{
  "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": 8.114700597292016,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.630662441253662,
      "logits/rejected": -2.588312864303589,
      "logps/chosen": -251.37826538085938,
      "logps/rejected": -245.56118774414062,
      "loss": 0.6932,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": 0.0003000783617608249,
      "rewards/margins": 3.4782169677782804e-05,
      "rewards/rejected": 0.00026529619935899973,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 9.2292553026982,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.6437935829162598,
      "logits/rejected": -2.5945396423339844,
      "logps/chosen": -305.3244323730469,
      "logps/rejected": -288.41082763671875,
      "loss": 0.692,
      "rewards/accuracies": 0.5562499761581421,
      "rewards/chosen": 0.0019361503655090928,
      "rewards/margins": 0.0025582597590982914,
      "rewards/rejected": -0.0006221095682121813,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 8.362484362577634,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6338467597961426,
      "logits/rejected": -2.5890305042266846,
      "logps/chosen": -290.099365234375,
      "logps/rejected": -311.11444091796875,
      "loss": 0.6875,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.012083572335541248,
      "rewards/margins": 0.01316638570278883,
      "rewards/rejected": -0.0010828140657395124,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 8.520101675940221,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.6183691024780273,
      "logits/rejected": -2.586637258529663,
      "logps/chosen": -266.4432067871094,
      "logps/rejected": -251.4285125732422,
      "loss": 0.6773,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.022834882140159607,
      "rewards/margins": 0.023432452231645584,
      "rewards/rejected": -0.0005975713720545173,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 9.078671253466315,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.6266233921051025,
      "logits/rejected": -2.5520858764648438,
      "logps/chosen": -304.1348571777344,
      "logps/rejected": -262.32061767578125,
      "loss": 0.6617,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": 0.002302716951817274,
      "rewards/margins": 0.0754440575838089,
      "rewards/rejected": -0.07314133644104004,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.561208486557007,
      "eval_logits/rejected": -2.4839041233062744,
      "eval_logps/chosen": -288.2777099609375,
      "eval_logps/rejected": -261.73126220703125,
      "eval_loss": 0.6429266333580017,
      "eval_rewards/accuracies": 0.6982758641242981,
      "eval_rewards/chosen": -0.031876176595687866,
      "eval_rewards/margins": 0.12021197378635406,
      "eval_rewards/rejected": -0.15208815038204193,
      "eval_runtime": 95.5992,
      "eval_samples_per_second": 19.017,
      "eval_steps_per_second": 0.303,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 11.596073380531864,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.547043561935425,
      "logits/rejected": -2.4718384742736816,
      "logps/chosen": -296.81402587890625,
      "logps/rejected": -309.17828369140625,
      "loss": 0.6408,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.10427121073007584,
      "rewards/margins": 0.10880544036626816,
      "rewards/rejected": -0.2130766659975052,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 14.345685973653804,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.57369065284729,
      "logits/rejected": -2.5374178886413574,
      "logps/chosen": -286.39215087890625,
      "logps/rejected": -286.45330810546875,
      "loss": 0.6181,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.11706429719924927,
      "rewards/margins": 0.20492573082447052,
      "rewards/rejected": -0.321990042924881,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 15.011490045096238,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.5018064975738525,
      "logits/rejected": -2.4190878868103027,
      "logps/chosen": -328.02960205078125,
      "logps/rejected": -310.5951843261719,
      "loss": 0.6089,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.22072939574718475,
      "rewards/margins": 0.22173993289470673,
      "rewards/rejected": -0.44246941804885864,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 12.064286648165384,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -2.316053628921509,
      "logits/rejected": -2.2170588970184326,
      "logps/chosen": -309.4535217285156,
      "logps/rejected": -293.6477966308594,
      "loss": 0.6033,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.34011608362197876,
      "rewards/margins": 0.24611127376556396,
      "rewards/rejected": -0.5862273573875427,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 18.008124319426113,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -2.297001838684082,
      "logits/rejected": -2.1643242835998535,
      "logps/chosen": -329.64617919921875,
      "logps/rejected": -325.0845947265625,
      "loss": 0.5671,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.3444996774196625,
      "rewards/margins": 0.4821701645851135,
      "rewards/rejected": -0.8266698718070984,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -1.9024137258529663,
      "eval_logits/rejected": -1.719163417816162,
      "eval_logps/chosen": -347.71966552734375,
      "eval_logps/rejected": -351.6959533691406,
      "eval_loss": 0.5741031765937805,
      "eval_rewards/accuracies": 0.6896551847457886,
      "eval_rewards/chosen": -0.6262954473495483,
      "eval_rewards/margins": 0.4254392087459564,
      "eval_rewards/rejected": -1.0517346858978271,
      "eval_runtime": 94.9312,
      "eval_samples_per_second": 19.151,
      "eval_steps_per_second": 0.305,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 18.209810501109782,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -1.37660813331604,
      "logits/rejected": -1.3109476566314697,
      "logps/chosen": -306.15673828125,
      "logps/rejected": -364.8440856933594,
      "loss": 0.5638,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.5968034267425537,
      "rewards/margins": 0.4451308250427246,
      "rewards/rejected": -1.0419342517852783,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 30.619035412465948,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": -0.5667544007301331,
      "logits/rejected": -0.11496512591838837,
      "logps/chosen": -352.6363830566406,
      "logps/rejected": -321.98138427734375,
      "loss": 0.5429,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.5322494506835938,
      "rewards/margins": 0.4666845202445984,
      "rewards/rejected": -0.9989339709281921,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 25.16692077643166,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": 0.31375107169151306,
      "logits/rejected": 0.7534220814704895,
      "logps/chosen": -333.3154602050781,
      "logps/rejected": -383.91571044921875,
      "loss": 0.5276,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.756668746471405,
      "rewards/margins": 0.7230855226516724,
      "rewards/rejected": -1.479754090309143,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 21.8628335306802,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": 0.12007780373096466,
      "logits/rejected": 0.32281678915023804,
      "logps/chosen": -341.0147705078125,
      "logps/rejected": -368.9652404785156,
      "loss": 0.5337,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.6602333188056946,
      "rewards/margins": 0.6314884424209595,
      "rewards/rejected": -1.2917217016220093,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 26.496344548145927,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": 0.6530567407608032,
      "logits/rejected": 1.3690438270568848,
      "logps/chosen": -375.08856201171875,
      "logps/rejected": -408.30181884765625,
      "loss": 0.5328,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.8673305511474609,
      "rewards/margins": 0.7217624187469482,
      "rewards/rejected": -1.5890929698944092,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": -0.10693416744470596,
      "eval_logits/rejected": 0.6796554327011108,
      "eval_logps/chosen": -340.38214111328125,
      "eval_logps/rejected": -367.83489990234375,
      "eval_loss": 0.5237244367599487,
      "eval_rewards/accuracies": 0.7155172228813171,
      "eval_rewards/chosen": -0.5529204607009888,
      "eval_rewards/margins": 0.6602039933204651,
      "eval_rewards/rejected": -1.2131245136260986,
      "eval_runtime": 95.5215,
      "eval_samples_per_second": 19.032,
      "eval_steps_per_second": 0.304,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 22.090085860189564,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": -0.6242966651916504,
      "logits/rejected": 0.1017322987318039,
      "logps/chosen": -337.90997314453125,
      "logps/rejected": -389.50592041015625,
      "loss": 0.5376,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.4958871901035309,
      "rewards/margins": 0.6162235140800476,
      "rewards/rejected": -1.1121107339859009,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 29.49090849875655,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": 0.3189530670642853,
      "logits/rejected": 0.7786465287208557,
      "logps/chosen": -349.08013916015625,
      "logps/rejected": -422.4642639160156,
      "loss": 0.5223,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.786384642124176,
      "rewards/margins": 0.660692036151886,
      "rewards/rejected": -1.4470767974853516,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 21.619610594454734,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": 0.5451809167861938,
      "logits/rejected": 1.3867518901824951,
      "logps/chosen": -417.93450927734375,
      "logps/rejected": -453.64581298828125,
      "loss": 0.5262,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.0858782529830933,
      "rewards/margins": 0.799088180065155,
      "rewards/rejected": -1.8849666118621826,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 22.879012626203956,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": 0.8325613141059875,
      "logits/rejected": 1.527114748954773,
      "logps/chosen": -372.27227783203125,
      "logps/rejected": -433.6874084472656,
      "loss": 0.5396,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.9668887853622437,
      "rewards/margins": 0.787070095539093,
      "rewards/rejected": -1.7539589405059814,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 19.71033831580365,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": 0.25724777579307556,
      "logits/rejected": 1.0539257526397705,
      "logps/chosen": -339.7835388183594,
      "logps/rejected": -395.4685363769531,
      "loss": 0.5339,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.7703949809074402,
      "rewards/margins": 0.7419241070747375,
      "rewards/rejected": -1.5123189687728882,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": 0.1333327293395996,
      "eval_logits/rejected": 1.1592669486999512,
      "eval_logps/chosen": -372.03033447265625,
      "eval_logps/rejected": -410.098388671875,
      "eval_loss": 0.5135313272476196,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -0.8694021701812744,
      "eval_rewards/margins": 0.7663572430610657,
      "eval_rewards/rejected": -1.6357594728469849,
      "eval_runtime": 94.9597,
      "eval_samples_per_second": 19.145,
      "eval_steps_per_second": 0.305,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 24.166340722703165,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": -0.02151186764240265,
      "logits/rejected": 0.7990398406982422,
      "logps/chosen": -386.0631408691406,
      "logps/rejected": -439.3416442871094,
      "loss": 0.5165,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.859545111656189,
      "rewards/margins": 0.791865348815918,
      "rewards/rejected": -1.651410460472107,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 29.831111320846208,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": 0.28069013357162476,
      "logits/rejected": 0.9275471568107605,
      "logps/chosen": -352.3778076171875,
      "logps/rejected": -410.13238525390625,
      "loss": 0.5119,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.7737405300140381,
      "rewards/margins": 0.7621678113937378,
      "rewards/rejected": -1.5359084606170654,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 27.12612279862232,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": 0.31911998987197876,
      "logits/rejected": 1.4982731342315674,
      "logps/chosen": -380.7137145996094,
      "logps/rejected": -428.25030517578125,
      "loss": 0.5326,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.8266562223434448,
      "rewards/margins": 0.8594551086425781,
      "rewards/rejected": -1.6861114501953125,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 30.270434326846676,
      "learning_rate": 2.5e-07,
      "logits/chosen": 0.08341093361377716,
      "logits/rejected": 0.8582345843315125,
      "logps/chosen": -345.74591064453125,
      "logps/rejected": -403.52093505859375,
      "loss": 0.5195,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.8471673727035522,
      "rewards/margins": 0.6442986726760864,
      "rewards/rejected": -1.4914662837982178,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 24.62506692170475,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": 0.6792846918106079,
      "logits/rejected": 1.6594598293304443,
      "logps/chosen": -357.1456298828125,
      "logps/rejected": -407.39794921875,
      "loss": 0.5206,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.9902165532112122,
      "rewards/margins": 0.7551018595695496,
      "rewards/rejected": -1.7453181743621826,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": 1.0596588850021362,
      "eval_logits/rejected": 2.125216007232666,
      "eval_logps/chosen": -399.8708190917969,
      "eval_logps/rejected": -448.6092529296875,
      "eval_loss": 0.5051079988479614,
      "eval_rewards/accuracies": 0.7456896305084229,
      "eval_rewards/chosen": -1.1478071212768555,
      "eval_rewards/margins": 0.8730602860450745,
      "eval_rewards/rejected": -2.020867347717285,
      "eval_runtime": 95.375,
      "eval_samples_per_second": 19.062,
      "eval_steps_per_second": 0.304,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 25.512119180845332,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": 1.0334103107452393,
      "logits/rejected": 1.9036356210708618,
      "logps/chosen": -393.27923583984375,
      "logps/rejected": -442.31634521484375,
      "loss": 0.5225,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.1329892873764038,
      "rewards/margins": 0.6801810264587402,
      "rewards/rejected": -1.8131701946258545,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 22.529726863175615,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": 0.06802092492580414,
      "logits/rejected": 0.6059505343437195,
      "logps/chosen": -379.96771240234375,
      "logps/rejected": -420.07244873046875,
      "loss": 0.5157,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.902859091758728,
      "rewards/margins": 0.6471339464187622,
      "rewards/rejected": -1.5499929189682007,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 29.61705869960123,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": 0.36484724283218384,
      "logits/rejected": 1.1159727573394775,
      "logps/chosen": -378.2865905761719,
      "logps/rejected": -446.976318359375,
      "loss": 0.5034,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.9187177419662476,
      "rewards/margins": 0.7557224035263062,
      "rewards/rejected": -1.6744401454925537,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 32.55159666302868,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": 0.4420185089111328,
      "logits/rejected": 1.4995023012161255,
      "logps/chosen": -382.71185302734375,
      "logps/rejected": -434.3760681152344,
      "loss": 0.5136,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.9455305933952332,
      "rewards/margins": 0.9165040254592896,
      "rewards/rejected": -1.862034559249878,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 23.09866796839295,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": 0.40113481879234314,
      "logits/rejected": 1.5598547458648682,
      "logps/chosen": -423.24615478515625,
      "logps/rejected": -433.6683044433594,
      "loss": 0.5161,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.1671854257583618,
      "rewards/margins": 0.6783391237258911,
      "rewards/rejected": -1.845524549484253,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": 0.468977689743042,
      "eval_logits/rejected": 1.6077337265014648,
      "eval_logps/chosen": -392.01226806640625,
      "eval_logps/rejected": -439.55352783203125,
      "eval_loss": 0.4995412528514862,
      "eval_rewards/accuracies": 0.7413793206214905,
      "eval_rewards/chosen": -1.0692216157913208,
      "eval_rewards/margins": 0.8610891103744507,
      "eval_rewards/rejected": -1.9303104877471924,
      "eval_runtime": 95.9852,
      "eval_samples_per_second": 18.94,
      "eval_steps_per_second": 0.302,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 23.93680772701649,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": 0.4794091284275055,
      "logits/rejected": 1.1960773468017578,
      "logps/chosen": -373.32708740234375,
      "logps/rejected": -457.6255798339844,
      "loss": 0.4872,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.9246217012405396,
      "rewards/margins": 0.8957231640815735,
      "rewards/rejected": -1.8203446865081787,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 33.75818947226454,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": 0.50605309009552,
      "logits/rejected": 1.3813035488128662,
      "logps/chosen": -395.261474609375,
      "logps/rejected": -462.6114196777344,
      "loss": 0.4904,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.1317824125289917,
      "rewards/margins": 0.8041917085647583,
      "rewards/rejected": -1.93597412109375,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 26.64078079133333,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": 0.7998399138450623,
      "logits/rejected": 1.5950431823730469,
      "logps/chosen": -358.0443420410156,
      "logps/rejected": -450.14898681640625,
      "loss": 0.4977,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.0693919658660889,
      "rewards/margins": 0.917527973651886,
      "rewards/rejected": -1.9869201183319092,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 25.809036872331053,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": 0.4978507161140442,
      "logits/rejected": 1.611681580543518,
      "logps/chosen": -386.6365661621094,
      "logps/rejected": -486.8155212402344,
      "loss": 0.4861,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.9396215677261353,
      "rewards/margins": 1.1603189706802368,
      "rewards/rejected": -2.099940538406372,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 25.432021542271517,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": 0.37401479482650757,
      "logits/rejected": 1.426959753036499,
      "logps/chosen": -393.84173583984375,
      "logps/rejected": -441.216796875,
      "loss": 0.5113,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -1.0162795782089233,
      "rewards/margins": 0.8663312792778015,
      "rewards/rejected": -1.8826109170913696,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": 0.28273966908454895,
      "eval_logits/rejected": 1.4599332809448242,
      "eval_logps/chosen": -384.1430969238281,
      "eval_logps/rejected": -436.3081359863281,
      "eval_loss": 0.4952593147754669,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -0.9905301332473755,
      "eval_rewards/margins": 0.9073269963264465,
      "eval_rewards/rejected": -1.8978571891784668,
      "eval_runtime": 98.0468,
      "eval_samples_per_second": 18.542,
      "eval_steps_per_second": 0.296,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 24.878023970166044,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": 0.011499330401420593,
      "logits/rejected": 1.0976569652557373,
      "logps/chosen": -381.95184326171875,
      "logps/rejected": -447.99334716796875,
      "loss": 0.4998,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.8815473318099976,
      "rewards/margins": 0.9280586242675781,
      "rewards/rejected": -1.8096059560775757,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 27.48602383662752,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": 0.2708785831928253,
      "logits/rejected": 0.9552985429763794,
      "logps/chosen": -397.5840148925781,
      "logps/rejected": -466.92999267578125,
      "loss": 0.4821,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.9426911473274231,
      "rewards/margins": 0.8755933046340942,
      "rewards/rejected": -1.8182843923568726,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 22.77493294832679,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": 0.3190244436264038,
      "logits/rejected": 1.6189016103744507,
      "logps/chosen": -393.47637939453125,
      "logps/rejected": -460.20428466796875,
      "loss": 0.5039,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -1.0399200916290283,
      "rewards/margins": 0.9450514912605286,
      "rewards/rejected": -1.9849714040756226,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 21.22789089972862,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": 0.4111763834953308,
      "logits/rejected": 1.2173652648925781,
      "logps/chosen": -375.0528869628906,
      "logps/rejected": -447.7088317871094,
      "loss": 0.5199,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.9053140878677368,
      "rewards/margins": 0.9649880528450012,
      "rewards/rejected": -1.8703022003173828,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 25.945198405726252,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": 0.3489355742931366,
      "logits/rejected": 1.733727216720581,
      "logps/chosen": -385.9686584472656,
      "logps/rejected": -407.83514404296875,
      "loss": 0.5006,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.9153203964233398,
      "rewards/margins": 0.9544361233711243,
      "rewards/rejected": -1.8697564601898193,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": 0.33653396368026733,
      "eval_logits/rejected": 1.545719861984253,
      "eval_logps/chosen": -384.1485290527344,
      "eval_logps/rejected": -438.1469421386719,
      "eval_loss": 0.49427667260169983,
      "eval_rewards/accuracies": 0.732758641242981,
      "eval_rewards/chosen": -0.9905844330787659,
      "eval_rewards/margins": 0.9256603717803955,
      "eval_rewards/rejected": -1.9162448644638062,
      "eval_runtime": 95.1799,
      "eval_samples_per_second": 19.101,
      "eval_steps_per_second": 0.305,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 29.414671515939542,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": 0.3652718663215637,
      "logits/rejected": 1.4336962699890137,
      "logps/chosen": -407.9559326171875,
      "logps/rejected": -456.4903259277344,
      "loss": 0.486,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.9434130787849426,
      "rewards/margins": 0.8941701650619507,
      "rewards/rejected": -1.837583303451538,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 23.283884603445415,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": 0.2882634997367859,
      "logits/rejected": 1.3258743286132812,
      "logps/chosen": -375.098876953125,
      "logps/rejected": -462.24639892578125,
      "loss": 0.4823,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.0388104915618896,
      "rewards/margins": 0.9280673861503601,
      "rewards/rejected": -1.9668779373168945,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 32.32447531053772,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": 0.6452657580375671,
      "logits/rejected": 1.3607791662216187,
      "logps/chosen": -378.0513916015625,
      "logps/rejected": -468.52301025390625,
      "loss": 0.504,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.1194813251495361,
      "rewards/margins": 0.9433539509773254,
      "rewards/rejected": -2.062835216522217,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.5444534902178914,
      "train_runtime": 11753.731,
      "train_samples_per_second": 4.744,
      "train_steps_per_second": 0.037
    }
  ],
  "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
}