{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 385,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.282051282051282e-07,
      "logits/chosen": -1.7278180122375488,
      "logits/rejected": -1.7377450466156006,
      "logps/chosen": -29.553977966308594,
      "logps/rejected": -42.813133239746094,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.282051282051282e-06,
      "logits/chosen": -1.8666962385177612,
      "logits/rejected": -1.8709977865219116,
      "logps/chosen": -36.98939514160156,
      "logps/rejected": -33.66963195800781,
      "loss": 0.6929,
      "rewards/accuracies": 0.5694444179534912,
      "rewards/chosen": 0.00017197892884723842,
      "rewards/margins": 0.0005675320862792432,
      "rewards/rejected": -0.0003955531574320048,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.564102564102564e-06,
      "logits/chosen": -1.9978935718536377,
      "logits/rejected": -2.000532627105713,
      "logps/chosen": -29.66562843322754,
      "logps/rejected": -29.045883178710938,
      "loss": 0.6934,
      "rewards/accuracies": 0.36250001192092896,
      "rewards/chosen": -0.00023434234026353806,
      "rewards/margins": -0.0004099405778106302,
      "rewards/rejected": 0.0001755982666509226,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": -1.9211324453353882,
      "logits/rejected": -1.9184545278549194,
      "logps/chosen": -31.41294288635254,
      "logps/rejected": -33.23053741455078,
      "loss": 0.6931,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 3.152530553052202e-05,
      "rewards/margins": 0.000152341352077201,
      "rewards/rejected": -0.00012081606837455183,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438434e-06,
      "logits/chosen": -2.017341375350952,
      "logits/rejected": -2.0086092948913574,
      "logps/chosen": -32.60146713256836,
      "logps/rejected": -32.49399185180664,
      "loss": 0.6934,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": -0.0002466029836796224,
      "rewards/margins": -0.0004333632532507181,
      "rewards/rejected": 0.00018676018225960433,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542187e-06,
      "logits/chosen": -1.862633466720581,
      "logits/rejected": -1.8518692255020142,
      "logps/chosen": -33.55931091308594,
      "logps/rejected": -35.44870376586914,
      "loss": 0.6932,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -1.831089502957184e-05,
      "rewards/margins": -5.47249146620743e-05,
      "rewards/rejected": 3.641402145149186e-05,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941119e-06,
      "logits/chosen": -1.9409154653549194,
      "logits/rejected": -1.9428699016571045,
      "logps/chosen": -32.53916549682617,
      "logps/rejected": -33.24130630493164,
      "loss": 0.6925,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.0006101715262047946,
      "rewards/margins": 0.0013094183523207903,
      "rewards/rejected": -0.0006992466514930129,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413549e-06,
      "logits/chosen": -2.070591926574707,
      "logits/rejected": -2.075544834136963,
      "logps/chosen": -34.023067474365234,
      "logps/rejected": -36.647151947021484,
      "loss": 0.6929,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.00034581663203425705,
      "rewards/margins": 0.0004369783273432404,
      "rewards/rejected": -0.0007827949011698365,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-06,
      "logits/chosen": -1.9308092594146729,
      "logits/rejected": -1.933943748474121,
      "logps/chosen": -34.318023681640625,
      "logps/rejected": -34.67802429199219,
      "loss": 0.6922,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.0010978971840813756,
      "rewards/margins": 0.0019540609791874886,
      "rewards/rejected": -0.000856163795106113,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.7367166013034295e-06,
      "logits/chosen": -1.9380912780761719,
      "logits/rejected": -1.9425855875015259,
      "logps/chosen": -32.38385009765625,
      "logps/rejected": -32.35346603393555,
      "loss": 0.6928,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.0008357145707122982,
      "rewards/margins": 0.0007813175907358527,
      "rewards/rejected": 5.439693995867856e-05,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.626245458345211e-06,
      "logits/chosen": -2.035137891769409,
      "logits/rejected": -2.0331528186798096,
      "logps/chosen": -32.112831115722656,
      "logps/rejected": -31.29166030883789,
      "loss": 0.6924,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.0012800416443496943,
      "rewards/margins": 0.0015345367137342691,
      "rewards/rejected": -0.0002544948656577617,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.2300801277160645,
      "eval_logits/rejected": -2.225238084793091,
      "eval_logps/chosen": -34.04683303833008,
      "eval_logps/rejected": -37.53927230834961,
      "eval_loss": 0.6930972337722778,
      "eval_rewards/accuracies": 0.5186877250671387,
      "eval_rewards/chosen": -0.00012280470400583,
      "eval_rewards/margins": 0.00010372586984885857,
      "eval_rewards/rejected": -0.0002265305956825614,
      "eval_runtime": 145.7259,
      "eval_samples_per_second": 2.354,
      "eval_steps_per_second": 0.295,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.498257201263691e-06,
      "logits/chosen": -1.989782691001892,
      "logits/rejected": -1.9873950481414795,
      "logps/chosen": -33.12385559082031,
      "logps/rejected": -34.011810302734375,
      "loss": 0.6926,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.0011996207758784294,
      "rewards/margins": 0.001024017808958888,
      "rewards/rejected": 0.00017560287960805,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777678e-06,
      "logits/chosen": -2.0008151531219482,
      "logits/rejected": -1.992500901222229,
      "logps/chosen": -32.320838928222656,
      "logps/rejected": -32.128170013427734,
      "loss": 0.6927,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.001240686746314168,
      "rewards/margins": 0.0009073130786418915,
      "rewards/rejected": 0.0003333735803607851,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.1940827077152755e-06,
      "logits/chosen": -2.0289230346679688,
      "logits/rejected": -2.020946502685547,
      "logps/chosen": -30.313907623291016,
      "logps/rejected": -32.086116790771484,
      "loss": 0.6922,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.0016180993989109993,
      "rewards/margins": 0.0019491963321343064,
      "rewards/rejected": -0.00033109664218500257,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.0204024186666215e-06,
      "logits/chosen": -1.9592479467391968,
      "logits/rejected": -1.9694607257843018,
      "logps/chosen": -31.223953247070312,
      "logps/rejected": -32.547454833984375,
      "loss": 0.6921,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.0019570994190871716,
      "rewards/margins": 0.002082846825942397,
      "rewards/rejected": -0.00012574761058203876,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.834196265035119e-06,
      "logits/chosen": -1.8708124160766602,
      "logits/rejected": -1.8719879388809204,
      "logps/chosen": -33.877174377441406,
      "logps/rejected": -34.78774642944336,
      "loss": 0.6914,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.003155181184411049,
      "rewards/margins": 0.0034600873477756977,
      "rewards/rejected": -0.00030490627977997065,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800573e-06,
      "logits/chosen": -1.9215673208236694,
      "logits/rejected": -1.9181665182113647,
      "logps/chosen": -36.011531829833984,
      "logps/rejected": -32.685707092285156,
      "loss": 0.6925,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.001601455733180046,
      "rewards/margins": 0.001231002388522029,
      "rewards/rejected": 0.00037045328645035625,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4304331721118078e-06,
      "logits/chosen": -2.021604061126709,
      "logits/rejected": -2.014291524887085,
      "logps/chosen": -33.482086181640625,
      "logps/rejected": -31.404422760009766,
      "loss": 0.6913,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.0031577465124428272,
      "rewards/margins": 0.003683448536321521,
      "rewards/rejected": -0.0005257020820863545,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.2162026428305436e-06,
      "logits/chosen": -2.027444839477539,
      "logits/rejected": -2.032665729522705,
      "logps/chosen": -32.183101654052734,
      "logps/rejected": -32.39936065673828,
      "loss": 0.6918,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.0036168727092444897,
      "rewards/margins": 0.0027590212412178516,
      "rewards/rejected": 0.0008578516426496208,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.996071664294641e-06,
      "logits/chosen": -2.027879238128662,
      "logits/rejected": -2.025132656097412,
      "logps/chosen": -31.258464813232422,
      "logps/rejected": -31.348388671875,
      "loss": 0.6919,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.0023048892617225647,
      "rewards/margins": 0.0026066480204463005,
      "rewards/rejected": -0.000301758642308414,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.7718537898066833e-06,
      "logits/chosen": -1.8983129262924194,
      "logits/rejected": -1.902967095375061,
      "logps/chosen": -31.276391983032227,
      "logps/rejected": -32.81935119628906,
      "loss": 0.6914,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.0032989257015287876,
      "rewards/margins": 0.00358308176510036,
      "rewards/rejected": -0.00028415597626008093,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.2249655723571777,
      "eval_logits/rejected": -2.220139503479004,
      "eval_logps/chosen": -34.04255294799805,
      "eval_logps/rejected": -37.55300521850586,
      "eval_loss": 0.6930080056190491,
      "eval_rewards/accuracies": 0.5245016813278198,
      "eval_rewards/chosen": -8.006239659152925e-05,
      "eval_rewards/margins": 0.00028380370349623263,
      "eval_rewards/rejected": -0.00036386612919159234,
      "eval_runtime": 145.5269,
      "eval_samples_per_second": 2.357,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402006e-06,
      "logits/chosen": -2.010593891143799,
      "logits/rejected": -2.021207332611084,
      "logps/chosen": -31.7437801361084,
      "logps/rejected": -33.93886947631836,
      "loss": 0.6916,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.0022752191871404648,
      "rewards/margins": 0.003036911366507411,
      "rewards/rejected": -0.000761692295782268,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.3185646976551794e-06,
      "logits/chosen": -1.9029136896133423,
      "logits/rejected": -1.9176632165908813,
      "logps/chosen": -29.78145408630371,
      "logps/rejected": -31.63638687133789,
      "loss": 0.6911,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.003300876123830676,
      "rewards/margins": 0.0040829661302268505,
      "rewards/rejected": -0.000782089657150209,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.0932279108998323e-06,
      "logits/chosen": -1.9591538906097412,
      "logits/rejected": -1.9631026983261108,
      "logps/chosen": -33.05189895629883,
      "logps/rejected": -31.594707489013672,
      "loss": 0.6911,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.003655704203993082,
      "rewards/margins": 0.004109731875360012,
      "rewards/rejected": -0.00045402703108265996,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279358e-06,
      "logits/chosen": -1.9572566747665405,
      "logits/rejected": -1.9354870319366455,
      "logps/chosen": -33.83857727050781,
      "logps/rejected": -35.12303924560547,
      "loss": 0.6907,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.0031517534516751766,
      "rewards/margins": 0.004874187987297773,
      "rewards/rejected": -0.0017224351176992059,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.6544367689701824e-06,
      "logits/chosen": -1.998875379562378,
      "logits/rejected": -1.9955555200576782,
      "logps/chosen": -32.72559356689453,
      "logps/rejected": -36.2435417175293,
      "loss": 0.6921,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.0019947518594563007,
      "rewards/margins": 0.0021809376776218414,
      "rewards/rejected": -0.0001861859782366082,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.4445974030621963e-06,
      "logits/chosen": -1.8661177158355713,
      "logits/rejected": -1.8636993169784546,
      "logps/chosen": -33.959014892578125,
      "logps/rejected": -35.526344299316406,
      "loss": 0.6919,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.0022938635665923357,
      "rewards/margins": 0.0025111136492341757,
      "rewards/rejected": -0.0002172500389860943,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.243452991757889e-06,
      "logits/chosen": -1.8511241674423218,
      "logits/rejected": -1.8487510681152344,
      "logps/chosen": -34.16337585449219,
      "logps/rejected": -31.830408096313477,
      "loss": 0.6917,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.002342230873182416,
      "rewards/margins": 0.0029330006800591946,
      "rewards/rejected": -0.0005907699232921004,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603523e-06,
      "logits/chosen": -1.9535648822784424,
      "logits/rejected": -1.943101167678833,
      "logps/chosen": -35.01304244995117,
      "logps/rejected": -31.87521743774414,
      "loss": 0.6913,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.0034359837882220745,
      "rewards/margins": 0.0037782168947160244,
      "rewards/rejected": -0.0003422332229092717,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071455e-07,
      "logits/chosen": -2.048783302307129,
      "logits/rejected": -2.0339112281799316,
      "logps/chosen": -30.716812133789062,
      "logps/rejected": -32.62614059448242,
      "loss": 0.6922,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.0020731096155941486,
      "rewards/margins": 0.001816184027120471,
      "rewards/rejected": 0.00025692558847367764,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-07,
      "logits/chosen": -1.9198474884033203,
      "logits/rejected": -1.917340636253357,
      "logps/chosen": -32.29683303833008,
      "logps/rejected": -30.91409683227539,
      "loss": 0.6895,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.0063027567230165005,
      "rewards/margins": 0.007275627460330725,
      "rewards/rejected": -0.0009728703880682588,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.221111297607422,
      "eval_logits/rejected": -2.2162926197052,
      "eval_logps/chosen": -34.0648307800293,
      "eval_logps/rejected": -37.58684158325195,
      "eval_loss": 0.6929495930671692,
      "eval_rewards/accuracies": 0.5419435501098633,
      "eval_rewards/chosen": -0.0003027978236787021,
      "eval_rewards/margins": 0.0003993964346591383,
      "eval_rewards/rejected": -0.0007021942874416709,
      "eval_runtime": 145.7415,
      "eval_samples_per_second": 2.353,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589035e-07,
      "logits/chosen": -1.905469298362732,
      "logits/rejected": -1.9022337198257446,
      "logps/chosen": -31.291845321655273,
      "logps/rejected": -33.816497802734375,
      "loss": 0.6912,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.003115057712420821,
      "rewards/margins": 0.003858551848679781,
      "rewards/rejected": -0.0007434942526742816,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380913e-07,
      "logits/chosen": -1.9548442363739014,
      "logits/rejected": -1.9426848888397217,
      "logps/chosen": -34.22846603393555,
      "logps/rejected": -33.660888671875,
      "loss": 0.6907,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.0036761020310223103,
      "rewards/margins": 0.004896566737443209,
      "rewards/rejected": -0.0012204641243442893,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-07,
      "logits/chosen": -1.9896023273468018,
      "logits/rejected": -1.9881556034088135,
      "logps/chosen": -33.096466064453125,
      "logps/rejected": -32.550086975097656,
      "loss": 0.6909,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.0038932852912694216,
      "rewards/margins": 0.004598308354616165,
      "rewards/rejected": -0.0007050234125927091,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.0579377374915805e-07,
      "logits/chosen": -2.0761990547180176,
      "logits/rejected": -2.0605435371398926,
      "logps/chosen": -33.7951774597168,
      "logps/rejected": -33.09763717651367,
      "loss": 0.6913,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.0038421000353991985,
      "rewards/margins": 0.0036381154786795378,
      "rewards/rejected": 0.00020398428023327142,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.2518018074041684e-07,
      "logits/chosen": -1.9492791891098022,
      "logits/rejected": -1.9484554529190063,
      "logps/chosen": -32.777854919433594,
      "logps/rejected": -32.48105239868164,
      "loss": 0.6906,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.005042382050305605,
      "rewards/margins": 0.00520141888409853,
      "rewards/rejected": -0.00015903706662356853,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-08,
      "logits/chosen": -1.9036123752593994,
      "logits/rejected": -1.9138826131820679,
      "logps/chosen": -31.845821380615234,
      "logps/rejected": -35.32244873046875,
      "loss": 0.6912,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.003588717896491289,
      "rewards/margins": 0.003925635013729334,
      "rewards/rejected": -0.00033691702992655337,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050325e-08,
      "logits/chosen": -2.043790102005005,
      "logits/rejected": -2.0373566150665283,
      "logps/chosen": -33.307098388671875,
      "logps/rejected": -29.227703094482422,
      "loss": 0.6914,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.003448712872341275,
      "rewards/margins": 0.0034615055192261934,
      "rewards/rejected": -1.2792841516784392e-05,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-09,
      "logits/chosen": -1.9043537378311157,
      "logits/rejected": -1.9065628051757812,
      "logps/chosen": -33.83999252319336,
      "logps/rejected": -30.96780776977539,
      "loss": 0.6907,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.003990581724792719,
      "rewards/margins": 0.004942373372614384,
      "rewards/rejected": -0.0009517916478216648,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.6918259131443965,
      "train_runtime": 3249.6401,
      "train_samples_per_second": 0.947,
      "train_steps_per_second": 0.118
    }
  ],
  "logging_steps": 10,
  "max_steps": 385,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}