/margins": 0.4704399108886719,
      "rewards/rejected": -0.9476984143257141,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -0.09357786923646927,
      "eval_logits/rejected": 0.4795497953891754,
      "eval_logps/chosen": -374.6986083984375,
      "eval_logps/rejected": -392.9424743652344,
      "eval_loss": 0.5813368558883667,
      "eval_rewards/accuracies": 0.6724137663841248,
      "eval_rewards/chosen": -0.9069267511367798,
      "eval_rewards/margins": 0.5660232305526733,
      "eval_rewards/rejected": -1.4729499816894531,
      "eval_runtime": 91.4662,
      "eval_samples_per_second": 19.876,
      "eval_steps_per_second": 0.317,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 14.7655268267239,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -0.23613190650939941,
      "logits/rejected": 0.2948758006095886,
      "logps/chosen": -371.15667724609375,
      "logps/rejected": -427.76885986328125,
      "loss": 0.5606,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.7391853332519531,
      "rewards/margins": 0.7208150625228882,
      "rewards/rejected": -1.4600005149841309,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 28.232913631245626,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": 0.3172193467617035,
      "logits/rejected": 1.228100299835205,
      "logps/chosen": -412.1929626464844,
      "logps/rejected": -428.08056640625,
      "loss": 0.5789,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -1.0178120136260986,
      "rewards/margins": 0.6625908613204956,
      "rewards/rejected": -1.6804027557373047,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 23.106046920597972,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": 1.296276330947876,
      "logits/rejected": 2.0952706336975098,
      "logps/chosen": -377.18572998046875,
      "logps/rejected": -435.5022888183594,
      "loss": 0.5712,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.8953048586845398,
      "rewards/margins": 0.8917394876480103,
      "rewards/rejected": -1.7870445251464844,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 21.3617509080007,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": 1.0354318618774414,
      "logits/rejected": 2.103768825531006,
      "logps/chosen": -374.42938232421875,
      "logps/rejected": -408.32342529296875,
      "loss": 0.5477,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.9035611152648926,
      "rewards/margins": 0.7404158115386963,
      "rewards/rejected": -1.643977165222168,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 21.53708307235743,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": 0.8976553678512573,
      "logits/rejected": 2.0599629878997803,
      "logps/chosen": -370.0615234375,
      "logps/rejected": -411.820068359375,
      "loss": 0.5512,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.7934576869010925,
      "rewards/margins": 0.716931939125061,
      "rewards/rejected": -1.5103896856307983,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": 1.9097994565963745,
      "eval_logits/rejected": 2.9840593338012695,
      "eval_logps/chosen": -382.4128112792969,
      "eval_logps/rejected": -426.216552734375,
      "eval_loss": 0.5533820390701294,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -0.9840683937072754,
      "eval_rewards/margins": 0.8216219544410706,
      "eval_rewards/rejected": -1.8056902885437012,
      "eval_runtime": 91.5586,
      "eval_samples_per_second": 19.856,
      "eval_steps_per_second": 0.317,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 17.634177990437703,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": 1.8957267999649048,
      "logits/rejected": 2.8357367515563965,
      "logps/chosen": -408.04705810546875,
      "logps/rejected": -450.3290100097656,
      "loss": 0.5609,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.1164201498031616,
      "rewards/margins": 0.8357731103897095,
      "rewards/rejected": -1.952193021774292,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 24.6798606158854,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": 1.3192155361175537,
      "logits/rejected": 2.3843648433685303,
      "logps/chosen": -356.1672058105469,
      "logps/rejected": -387.1358947753906,
      "loss": 0.5414,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.9024198651313782,
      "rewards/margins": 0.7546060681343079,
      "rewards/rejected": -1.6570260524749756,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 22.218831723434445,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": 1.3455697298049927,
      "logits/rejected": 2.5438590049743652,
      "logps/chosen": -351.62774658203125,
      "logps/rejected": -430.138427734375,
      "loss": 0.5396,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.9148648381233215,
      "rewards/margins": 0.9510505795478821,
      "rewards/rejected": -1.8659156560897827,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 25.297690497973328,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": 1.7387921810150146,
      "logits/rejected": 2.998396396636963,
      "logps/chosen": -333.116455078125,
      "logps/rejected": -437.8639221191406,
      "loss": 0.5399,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.7444877624511719,
      "rewards/margins": 1.007294774055481,
      "rewards/rejected": -1.7517824172973633,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 25.947440207471573,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": 1.9480648040771484,
      "logits/rejected": 2.9033870697021484,
      "logps/chosen": -380.21185302734375,
      "logps/rejected": -424.3128967285156,
      "loss": 0.5364,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -1.1435054540634155,
      "rewards/margins": 0.736918032169342,
      "rewards/rejected": -1.8804235458374023,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": 2.4303267002105713,
      "eval_logits/rejected": 3.876626491546631,
      "eval_logps/chosen": -425.8599853515625,
      "eval_logps/rejected": -481.8061828613281,
      "eval_loss": 0.5367683172225952,
      "eval_rewards/accuracies": 0.732758641242981,
      "eval_rewards/chosen": -1.4185398817062378,
      "eval_rewards/margins": 0.9430465698242188,
      "eval_rewards/rejected": -2.361586570739746,
      "eval_runtime": 91.0832,
      "eval_samples_per_second": 19.96,
      "eval_steps_per_second": 0.318,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 19.563478098052215,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": 3.3911328315734863,
      "logits/rejected": 4.194566249847412,
      "logps/chosen": -451.25238037109375,
      "logps/rejected": -539.4793701171875,
      "loss": 0.5321,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.744564414024353,
      "rewards/margins": 0.8036998510360718,
      "rewards/rejected": -2.548264503479004,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 24.94888206530693,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": 2.1577422618865967,
      "logits/rejected": 3.670943021774292,
      "logps/chosen": -426.3487854003906,
      "logps/rejected": -508.80780029296875,
      "loss": 0.5365,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -1.5799643993377686,
      "rewards/margins": 0.9677878618240356,
      "rewards/rejected": -2.5477521419525146,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 23.826139773858404,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": 0.8777297735214233,
      "logits/rejected": 2.3443570137023926,
      "logps/chosen": -384.8174743652344,
      "logps/rejected": -467.87298583984375,
      "loss": 0.5416,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.1064906120300293,
      "rewards/margins": 1.0363706350326538,
      "rewards/rejected": -2.1428613662719727,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 18.150746967508407,
      "learning_rate": 2.5e-07,
      "logits/chosen": 2.9278922080993652,
      "logits/rejected": 3.511791706085205,
      "logps/chosen": -450.8653259277344,
      "logps/rejected": -547.5108642578125,
      "loss": 0.5638,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -1.810927391052246,
      "rewards/margins": 1.0066088438034058,
      "rewards/rejected": -2.8175363540649414,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 21.73286552315769,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": 2.6097209453582764,
      "logits/rejected": 3.895547389984131,
      "logps/chosen": -484.319091796875,
      "logps/rejected": -561.3115234375,
      "loss": 0.5308,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -2.1941843032836914,
      "rewards/margins": 0.9737985730171204,
      "rewards/rejected": -3.167982578277588,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": 2.454780101776123,
      "eval_logits/rejected": 4.2408928871154785,
      "eval_logps/chosen": -519.5304565429688,
      "eval_logps/rejected": -602.6265869140625,
      "eval_loss": 0.5234553217887878,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -2.3552448749542236,
      "eval_rewards/margins": 1.2145458459854126,
      "eval_rewards/rejected": -3.5697906017303467,
      "eval_runtime": 91.7528,
      "eval_samples_per_second": 19.814,
      "eval_steps_per_second": 0.316,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 29.190042715796082,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": 2.5975215435028076,
      "logits/rejected": 4.360453128814697,
      "logps/chosen": -554.3533935546875,
      "logps/rejected": -655.7716064453125,
      "loss": 0.5307,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -2.5185165405273438,
      "rewards/margins": 1.447409987449646,
      "rewards/rejected": -3.9659264087677,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 20.081922974931803,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": 2.110996723175049,
      "logits/rejected": 3.4121272563934326,
      "logps/chosen": -542.6881103515625,
      "logps/rejected": -613.2174682617188,
      "loss": 0.5514,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -2.664341449737549,
      "rewards/margins": 0.9934176206588745,
      "rewards/rejected": -3.657759189605713,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 20.797156741141926,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": 1.95541250705719,
      "logits/rejected": 3.574702024459839,
      "logps/chosen": -541.0426025390625,
      "logps/rejected": -644.5146484375,
      "loss": 0.522,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -2.637174129486084,
      "rewards/margins": 1.117336630821228,
      "rewards/rejected": -3.7545104026794434,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 22.250484161252675,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": 1.9210926294326782,
      "logits/rejected": 3.2735812664031982,
      "logps/chosen": -576.992431640625,
      "logps/rejected": -663.1363525390625,
      "loss": 0.5191,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -2.8690013885498047,
      "rewards/margins": 1.0208569765090942,
      "rewards/rejected": -3.8898582458496094,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 24.39105290704078,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": 1.646095871925354,
      "logits/rejected": 3.443913221359253,
      "logps/chosen": -558.287353515625,
      "logps/rejected": -657.3651123046875,
      "loss": 0.532,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -2.7306606769561768,
      "rewards/margins": 1.3434104919433594,
      "rewards/rejected": -4.074070930480957,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": 1.1080348491668701,
      "eval_logits/rejected": 3.015399694442749,
      "eval_logps/chosen": -535.3407592773438,
      "eval_logps/rejected": -617.0262451171875,
      "eval_loss": 0.5116756558418274,
      "eval_rewards/accuracies": 0.7198275923728943,
      "eval_rewards/chosen": -2.513347864151001,
      "eval_rewards/margins": 1.200439691543579,
      "eval_rewards/rejected": -3.713787794113159,
      "eval_runtime": 91.1655,
      "eval_samples_per_second": 19.942,
      "eval_steps_per_second": 0.318,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 27.989896858900266,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": 2.1850762367248535,
      "logits/rejected": 3.5708484649658203,
      "logps/chosen": -533.7103271484375,
      "logps/rejected": -638.1661376953125,
      "loss": 0.54,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -2.7955188751220703,
      "rewards/margins": 1.163648247718811,
      "rewards/rejected": -3.959167003631592,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 26.423781855718417,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": 1.8076130151748657,
      "logits/rejected": 4.071971893310547,
      "logps/chosen": -614.3458251953125,
      "logps/rejected": -702.5189819335938,
      "loss": 0.5138,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -3.0845131874084473,
      "rewards/margins": 1.4407538175582886,
      "rewards/rejected": -4.525267601013184,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 24.484829935546955,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": 2.0375962257385254,
      "logits/rejected": 3.5112037658691406,
      "logps/chosen": -607.512939453125,
      "logps/rejected": -720.7943115234375,
      "loss": 0.5187,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -3.2878499031066895,
      "rewards/margins": 1.2010066509246826,
      "rewards/rejected": -4.488856315612793,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 23.41145949202815,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": 1.657248854637146,
      "logits/rejected": 3.65797758102417,
      "logps/chosen": -553.1179809570312,
      "logps/rejected": -692.9093017578125,
      "loss": 0.5258,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -2.909062623977661,
      "rewards/margins": 1.4310705661773682,
      "rewards/rejected": -4.340132713317871,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 28.908000369301963,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": 0.9193560481071472,
      "logits/rejected": 3.4261555671691895,
      "logps/chosen": -548.4246215820312,
      "logps/rejected": -669.2120361328125,
      "loss": 0.5064,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -2.572448968887329,
      "rewards/margins": 1.67291259765625,
      "rewards/rejected": -4.245361804962158,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": 0.7126501202583313,
      "eval_logits/rejected": 2.685429096221924,
      "eval_logps/chosen": -547.6198120117188,
      "eval_logps/rejected": -636.6490478515625,
      "eval_loss": 0.5116574168205261,
      "eval_rewards/accuracies": 0.7241379022598267,
      "eval_rewards/chosen": -2.636138677597046,
      "eval_rewards/margins": 1.2738765478134155,
      "eval_rewards/rejected": -3.9100148677825928,
      "eval_runtime": 90.9536,
      "eval_samples_per_second": 19.988,
      "eval_steps_per_second": 0.319,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 24.395714355317615,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": 0.9300888180732727,
      "logits/rejected": 2.3748581409454346,
      "logps/chosen": -543.5307006835938,
      "logps/rejected": -659.3692016601562,
      "loss": 0.5084,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -2.6378746032714844,
      "rewards/margins": 1.2973625659942627,
      "rewards/rejected": -3.935237169265747,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 22.09369129566989,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": 0.2887948155403137,
      "logits/rejected": 2.4889461994171143,
      "logps/chosen": -591.7805786132812,
      "logps/rejected": -667.9407348632812,
      "loss": 0.5109,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -2.5994677543640137,
      "rewards/margins": 1.3468105792999268,
      "rewards/rejected": -3.9462783336639404,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 26.209345843998328,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": 0.8550162315368652,
      "logits/rejected": 3.1779205799102783,
      "logps/chosen": -580.2351684570312,
      "logps/rejected": -637.8566284179688,
      "loss": 0.5075,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -2.8695898056030273,
      "rewards/margins": 1.257644534111023,
      "rewards/rejected": -4.127234935760498,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 21.12858235158005,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": 0.47364893555641174,
      "logits/rejected": 2.78879451751709,
      "logps/chosen": -587.971435546875,
      "logps/rejected": -724.6598510742188,
      "loss": 0.5014,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": -2.7999932765960693,
      "rewards/margins": 1.9162429571151733,
      "rewards/rejected": -4.716236591339111,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 23.688779288096637,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": 1.2185232639312744,
      "logits/rejected": 2.7246413230895996,
      "logps/chosen": -606.762939453125,
      "logps/rejected": -717.7750244140625,
      "loss": 0.5105,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -3.1490612030029297,
      "rewards/margins": 1.1964399814605713,
      "rewards/rejected": -4.345500946044922,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": 0.863433837890625,
      "eval_logits/rejected": 2.9146454334259033,
      "eval_logps/chosen": -575.282958984375,
      "eval_logps/rejected": -674.3233032226562,
      "eval_loss": 0.5098804235458374,
      "eval_rewards/accuracies": 0.732758641242981,
      "eval_rewards/chosen": -2.9127700328826904,
      "eval_rewards/margins": 1.3739889860153198,
      "eval_rewards/rejected": -4.286758899688721,
      "eval_runtime": 91.1821,
      "eval_samples_per_second": 19.938,
      "eval_steps_per_second": 0.318,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 23.71509307695075,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": 1.313011884689331,
      "logits/rejected": 2.788435459136963,
      "logps/chosen": -609.2062377929688,
      "logps/rejected": -684.1138305664062,
      "loss": 0.5379,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -3.2180449962615967,
      "rewards/margins": 1.0637754201889038,
      "rewards/rejected": -4.281820297241211,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 27.57431143958426,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": 1.1412100791931152,
      "logits/rejected": 3.4668610095977783,
      "logps/chosen": -616.3365478515625,
      "logps/rejected": -685.5581665039062,
      "loss": 0.5129,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -3.115288496017456,
      "rewards/margins": 1.404497504234314,
      "rewards/rejected": -4.5197858810424805,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 19.004038331375536,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": 1.3194568157196045,
      "logits/rejected": 2.8764185905456543,
      "logps/chosen": -561.5572509765625,
      "logps/rejected": -680.3653564453125,
      "loss": 0.5165,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -2.9964187145233154,
      "rewards/margins": 1.2785086631774902,
      "rewards/rejected": -4.274927616119385,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.5604413999330013,
      "train_runtime": 11415.5934,
      "train_samples_per_second": 4.884,
      "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
}