{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.28993385883845246,
  "eval_steps": 50,
  "global_step": 400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007248346470961312,
      "grad_norm": 0.07235438376665115,
      "learning_rate": 4.998766400914329e-06,
      "logits/chosen": -1.9785633087158203,
      "logits/rejected": -2.5380234718322754,
      "logps/chosen": -0.28137442469596863,
      "logps/rejected": -0.37793582677841187,
      "loss": 7.3908,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.42206162214279175,
      "rewards/margins": 0.14484205842018127,
      "rewards/rejected": -0.5669037103652954,
      "step": 10
    },
    {
      "epoch": 0.014496692941922623,
      "grad_norm": 0.07455909997224808,
      "learning_rate": 4.9950668210706795e-06,
      "logits/chosen": -2.0576319694519043,
      "logits/rejected": -2.499069929122925,
      "logps/chosen": -0.2769497036933899,
      "logps/rejected": -0.33560773730278015,
      "loss": 7.3753,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.41542449593544006,
      "rewards/margins": 0.08798708021640778,
      "rewards/rejected": -0.503411591053009,
      "step": 20
    },
    {
      "epoch": 0.021745039412883936,
      "grad_norm": 0.09561269730329514,
      "learning_rate": 4.9889049115077e-06,
      "logits/chosen": -2.1018643379211426,
      "logits/rejected": -2.377673625946045,
      "logps/chosen": -0.2673099935054779,
      "logps/rejected": -0.3057115972042084,
      "loss": 7.4516,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.40096497535705566,
      "rewards/margins": 0.05760239437222481,
      "rewards/rejected": -0.45856744050979614,
      "step": 30
    },
    {
      "epoch": 0.028993385883845247,
      "grad_norm": 0.08169445395469666,
      "learning_rate": 4.980286753286196e-06,
      "logits/chosen": -2.1627352237701416,
      "logits/rejected": -2.3873391151428223,
      "logps/chosen": -0.27636194229125977,
      "logps/rejected": -0.3703404664993286,
      "loss": 7.3888,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.4145428538322449,
      "rewards/margins": 0.14096775650978088,
      "rewards/rejected": -0.5555106401443481,
      "step": 40
    },
    {
      "epoch": 0.03624173235480656,
      "grad_norm": 0.08877147734165192,
      "learning_rate": 4.9692208514878445e-06,
      "logits/chosen": -2.1400368213653564,
      "logits/rejected": -2.44627046585083,
      "logps/chosen": -0.25384199619293213,
      "logps/rejected": -0.309225857257843,
      "loss": 7.518,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -0.3807629942893982,
      "rewards/margins": 0.08307582885026932,
      "rewards/rejected": -0.4638388752937317,
      "step": 50
    },
    {
      "epoch": 0.03624173235480656,
      "eval_logits/chosen": -2.1149837970733643,
      "eval_logits/rejected": -2.4759538173675537,
      "eval_logps/chosen": -0.28288090229034424,
      "eval_logps/rejected": -0.3434317111968994,
      "eval_loss": 0.9120966196060181,
      "eval_rewards/accuracies": 0.5089285969734192,
      "eval_rewards/chosen": -0.42432135343551636,
      "eval_rewards/margins": 0.09082622081041336,
      "eval_rewards/rejected": -0.5151475667953491,
      "eval_runtime": 30.2473,
      "eval_samples_per_second": 29.49,
      "eval_steps_per_second": 3.703,
      "step": 50
    },
    {
      "epoch": 0.04349007882576787,
      "grad_norm": 0.11774340271949768,
      "learning_rate": 4.9557181268217225e-06,
      "logits/chosen": -2.0567967891693115,
      "logits/rejected": -2.545668125152588,
      "logps/chosen": -0.2460743933916092,
      "logps/rejected": -0.32390663027763367,
      "loss": 7.3929,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.3691115975379944,
      "rewards/margins": 0.1167483925819397,
      "rewards/rejected": -0.4858599603176117,
      "step": 60
    },
    {
      "epoch": 0.05073842529672919,
      "grad_norm": 0.10012238472700119,
      "learning_rate": 4.939791904846869e-06,
      "logits/chosen": -1.9286606311798096,
      "logits/rejected": -2.4620895385742188,
      "logps/chosen": -0.25992274284362793,
      "logps/rejected": -0.3665553629398346,
      "loss": 7.1863,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.3898841440677643,
      "rewards/margins": 0.15994893014431,
      "rewards/rejected": -0.5498330593109131,
      "step": 70
    },
    {
      "epoch": 0.057986771767690494,
      "grad_norm": 0.07387609034776688,
      "learning_rate": 4.921457902821578e-06,
      "logits/chosen": -2.0778262615203857,
      "logits/rejected": -2.4614813327789307,
      "logps/chosen": -0.24521782994270325,
      "logps/rejected": -0.3450419306755066,
      "loss": 7.3662,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.3678267002105713,
      "rewards/margins": 0.1497361809015274,
      "rewards/rejected": -0.5175628662109375,
      "step": 80
    },
    {
      "epoch": 0.0652351182386518,
      "grad_norm": 0.09718403220176697,
      "learning_rate": 4.900734214192358e-06,
      "logits/chosen": -1.998186707496643,
      "logits/rejected": -2.4382071495056152,
      "logps/chosen": -0.24080593883991241,
      "logps/rejected": -0.3280099332332611,
      "loss": 7.268,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.36120888590812683,
      "rewards/margins": 0.13080602884292603,
      "rewards/rejected": -0.49201488494873047,
      "step": 90
    },
    {
      "epoch": 0.07248346470961312,
      "grad_norm": 0.08059138059616089,
      "learning_rate": 4.8776412907378845e-06,
      "logits/chosen": -1.894728422164917,
      "logits/rejected": -2.475407838821411,
      "logps/chosen": -0.2019994705915451,
      "logps/rejected": -0.29733040928840637,
      "loss": 7.2544,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.30299919843673706,
      "rewards/margins": 0.14299637079238892,
      "rewards/rejected": -0.445995569229126,
      "step": 100
    },
    {
      "epoch": 0.07248346470961312,
      "eval_logits/chosen": -2.1233773231506348,
      "eval_logits/rejected": -2.4810428619384766,
      "eval_logps/chosen": -0.24400465190410614,
      "eval_logps/rejected": -0.3260345160961151,
      "eval_loss": 0.8920583724975586,
      "eval_rewards/accuracies": 0.5625,
      "eval_rewards/chosen": -0.3660070598125458,
      "eval_rewards/margins": 0.12304472178220749,
      "eval_rewards/rejected": -0.4890517592430115,
      "eval_runtime": 30.28,
      "eval_samples_per_second": 29.458,
      "eval_steps_per_second": 3.699,
      "step": 100
    },
    {
      "epoch": 0.07973181118057443,
      "grad_norm": 0.060006480664014816,
      "learning_rate": 4.852201922385564e-06,
      "logits/chosen": -2.1128883361816406,
      "logits/rejected": -2.5161118507385254,
      "logps/chosen": -0.23731942474842072,
      "logps/rejected": -0.31481316685676575,
      "loss": 7.1762,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.3559790849685669,
      "rewards/margins": 0.11624068021774292,
      "rewards/rejected": -0.4722197651863098,
      "step": 110
    },
    {
      "epoch": 0.08698015765153574,
      "grad_norm": 0.0602690726518631,
      "learning_rate": 4.824441214720629e-06,
      "logits/chosen": -2.0212602615356445,
      "logits/rejected": -2.4894890785217285,
      "logps/chosen": -0.18956038355827332,
      "logps/rejected": -0.31935763359069824,
      "loss": 7.1442,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.2843405604362488,
      "rewards/margins": 0.19469590485095978,
      "rewards/rejected": -0.47903648018836975,
      "step": 120
    },
    {
      "epoch": 0.09422850412249706,
      "grad_norm": 0.06425223499536514,
      "learning_rate": 4.794386564209953e-06,
      "logits/chosen": -2.06717848777771,
      "logits/rejected": -2.5086405277252197,
      "logps/chosen": -0.2075866460800171,
      "logps/rejected": -0.3124083876609802,
      "loss": 7.1449,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.31137990951538086,
      "rewards/margins": 0.15723267197608948,
      "rewards/rejected": -0.46861258149147034,
      "step": 130
    },
    {
      "epoch": 0.10147685059345837,
      "grad_norm": 0.08793757855892181,
      "learning_rate": 4.762067631165049e-06,
      "logits/chosen": -2.0242953300476074,
      "logits/rejected": -2.4943737983703613,
      "logps/chosen": -0.19989363849163055,
      "logps/rejected": -0.32057636976242065,
      "loss": 7.122,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.29984045028686523,
      "rewards/margins": 0.18102414906024933,
      "rewards/rejected": -0.48086461424827576,
      "step": 140
    },
    {
      "epoch": 0.10872519706441967,
      "grad_norm": 0.09698841720819473,
      "learning_rate": 4.72751631047092e-06,
      "logits/chosen": -2.098658800125122,
      "logits/rejected": -2.582111120223999,
      "logps/chosen": -0.18897351622581482,
      "logps/rejected": -0.363709419965744,
      "loss": 7.1253,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.28346025943756104,
      "rewards/margins": 0.2621038556098938,
      "rewards/rejected": -0.5455641150474548,
      "step": 150
    },
    {
      "epoch": 0.10872519706441967,
      "eval_logits/chosen": -2.1921768188476562,
      "eval_logits/rejected": -2.556370973587036,
      "eval_logps/chosen": -0.22214026749134064,
      "eval_logps/rejected": -0.3240993022918701,
      "eval_loss": 0.8764163851737976,
      "eval_rewards/accuracies": 0.5803571343421936,
      "eval_rewards/chosen": -0.33321040868759155,
      "eval_rewards/margins": 0.15293852984905243,
      "eval_rewards/rejected": -0.48614898324012756,
      "eval_runtime": 30.273,
      "eval_samples_per_second": 29.465,
      "eval_steps_per_second": 3.7,
      "step": 150
    },
    {
      "epoch": 0.11597354353538099,
      "grad_norm": 0.09119638800621033,
      "learning_rate": 4.690766700109659e-06,
      "logits/chosen": -2.2321066856384277,
      "logits/rejected": -2.5475690364837646,
      "logps/chosen": -0.19602122902870178,
      "logps/rejected": -0.27682799100875854,
      "loss": 7.0307,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.2940318286418915,
      "rewards/margins": 0.12121014297008514,
      "rewards/rejected": -0.4152420163154602,
      "step": 160
    },
    {
      "epoch": 0.1232218900063423,
      "grad_norm": 0.11496366560459137,
      "learning_rate": 4.65185506750986e-06,
      "logits/chosen": -2.1506271362304688,
      "logits/rejected": -2.506520986557007,
      "logps/chosen": -0.18018664419651031,
      "logps/rejected": -0.3084454834461212,
      "loss": 7.1652,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.27028003334999084,
      "rewards/margins": 0.19238826632499695,
      "rewards/rejected": -0.462668240070343,
      "step": 170
    },
    {
      "epoch": 0.1304702364773036,
      "grad_norm": 0.09201692789793015,
      "learning_rate": 4.610819813755038e-06,
      "logits/chosen": -2.285245895385742,
      "logits/rejected": -2.573941707611084,
      "logps/chosen": -0.1844937801361084,
      "logps/rejected": -0.3112415075302124,
      "loss": 7.0022,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.2767406404018402,
      "rewards/margins": 0.1901216208934784,
      "rewards/rejected": -0.4668622612953186,
      "step": 180
    },
    {
      "epoch": 0.13771858294826492,
      "grad_norm": 0.13191230595111847,
      "learning_rate": 4.567701435686405e-06,
      "logits/chosen": -2.265538454055786,
      "logits/rejected": -2.569638252258301,
      "logps/chosen": -0.21122264862060547,
      "logps/rejected": -0.3469196856021881,
      "loss": 7.004,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.3168340027332306,
      "rewards/margins": 0.20354552567005157,
      "rewards/rejected": -0.5203795433044434,
      "step": 190
    },
    {
      "epoch": 0.14496692941922623,
      "grad_norm": 0.1503789722919464,
      "learning_rate": 4.522542485937369e-06,
      "logits/chosen": -2.2021026611328125,
      "logits/rejected": -2.655961513519287,
      "logps/chosen": -0.19664816558361053,
      "logps/rejected": -0.34679359197616577,
      "loss": 7.0539,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.2949722409248352,
      "rewards/margins": 0.22521813213825226,
      "rewards/rejected": -0.5201903581619263,
      "step": 200
    },
    {
      "epoch": 0.14496692941922623,
      "eval_logits/chosen": -2.325611114501953,
      "eval_logits/rejected": -2.715676784515381,
      "eval_logps/chosen": -0.2261900007724762,
      "eval_logps/rejected": -0.3570065200328827,
      "eval_loss": 0.8565592169761658,
      "eval_rewards/accuracies": 0.5714285969734192,
      "eval_rewards/chosen": -0.3392849862575531,
      "eval_rewards/margins": 0.19622473418712616,
      "eval_rewards/rejected": -0.5355097055435181,
      "eval_runtime": 30.2681,
      "eval_samples_per_second": 29.47,
      "eval_steps_per_second": 3.7,
      "step": 200
    },
    {
      "epoch": 0.15221527589018755,
      "grad_norm": 0.12589682638645172,
      "learning_rate": 4.475387530939226e-06,
      "logits/chosen": -2.24491810798645,
      "logits/rejected": -2.739323616027832,
      "logps/chosen": -0.19889305531978607,
      "logps/rejected": -0.32832199335098267,
      "loss": 6.9013,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.2983395457267761,
      "rewards/margins": 0.1941433995962143,
      "rewards/rejected": -0.492482990026474,
      "step": 210
    },
    {
      "epoch": 0.15946362236114886,
      "grad_norm": 0.1417039930820465,
      "learning_rate": 4.426283106939474e-06,
      "logits/chosen": -2.278883695602417,
      "logits/rejected": -2.7108216285705566,
      "logps/chosen": -0.2170490026473999,
      "logps/rejected": -0.3783469796180725,
      "loss": 6.9366,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.32557350397109985,
      "rewards/margins": 0.2419470101594925,
      "rewards/rejected": -0.5675204992294312,
      "step": 220
    },
    {
      "epoch": 0.16671196883211017,
      "grad_norm": 0.1472843438386917,
      "learning_rate": 4.3752776740761495e-06,
      "logits/chosen": -2.16560697555542,
      "logits/rejected": -2.7911431789398193,
      "logps/chosen": -0.20280452072620392,
      "logps/rejected": -0.3874067962169647,
      "loss": 6.8823,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.3042067885398865,
      "rewards/margins": 0.2769034504890442,
      "rewards/rejected": -0.5811101794242859,
      "step": 230
    },
    {
      "epoch": 0.1739603153030715,
      "grad_norm": 0.15212437510490417,
      "learning_rate": 4.322421568553529e-06,
      "logits/chosen": -2.2929282188415527,
      "logits/rejected": -2.8649649620056152,
      "logps/chosen": -0.1841541975736618,
      "logps/rejected": -0.3838128447532654,
      "loss": 6.7139,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.2762312889099121,
      "rewards/margins": 0.29948797821998596,
      "rewards/rejected": -0.5757192373275757,
      "step": 240
    },
    {
      "epoch": 0.1812086617740328,
      "grad_norm": 0.22261729836463928,
      "learning_rate": 4.267766952966369e-06,
      "logits/chosen": -2.436549663543701,
      "logits/rejected": -2.827937126159668,
      "logps/chosen": -0.2322833091020584,
      "logps/rejected": -0.39587104320526123,
      "loss": 6.7737,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.3484249711036682,
      "rewards/margins": 0.24538159370422363,
      "rewards/rejected": -0.5938066244125366,
      "step": 250
    },
    {
      "epoch": 0.1812086617740328,
      "eval_logits/chosen": -2.4983742237091064,
      "eval_logits/rejected": -2.8988120555877686,
      "eval_logps/chosen": -0.269414484500885,
      "eval_logps/rejected": -0.45964303612709045,
      "eval_loss": 0.816527247428894,
      "eval_rewards/accuracies": 0.5714285969734192,
      "eval_rewards/chosen": -0.4041217863559723,
      "eval_rewards/margins": 0.2853427231311798,
      "eval_rewards/rejected": -0.6894644498825073,
      "eval_runtime": 30.2705,
      "eval_samples_per_second": 29.468,
      "eval_steps_per_second": 3.7,
      "step": 250
    },
    {
      "epoch": 0.18845700824499412,
      "grad_norm": 0.25962531566619873,
      "learning_rate": 4.211367764821722e-06,
      "logits/chosen": -2.3911795616149902,
      "logits/rejected": -2.9030754566192627,
      "logps/chosen": -0.23631009459495544,
      "logps/rejected": -0.44218096137046814,
      "loss": 6.6574,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.35446515679359436,
      "rewards/margins": 0.3088063597679138,
      "rewards/rejected": -0.6632715463638306,
      "step": 260
    },
    {
      "epoch": 0.19570535471595543,
      "grad_norm": 0.29614314436912537,
      "learning_rate": 4.15327966330913e-06,
      "logits/chosen": -2.39563250541687,
      "logits/rejected": -2.8926572799682617,
      "logps/chosen": -0.2558010220527649,
      "logps/rejected": -0.5495272278785706,
      "loss": 6.6017,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.38370150327682495,
      "rewards/margins": 0.4405893385410309,
      "rewards/rejected": -0.8242908716201782,
      "step": 270
    },
    {
      "epoch": 0.20295370118691675,
      "grad_norm": 0.2586583197116852,
      "learning_rate": 4.093559974371725e-06,
      "logits/chosen": -2.5368762016296387,
      "logits/rejected": -2.8856003284454346,
      "logps/chosen": -0.2823846638202667,
      "logps/rejected": -0.4850040376186371,
      "loss": 6.3849,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.4235769808292389,
      "rewards/margins": 0.30392909049987793,
      "rewards/rejected": -0.7275060415267944,
      "step": 280
    },
    {
      "epoch": 0.21020204765787803,
      "grad_norm": 0.3820374608039856,
      "learning_rate": 4.032267634132442e-06,
      "logits/chosen": -2.404680013656616,
      "logits/rejected": -2.8940536975860596,
      "logps/chosen": -0.3108128011226654,
      "logps/rejected": -0.7274529337882996,
      "loss": 6.0103,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.46621912717819214,
      "rewards/margins": 0.6249603033065796,
      "rewards/rejected": -1.0911794900894165,
      "step": 290
    },
    {
      "epoch": 0.21745039412883935,
      "grad_norm": 0.572814404964447,
      "learning_rate": 3.969463130731183e-06,
      "logits/chosen": -2.3861801624298096,
      "logits/rejected": -2.8026037216186523,
      "logps/chosen": -0.40264564752578735,
      "logps/rejected": -0.8128012418746948,
      "loss": 5.9678,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.6039685010910034,
      "rewards/margins": 0.6152334809303284,
      "rewards/rejected": -1.219201922416687,
      "step": 300
    },
    {
      "epoch": 0.21745039412883935,
      "eval_logits/chosen": -2.519517421722412,
      "eval_logits/rejected": -2.8339390754699707,
      "eval_logps/chosen": -0.4459304213523865,
      "eval_logps/rejected": -0.8040717840194702,
      "eval_loss": 0.7266466617584229,
      "eval_rewards/accuracies": 0.5625,
      "eval_rewards/chosen": -0.6688956618309021,
      "eval_rewards/margins": 0.5372119545936584,
      "eval_rewards/rejected": -1.206107497215271,
      "eval_runtime": 30.2692,
      "eval_samples_per_second": 29.469,
      "eval_steps_per_second": 3.7,
      "step": 300
    },
    {
      "epoch": 0.22469874059980066,
      "grad_norm": 0.3507815897464752,
      "learning_rate": 3.905208444630326e-06,
      "logits/chosen": -2.4855399131774902,
      "logits/rejected": -2.8341915607452393,
      "logps/chosen": -0.36750203371047974,
      "logps/rejected": -0.8827482461929321,
      "loss": 5.5511,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.5512530207633972,
      "rewards/margins": 0.7728692889213562,
      "rewards/rejected": -1.324122428894043,
      "step": 310
    },
    {
      "epoch": 0.23194708707076198,
      "grad_norm": 0.34762975573539734,
      "learning_rate": 3.839566987447492e-06,
      "logits/chosen": -2.312408924102783,
      "logits/rejected": -2.7844693660736084,
      "logps/chosen": -0.465009868144989,
      "logps/rejected": -1.074065923690796,
      "loss": 5.5495,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.6975148320198059,
      "rewards/margins": 0.9135842323303223,
      "rewards/rejected": -1.6110990047454834,
      "step": 320
    },
    {
      "epoch": 0.2391954335417233,
      "grad_norm": 0.5242094397544861,
      "learning_rate": 3.772603539375929e-06,
      "logits/chosen": -2.4517929553985596,
      "logits/rejected": -2.7228286266326904,
      "logps/chosen": -0.5588141679763794,
      "logps/rejected": -1.1725430488586426,
      "loss": 5.526,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.8382211923599243,
      "rewards/margins": 0.9205933809280396,
      "rewards/rejected": -1.7588145732879639,
      "step": 330
    },
    {
      "epoch": 0.2464437800126846,
      "grad_norm": 0.7856224179267883,
      "learning_rate": 3.7043841852542884e-06,
      "logits/chosen": -2.5230085849761963,
      "logits/rejected": -2.8241894245147705,
      "logps/chosen": -0.527363657951355,
      "logps/rejected": -1.1461597681045532,
      "loss": 5.0865,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.7910455465316772,
      "rewards/margins": 0.9281940460205078,
      "rewards/rejected": -1.7192394733428955,
      "step": 340
    },
    {
      "epoch": 0.2536921264836459,
      "grad_norm": 0.5080834031105042,
      "learning_rate": 3.634976249348867e-06,
      "logits/chosen": -2.414911985397339,
      "logits/rejected": -2.8537158966064453,
      "logps/chosen": -0.5747151374816895,
      "logps/rejected": -1.4558517932891846,
      "loss": 5.0036,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.862072765827179,
      "rewards/margins": 1.3217047452926636,
      "rewards/rejected": -2.1837775707244873,
      "step": 350
    },
    {
      "epoch": 0.2536921264836459,
      "eval_logits/chosen": -2.647620677947998,
      "eval_logits/rejected": -2.910278558731079,
      "eval_logps/chosen": -0.6226872205734253,
      "eval_logps/rejected": -1.3186976909637451,
      "eval_loss": 0.6240565776824951,
      "eval_rewards/accuracies": 0.5982142686843872,
      "eval_rewards/chosen": -0.9340308308601379,
      "eval_rewards/margins": 1.0440157651901245,
      "eval_rewards/rejected": -1.9780464172363281,
      "eval_runtime": 30.2634,
      "eval_samples_per_second": 29.475,
      "eval_steps_per_second": 3.701,
      "step": 350
    },
    {
      "epoch": 0.2609404729546072,
      "grad_norm": 0.5225710868835449,
      "learning_rate": 3.564448228912682e-06,
      "logits/chosen": -2.5971312522888184,
      "logits/rejected": -2.8506340980529785,
      "logps/chosen": -0.6313252449035645,
      "logps/rejected": -1.4177578687667847,
      "loss": 5.2602,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.9469879269599915,
      "rewards/margins": 1.179648995399475,
      "rewards/rejected": -2.1266369819641113,
      "step": 360
    },
    {
      "epoch": 0.26818881942556855,
      "grad_norm": 0.44064539670944214,
      "learning_rate": 3.4928697265869516e-06,
      "logits/chosen": -2.645890712738037,
      "logits/rejected": -2.8477931022644043,
      "logps/chosen": -0.6448010206222534,
      "logps/rejected": -1.6615798473358154,
      "loss": 4.9912,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.9672015905380249,
      "rewards/margins": 1.5251682996749878,
      "rewards/rejected": -2.492370128631592,
      "step": 370
    },
    {
      "epoch": 0.27543716589652983,
      "grad_norm": 0.6242617964744568,
      "learning_rate": 3.4203113817116955e-06,
      "logits/chosen": -2.685136318206787,
      "logits/rejected": -2.8105804920196533,
      "logps/chosen": -0.6792098879814148,
      "logps/rejected": -1.753614068031311,
      "loss": 4.7972,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.0188149213790894,
      "rewards/margins": 1.611606240272522,
      "rewards/rejected": -2.6304211616516113,
      "step": 380
    },
    {
      "epoch": 0.2826855123674912,
      "grad_norm": 0.5383561253547668,
      "learning_rate": 3.346844800613229e-06,
      "logits/chosen": -2.6124706268310547,
      "logits/rejected": -2.8373632431030273,
      "logps/chosen": -0.6230086088180542,
      "logps/rejected": -2.398132801055908,
      "loss": 4.7275,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.9345127940177917,
      "rewards/margins": 2.662686824798584,
      "rewards/rejected": -3.5971992015838623,
      "step": 390
    },
    {
      "epoch": 0.28993385883845246,
      "grad_norm": 0.6497470736503601,
      "learning_rate": 3.272542485937369e-06,
      "logits/chosen": -2.5315146446228027,
      "logits/rejected": -2.892939805984497,
      "logps/chosen": -0.7851268649101257,
      "logps/rejected": -2.1644442081451416,
      "loss": 4.5874,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -1.1776902675628662,
      "rewards/margins": 2.0689759254455566,
      "rewards/rejected": -3.246666431427002,
      "step": 400
    },
    {
      "epoch": 0.28993385883845246,
      "eval_logits/chosen": -2.704826831817627,
      "eval_logits/rejected": -2.969177007675171,
      "eval_logps/chosen": -0.7823955416679382,
      "eval_logps/rejected": -2.0450375080108643,
      "eval_loss": 0.5390450358390808,
      "eval_rewards/accuracies": 0.5982142686843872,
      "eval_rewards/chosen": -1.1735934019088745,
      "eval_rewards/margins": 1.8939628601074219,
      "eval_rewards/rejected": -3.067556142807007,
      "eval_runtime": 30.267,
      "eval_samples_per_second": 29.471,
      "eval_steps_per_second": 3.7,
      "step": 400
    }
  ],
  "logging_steps": 10,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.4824563793807278e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}