{
  "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": 5.786441925419282,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.619342088699341,
      "logits/rejected": -2.5526421070098877,
      "logps/chosen": -265.45428466796875,
      "logps/rejected": -236.1463165283203,
      "loss": 0.6931,
      "rewards/accuracies": 0.41874998807907104,
      "rewards/chosen": -8.460578828817233e-05,
      "rewards/margins": 7.87067401688546e-05,
      "rewards/rejected": -0.000163312564836815,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 5.491351697550114,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.6576342582702637,
      "logits/rejected": -2.5759005546569824,
      "logps/chosen": -298.7987060546875,
      "logps/rejected": -274.3077392578125,
      "loss": 0.6923,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.00013454208965413272,
      "rewards/margins": 0.0018433767836540937,
      "rewards/rejected": -0.0019779191352427006,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 5.277174601748908,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6759681701660156,
      "logits/rejected": -2.602142333984375,
      "logps/chosen": -290.32366943359375,
      "logps/rejected": -234.38919067382812,
      "loss": 0.6891,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.010090610012412071,
      "rewards/margins": 0.014163595624268055,
      "rewards/rejected": -0.004072986543178558,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 5.753580388451781,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.660831928253174,
      "logits/rejected": -2.61124324798584,
      "logps/chosen": -280.9427490234375,
      "logps/rejected": -267.7388000488281,
      "loss": 0.68,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.042264439165592194,
      "rewards/margins": 0.04172234237194061,
      "rewards/rejected": 0.0005421031382866204,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 6.220985202592042,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.61970591545105,
      "logits/rejected": -2.6113502979278564,
      "logps/chosen": -294.2338562011719,
      "logps/rejected": -305.00823974609375,
      "loss": 0.6653,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.020173203200101852,
      "rewards/margins": 0.07841379195451736,
      "rewards/rejected": -0.058240581303834915,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.5731215476989746,
      "eval_logits/rejected": -2.4923324584960938,
      "eval_logps/chosen": -285.4681701660156,
      "eval_logps/rejected": -260.73919677734375,
      "eval_loss": 0.6512896418571472,
      "eval_rewards/accuracies": 0.693965494632721,
      "eval_rewards/chosen": -0.003780907019972801,
      "eval_rewards/margins": 0.13838669657707214,
      "eval_rewards/rejected": -0.1421675980091095,
      "eval_runtime": 91.4355,
      "eval_samples_per_second": 19.883,
      "eval_steps_per_second": 0.317,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 8.996643288143726,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.5592479705810547,
      "logits/rejected": -2.5012314319610596,
      "logps/chosen": -296.82501220703125,
      "logps/rejected": -279.3923645019531,
      "loss": 0.6437,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.07340322434902191,
      "rewards/margins": 0.14495481550693512,
      "rewards/rejected": -0.21835803985595703,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 9.499312526523301,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.4377551078796387,
      "logits/rejected": -2.392364740371704,
      "logps/chosen": -343.14935302734375,
      "logps/rejected": -305.904541015625,
      "loss": 0.6225,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.15578363835811615,
      "rewards/margins": 0.2891853451728821,
      "rewards/rejected": -0.4449689984321594,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 13.383949209241152,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.324429512023926,
      "logits/rejected": -2.2132680416107178,
      "logps/chosen": -312.23431396484375,
      "logps/rejected": -314.443115234375,
      "loss": 0.6094,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.2638443112373352,
      "rewards/margins": 0.35402628779411316,
      "rewards/rejected": -0.617870569229126,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 13.193135435969568,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -1.3427231311798096,
      "logits/rejected": -1.2150843143463135,
      "logps/chosen": -337.8877258300781,
      "logps/rejected": -351.61114501953125,
      "loss": 0.588,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.5029428601264954,
      "rewards/margins": 0.5451563596725464,
      "rewards/rejected": -1.048099160194397,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 18.776758250311076,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -0.1327299326658249,
      "logits/rejected": 0.009161601774394512,
      "logps/chosen": -333.9382019042969,
      "logps/rejected": -386.19952392578125,
      "loss": 0.5699,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.6151348352432251,
      "rewards/margins": 0.5275757312774658,
      "rewards/rejected": -1.142710566520691,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": 0.3393348455429077,
      "eval_logits/rejected": 0.8771037459373474,
      "eval_logps/chosen": -373.8545227050781,
      "eval_logps/rejected": -407.7341613769531,
      "eval_loss": 0.5577893257141113,
      "eval_rewards/accuracies": 0.6853448152542114,
      "eval_rewards/chosen": -0.8876444697380066,
      "eval_rewards/margins": 0.7244728207588196,
      "eval_rewards/rejected": -1.6121174097061157,
      "eval_runtime": 92.82,
      "eval_samples_per_second": 19.586,
      "eval_steps_per_second": 0.312,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 16.828561137860333,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": 0.10855080932378769,
      "logits/rejected": 0.6684142351150513,
      "logps/chosen": -363.28369140625,
      "logps/rejected": -392.91302490234375,
      "loss": 0.5591,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.9809554815292358,
      "rewards/margins": 0.6108429431915283,
      "rewards/rejected": -1.5917984247207642,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 19.631970765133257,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": 0.46126747131347656,
      "logits/rejected": 0.8139599561691284,
      "logps/chosen": -382.4955139160156,
      "logps/rejected": -427.80206298828125,
      "loss": 0.5619,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -1.0667617321014404,
      "rewards/margins": 0.6966198682785034,
      "rewards/rejected": -1.7633816003799438,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 23.999804178074818,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": 1.1685855388641357,
      "logits/rejected": 1.665592908859253,
      "logps/chosen": -352.095458984375,
      "logps/rejected": -419.9580078125,
      "loss": 0.5499,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.9591536521911621,
      "rewards/margins": 0.8143006563186646,
      "rewards/rejected": -1.7734540700912476,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 18.64421738880492,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": 1.145119547843933,
      "logits/rejected": 1.878694772720337,
      "logps/chosen": -350.3140563964844,
      "logps/rejected": -365.4109802246094,
      "loss": 0.5568,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.7412891387939453,
      "rewards/margins": 0.6698554754257202,
      "rewards/rejected": -1.411144733428955,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 20.200927048362935,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": 1.7342513799667358,
      "logits/rejected": 2.4978702068328857,
      "logps/chosen": -387.82183837890625,
      "logps/rejected": -444.968505859375,
      "loss": 0.5416,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -1.2127825021743774,
      "rewards/margins": 0.9050270915031433,
      "rewards/rejected": -2.117809772491455,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": 1.2442059516906738,
      "eval_logits/rejected": 2.1977624893188477,
      "eval_logps/chosen": -399.5646667480469,
      "eval_logps/rejected": -458.08807373046875,
      "eval_loss": 0.5319975018501282,
      "eval_rewards/accuracies": 0.7025862336158752,
      "eval_rewards/chosen": -1.1447453498840332,
      "eval_rewards/margins": 0.9709104895591736,
      "eval_rewards/rejected": -2.1156558990478516,
      "eval_runtime": 91.9947,
      "eval_samples_per_second": 19.762,
      "eval_steps_per_second": 0.315,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 21.063284567870923,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": 1.3700188398361206,
      "logits/rejected": 1.9712657928466797,
      "logps/chosen": -422.37750244140625,
      "logps/rejected": -497.3843688964844,
      "loss": 0.5292,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -1.2582637071609497,
      "rewards/margins": 0.8558940887451172,
      "rewards/rejected": -2.1141579151153564,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 20.8460851273499,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": 1.7427918910980225,
      "logits/rejected": 2.409679412841797,
      "logps/chosen": -401.26861572265625,
      "logps/rejected": -495.7734375,
      "loss": 0.5061,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.2966772317886353,
      "rewards/margins": 0.9603677988052368,
      "rewards/rejected": -2.257045269012451,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 17.79571920632354,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": 1.8351173400878906,
      "logits/rejected": 2.8393807411193848,
      "logps/chosen": -433.7701110839844,
      "logps/rejected": -456.1665954589844,
      "loss": 0.5391,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.3041222095489502,
      "rewards/margins": 0.8725945353507996,
      "rewards/rejected": -2.1767165660858154,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 20.925215758529603,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": 2.0596728324890137,
      "logits/rejected": 2.587247371673584,
      "logps/chosen": -375.49212646484375,
      "logps/rejected": -489.371826171875,
      "loss": 0.517,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -1.1737760305404663,
      "rewards/margins": 1.0764976739883423,
      "rewards/rejected": -2.2502734661102295,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 21.33737350989529,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": 2.203148365020752,
      "logits/rejected": 2.9923298358917236,
      "logps/chosen": -424.408203125,
      "logps/rejected": -482.44879150390625,
      "loss": 0.5318,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.4612659215927124,
      "rewards/margins": 0.9246621131896973,
      "rewards/rejected": -2.38592791557312,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": 1.3448154926300049,
      "eval_logits/rejected": 2.4483416080474854,
      "eval_logps/chosen": -404.1459655761719,
      "eval_logps/rejected": -469.58026123046875,
      "eval_loss": 0.5122287273406982,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -1.1905584335327148,
      "eval_rewards/margins": 1.0400197505950928,
      "eval_rewards/rejected": -2.2305781841278076,
      "eval_runtime": 92.2657,
      "eval_samples_per_second": 19.704,
      "eval_steps_per_second": 0.314,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 22.48419176653969,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": 1.9083038568496704,
      "logits/rejected": 2.511636734008789,
      "logps/chosen": -405.66839599609375,
      "logps/rejected": -489.11663818359375,
      "loss": 0.5124,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.3236396312713623,
      "rewards/margins": 1.0241048336029053,
      "rewards/rejected": -2.3477444648742676,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 24.286637463373633,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": 1.8827180862426758,
      "logits/rejected": 2.9742465019226074,
      "logps/chosen": -462.0406799316406,
      "logps/rejected": -545.8687133789062,
      "loss": 0.503,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -1.3936035633087158,
      "rewards/margins": 1.215381383895874,
      "rewards/rejected": -2.6089847087860107,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 19.94988074398488,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": 1.1973925828933716,
      "logits/rejected": 1.7874739170074463,
      "logps/chosen": -377.791748046875,
      "logps/rejected": -472.1722717285156,
      "loss": 0.5198,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.0984822511672974,
      "rewards/margins": 0.9670408368110657,
      "rewards/rejected": -2.0655229091644287,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 24.247680553138093,
      "learning_rate": 2.5e-07,
      "logits/chosen": 1.5466783046722412,
      "logits/rejected": 2.6804802417755127,
      "logps/chosen": -449.4345703125,
      "logps/rejected": -522.0684814453125,
      "loss": 0.5279,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -1.6190277338027954,
      "rewards/margins": 0.9881566762924194,
      "rewards/rejected": -2.6071841716766357,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 18.449169539289652,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": 1.7379165887832642,
      "logits/rejected": 2.9272639751434326,
      "logps/chosen": -468.9725646972656,
      "logps/rejected": -532.5216674804688,
      "loss": 0.5178,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.6517877578735352,
      "rewards/margins": 1.06899094581604,
      "rewards/rejected": -2.720778226852417,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": 1.3214832544326782,
      "eval_logits/rejected": 2.5336461067199707,
      "eval_logps/chosen": -429.1148681640625,
      "eval_logps/rejected": -502.6709289550781,
      "eval_loss": 0.5028622150421143,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -1.440247893333435,
      "eval_rewards/margins": 1.1212375164031982,
      "eval_rewards/rejected": -2.561485528945923,
      "eval_runtime": 92.4677,
      "eval_samples_per_second": 19.661,
      "eval_steps_per_second": 0.314,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 18.218334579311854,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": 1.2481517791748047,
      "logits/rejected": 2.1310501098632812,
      "logps/chosen": -418.47222900390625,
      "logps/rejected": -541.3790283203125,
      "loss": 0.5073,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -1.4775768518447876,
      "rewards/margins": 1.1922760009765625,
      "rewards/rejected": -2.6698527336120605,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 18.269676157352414,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": 1.8889141082763672,
      "logits/rejected": 2.4701380729675293,
      "logps/chosen": -426.7981872558594,
      "logps/rejected": -536.7744750976562,
      "loss": 0.504,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.607358694076538,
      "rewards/margins": 1.241914987564087,
      "rewards/rejected": -2.849273920059204,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 21.937738574869034,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": 2.0935568809509277,
      "logits/rejected": 3.0940475463867188,
      "logps/chosen": -444.2198181152344,
      "logps/rejected": -559.611328125,
      "loss": 0.5276,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.7122100591659546,
      "rewards/margins": 1.2459454536437988,
      "rewards/rejected": -2.958155393600464,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 23.622077882281282,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": 2.260361433029175,
      "logits/rejected": 3.3843066692352295,
      "logps/chosen": -451.55938720703125,
      "logps/rejected": -546.6029052734375,
      "loss": 0.5071,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -1.6382853984832764,
      "rewards/margins": 1.288576364517212,
      "rewards/rejected": -2.9268617630004883,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 20.079829542188833,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": 1.4699538946151733,
      "logits/rejected": 2.1643927097320557,
      "logps/chosen": -453.72039794921875,
      "logps/rejected": -500.624267578125,
      "loss": 0.519,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -1.4777113199234009,
      "rewards/margins": 0.939932644367218,
      "rewards/rejected": -2.4176440238952637,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": 1.3894954919815063,
      "eval_logits/rejected": 2.688563346862793,
      "eval_logps/chosen": -433.8905944824219,
      "eval_logps/rejected": -514.7540283203125,
      "eval_loss": 0.4985021948814392,
      "eval_rewards/accuracies": 0.7370689511299133,
      "eval_rewards/chosen": -1.488004446029663,
      "eval_rewards/margins": 1.1943109035491943,
      "eval_rewards/rejected": -2.6823153495788574,
      "eval_runtime": 91.7036,
      "eval_samples_per_second": 19.825,
      "eval_steps_per_second": 0.316,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 20.41756706068905,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": 1.4253332614898682,
      "logits/rejected": 2.419015407562256,
      "logps/chosen": -438.324462890625,
      "logps/rejected": -531.0131225585938,
      "loss": 0.5304,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.603592872619629,
      "rewards/margins": 1.0420410633087158,
      "rewards/rejected": -2.6456339359283447,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 23.710557812034125,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": 1.8713343143463135,
      "logits/rejected": 3.118149757385254,
      "logps/chosen": -430.84912109375,
      "logps/rejected": -534.192626953125,
      "loss": 0.5078,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -1.5661717653274536,
      "rewards/margins": 1.4308522939682007,
      "rewards/rejected": -2.9970240592956543,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 22.190813187925453,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": 2.05544376373291,
      "logits/rejected": 2.8256397247314453,
      "logps/chosen": -446.35931396484375,
      "logps/rejected": -508.9071350097656,
      "loss": 0.5035,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -1.7183713912963867,
      "rewards/margins": 0.8996985554695129,
      "rewards/rejected": -2.618070125579834,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 26.678335434261424,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": 1.8211781978607178,
      "logits/rejected": 2.5711045265197754,
      "logps/chosen": -450.910400390625,
      "logps/rejected": -517.4847412109375,
      "loss": 0.501,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.6291780471801758,
      "rewards/margins": 0.9458298683166504,
      "rewards/rejected": -2.575007677078247,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 22.960680980618804,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": 1.8633334636688232,
      "logits/rejected": 3.0288431644439697,
      "logps/chosen": -437.11260986328125,
      "logps/rejected": -511.2594299316406,
      "loss": 0.5137,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.6262149810791016,
      "rewards/margins": 1.1097644567489624,
      "rewards/rejected": -2.7359795570373535,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": 1.8716328144073486,
      "eval_logits/rejected": 3.1580371856689453,
      "eval_logps/chosen": -446.3657531738281,
      "eval_logps/rejected": -532.5296020507812,
      "eval_loss": 0.49309831857681274,
      "eval_rewards/accuracies": 0.732758641242981,
      "eval_rewards/chosen": -1.612756371498108,
      "eval_rewards/margins": 1.247314453125,
      "eval_rewards/rejected": -2.8600711822509766,
      "eval_runtime": 91.4764,
      "eval_samples_per_second": 19.874,
      "eval_steps_per_second": 0.317,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 25.170197026630024,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": 1.9063358306884766,
      "logits/rejected": 2.6921870708465576,
      "logps/chosen": -398.64984130859375,
      "logps/rejected": -524.7725830078125,
      "loss": 0.4898,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.4632974863052368,
      "rewards/margins": 1.1990567445755005,
      "rewards/rejected": -2.662353992462158,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 23.25403059611051,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": 1.4586713314056396,
      "logits/rejected": 2.6852428913116455,
      "logps/chosen": -454.4266052246094,
      "logps/rejected": -542.4071044921875,
      "loss": 0.4926,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.5484213829040527,
      "rewards/margins": 1.2499048709869385,
      "rewards/rejected": -2.798326015472412,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 28.54168436834153,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": 1.7404067516326904,
      "logits/rejected": 3.0292108058929443,
      "logps/chosen": -451.39849853515625,
      "logps/rejected": -544.4757080078125,
      "loss": 0.5001,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.5502045154571533,
      "rewards/margins": 1.3179900646209717,
      "rewards/rejected": -2.868194818496704,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 25.43427497983774,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": 2.222611904144287,
      "logits/rejected": 3.231661558151245,
      "logps/chosen": -413.8394470214844,
      "logps/rejected": -501.4020080566406,
      "loss": 0.515,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.6188195943832397,
      "rewards/margins": 1.1021382808685303,
      "rewards/rejected": -2.7209575176239014,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 24.959131634596684,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": 2.535111427307129,
      "logits/rejected": 3.2772762775421143,
      "logps/chosen": -382.0926208496094,
      "logps/rejected": -499.1607971191406,
      "loss": 0.5033,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.5753300189971924,
      "rewards/margins": 1.1241223812103271,
      "rewards/rejected": -2.6994521617889404,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": 2.0773062705993652,
      "eval_logits/rejected": 3.358436346054077,
      "eval_logps/chosen": -449.9159851074219,
      "eval_logps/rejected": -541.3424072265625,
      "eval_loss": 0.49311476945877075,
      "eval_rewards/accuracies": 0.7413793206214905,
      "eval_rewards/chosen": -1.6482590436935425,
      "eval_rewards/margins": 1.299940824508667,
      "eval_rewards/rejected": -2.948199987411499,
      "eval_runtime": 92.5987,
      "eval_samples_per_second": 19.633,
      "eval_steps_per_second": 0.313,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 28.32448783876109,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": 2.5014076232910156,
      "logits/rejected": 3.729045867919922,
      "logps/chosen": -446.04974365234375,
      "logps/rejected": -546.2440185546875,
      "loss": 0.4985,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -1.7509788274765015,
      "rewards/margins": 1.3309944868087769,
      "rewards/rejected": -3.0819735527038574,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 31.377213722949094,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": 2.54590106010437,
      "logits/rejected": 3.5291149616241455,
      "logps/chosen": -431.8235778808594,
      "logps/rejected": -553.3231201171875,
      "loss": 0.4966,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -1.723846197128296,
      "rewards/margins": 1.350610613822937,
      "rewards/rejected": -3.0744569301605225,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 25.27983521226701,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": 2.104177713394165,
      "logits/rejected": 3.586791515350342,
      "logps/chosen": -467.50860595703125,
      "logps/rejected": -525.298828125,
      "loss": 0.4944,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.64333176612854,
      "rewards/margins": 1.1692912578582764,
      "rewards/rejected": -2.8126227855682373,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.5464360101507344,
      "train_runtime": 11333.4438,
      "train_samples_per_second": 4.92,
      "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
}