{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "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": 9.13641688324014,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.6193928718566895,
      "logits/rejected": -2.552712917327881,
      "logps/chosen": -265.43743896484375,
      "logps/rejected": -236.1606903076172,
      "loss": 0.6931,
      "rewards/accuracies": 0.4312500059604645,
      "rewards/chosen": 8.363189408555627e-05,
      "rewards/margins": 0.0003909034130629152,
      "rewards/rejected": -0.00030727163539268076,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 8.694428459805497,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.6578612327575684,
      "logits/rejected": -2.575941324234009,
      "logps/chosen": -298.83441162109375,
      "logps/rejected": -274.2864685058594,
      "loss": 0.6921,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.0004915857571177185,
      "rewards/margins": 0.0012736506760120392,
      "rewards/rejected": -0.0017652364913374186,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 8.216933133006783,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.675276279449463,
      "logits/rejected": -2.601776599884033,
      "logps/chosen": -290.555908203125,
      "logps/rejected": -234.41598510742188,
      "loss": 0.6878,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.007768664509057999,
      "rewards/margins": 0.0121098468080163,
      "rewards/rejected": -0.0043411822989583015,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 8.711915321379939,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.660461902618408,
      "logits/rejected": -2.6112020015716553,
      "logps/chosen": -281.26617431640625,
      "logps/rejected": -267.74810791015625,
      "loss": 0.6757,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": 0.039030514657497406,
      "rewards/margins": 0.03858140856027603,
      "rewards/rejected": 0.0004491090658120811,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 9.886855722455959,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.6204681396484375,
      "logits/rejected": -2.6130080223083496,
      "logps/chosen": -294.60052490234375,
      "logps/rejected": -304.6225280761719,
      "loss": 0.6583,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.016506755724549294,
      "rewards/margins": 0.07089035958051682,
      "rewards/rejected": -0.05438361316919327,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 11.187614568198285,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.575127124786377,
      "logits/rejected": -2.5189614295959473,
      "logps/chosen": -294.8846130371094,
      "logps/rejected": -274.777587890625,
      "loss": 0.6358,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.05399893596768379,
      "rewards/margins": 0.11821160465478897,
      "rewards/rejected": -0.17221052944660187,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 12.12486219819406,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.4791793823242188,
      "logits/rejected": -2.43456768989563,
      "logps/chosen": -341.71612548828125,
      "logps/rejected": -300.43731689453125,
      "loss": 0.6121,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.1414516121149063,
      "rewards/margins": 0.24884521961212158,
      "rewards/rejected": -0.3902968466281891,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 12.125843765946055,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.398482084274292,
      "logits/rejected": -2.2879607677459717,
      "logps/chosen": -305.4925537109375,
      "logps/rejected": -298.49224853515625,
      "loss": 0.6049,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.1964271366596222,
      "rewards/margins": 0.2619345188140869,
      "rewards/rejected": -0.4583616256713867,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 14.181798116852052,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -2.314242124557495,
      "logits/rejected": -2.2462363243103027,
      "logps/chosen": -307.9869079589844,
      "logps/rejected": -303.9811706542969,
      "loss": 0.5963,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.20393487811088562,
      "rewards/margins": 0.3678644895553589,
      "rewards/rejected": -0.5717993974685669,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 21.310278276106065,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -1.5569207668304443,
      "logits/rejected": -1.490770936012268,
      "logps/chosen": -306.3553161621094,
      "logps/rejected": -343.92852783203125,
      "loss": 0.5718,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.3393058478832245,
      "rewards/margins": 0.3806948661804199,
      "rewards/rejected": -0.7200007438659668,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -1.366591215133667,
      "eval_logits/rejected": -1.1398286819458008,
      "eval_logps/chosen": -323.2030944824219,
      "eval_logps/rejected": -340.7015075683594,
      "eval_loss": 0.5524524450302124,
      "eval_rewards/accuracies": 0.693965494632721,
      "eval_rewards/chosen": -0.3811298906803131,
      "eval_rewards/margins": 0.5606611967086792,
      "eval_rewards/rejected": -0.9417910575866699,
      "eval_runtime": 95.9209,
      "eval_samples_per_second": 18.953,
      "eval_steps_per_second": 0.302,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 22.6242102385037,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -1.3675466775894165,
      "logits/rejected": -1.107060432434082,
      "logps/chosen": -325.51593017578125,
      "logps/rejected": -339.6572265625,
      "loss": 0.5522,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.6032779216766357,
      "rewards/margins": 0.45596203207969666,
      "rewards/rejected": -1.0592399835586548,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 26.734564490123315,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": -1.174917459487915,
      "logits/rejected": -1.0646188259124756,
      "logps/chosen": -337.2206726074219,
      "logps/rejected": -367.90960693359375,
      "loss": 0.5593,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.6140136122703552,
      "rewards/margins": 0.5504432320594788,
      "rewards/rejected": -1.1644567251205444,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 25.5727518504957,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": -0.7371357083320618,
      "logits/rejected": -0.4288739562034607,
      "logps/chosen": -328.1370544433594,
      "logps/rejected": -368.2788391113281,
      "loss": 0.5539,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.7195693850517273,
      "rewards/margins": 0.5370928645133972,
      "rewards/rejected": -1.256662130355835,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 27.718146660570127,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": -0.9468735456466675,
      "logits/rejected": -0.6918075680732727,
      "logps/chosen": -324.7828063964844,
      "logps/rejected": -322.2585144042969,
      "loss": 0.5515,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.48597660660743713,
      "rewards/margins": 0.4936434328556061,
      "rewards/rejected": -0.9796198606491089,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 27.968029844080455,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": -0.30871185660362244,
      "logits/rejected": 0.24544291198253632,
      "logps/chosen": -354.21392822265625,
      "logps/rejected": -385.76666259765625,
      "loss": 0.5389,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.8767034411430359,
      "rewards/margins": 0.6490874290466309,
      "rewards/rejected": -1.5257909297943115,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 28.129228197799993,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": -0.545559287071228,
      "logits/rejected": -0.1624789535999298,
      "logps/chosen": -366.7821350097656,
      "logps/rejected": -417.90545654296875,
      "loss": 0.5313,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.7023102045059204,
      "rewards/margins": 0.6170581579208374,
      "rewards/rejected": -1.3193682432174683,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 22.14745919356143,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": 0.30408772826194763,
      "logits/rejected": 0.8390473127365112,
      "logps/chosen": -362.6156311035156,
      "logps/rejected": -431.7569274902344,
      "loss": 0.5007,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.9101476669311523,
      "rewards/margins": 0.7067325711250305,
      "rewards/rejected": -1.616880178451538,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 21.278511576979266,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": 0.29116564989089966,
      "logits/rejected": 1.1298694610595703,
      "logps/chosen": -386.0079650878906,
      "logps/rejected": -382.1528625488281,
      "loss": 0.5349,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.8265001177787781,
      "rewards/margins": 0.6100791692733765,
      "rewards/rejected": -1.4365794658660889,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 27.634146578624538,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": 0.34451404213905334,
      "logits/rejected": 0.971908688545227,
      "logps/chosen": -325.24951171875,
      "logps/rejected": -408.52886962890625,
      "loss": 0.5132,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.6713501811027527,
      "rewards/margins": 0.7704941034317017,
      "rewards/rejected": -1.4418442249298096,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 23.689525185525035,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": 0.3892287611961365,
      "logits/rejected": 1.2381625175476074,
      "logps/chosen": -350.74658203125,
      "logps/rejected": -392.03265380859375,
      "loss": 0.5281,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.7246497869491577,
      "rewards/margins": 0.7571171522140503,
      "rewards/rejected": -1.4817668199539185,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": -0.12046875804662704,
      "eval_logits/rejected": 1.0060540437698364,
      "eval_logps/chosen": -347.0378112792969,
      "eval_logps/rejected": -388.8913269042969,
      "eval_loss": 0.5102300047874451,
      "eval_rewards/accuracies": 0.7456896305084229,
      "eval_rewards/chosen": -0.6194772124290466,
      "eval_rewards/margins": 0.804211437702179,
      "eval_rewards/rejected": -1.4236886501312256,
      "eval_runtime": 96.4808,
      "eval_samples_per_second": 18.843,
      "eval_steps_per_second": 0.301,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 29.46326810590551,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": 0.42962178587913513,
      "logits/rejected": 1.0683366060256958,
      "logps/chosen": -348.16741943359375,
      "logps/rejected": -402.72344970703125,
      "loss": 0.5081,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.7486302256584167,
      "rewards/margins": 0.7351824045181274,
      "rewards/rejected": -1.4838125705718994,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 25.818080936093143,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": 1.011344075202942,
      "logits/rejected": 2.023766279220581,
      "logps/chosen": -427.4793395996094,
      "logps/rejected": -478.3731994628906,
      "loss": 0.4988,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -1.0479902029037476,
      "rewards/margins": 0.8860396146774292,
      "rewards/rejected": -1.9340295791625977,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 22.47549460336781,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": 0.21876761317253113,
      "logits/rejected": 0.814139723777771,
      "logps/chosen": -328.03802490234375,
      "logps/rejected": -390.96160888671875,
      "loss": 0.5217,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.600945234298706,
      "rewards/margins": 0.6524718403816223,
      "rewards/rejected": -1.2534170150756836,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 21.171293378773015,
      "learning_rate": 2.5e-07,
      "logits/chosen": -0.24384799599647522,
      "logits/rejected": 0.6541222333908081,
      "logps/chosen": -352.31732177734375,
      "logps/rejected": -387.68975830078125,
      "loss": 0.5355,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.6478549838066101,
      "rewards/margins": 0.6155428886413574,
      "rewards/rejected": -1.2633978128433228,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 20.854041628569117,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": 0.6312379837036133,
      "logits/rejected": 1.7261543273925781,
      "logps/chosen": -404.70721435546875,
      "logps/rejected": -441.50775146484375,
      "loss": 0.5067,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -1.0091339349746704,
      "rewards/margins": 0.801504909992218,
      "rewards/rejected": -1.8106390237808228,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 22.834810900622486,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": 1.0223064422607422,
      "logits/rejected": 1.8627128601074219,
      "logps/chosen": -392.29034423828125,
      "logps/rejected": -494.05767822265625,
      "loss": 0.5001,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -1.215757966041565,
      "rewards/margins": 0.9808815121650696,
      "rewards/rejected": -2.1966395378112793,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 24.01182646236212,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": 0.8764435052871704,
      "logits/rejected": 1.520572304725647,
      "logps/chosen": -372.30413818359375,
      "logps/rejected": -447.957763671875,
      "loss": 0.4994,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.0624181032180786,
      "rewards/margins": 0.8986889123916626,
      "rewards/rejected": -1.9611070156097412,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 27.05306764931255,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": 0.5640047788619995,
      "logits/rejected": 1.584166169166565,
      "logps/chosen": -370.31890869140625,
      "logps/rejected": -454.9625549316406,
      "loss": 0.5238,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.9732006192207336,
      "rewards/margins": 0.9384667277336121,
      "rewards/rejected": -1.9116674661636353,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 26.88772523041706,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": 0.8203527331352234,
      "logits/rejected": 1.9352061748504639,
      "logps/chosen": -392.50665283203125,
      "logps/rejected": -447.2642517089844,
      "loss": 0.5027,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -1.0477584600448608,
      "rewards/margins": 0.8857167363166809,
      "rewards/rejected": -1.9334752559661865,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 23.287905403576715,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": 0.5250275731086731,
      "logits/rejected": 1.2394458055496216,
      "logps/chosen": -408.4005432128906,
      "logps/rejected": -430.564453125,
      "loss": 0.5213,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.0245119333267212,
      "rewards/margins": 0.6925337314605713,
      "rewards/rejected": -1.717045545578003,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": 0.3984619081020355,
      "eval_logits/rejected": 1.7280592918395996,
      "eval_logps/chosen": -379.9385070800781,
      "eval_logps/rejected": -434.2354431152344,
      "eval_loss": 0.4939006567001343,
      "eval_rewards/accuracies": 0.7543103694915771,
      "eval_rewards/chosen": -0.9484842419624329,
      "eval_rewards/margins": 0.9286458492279053,
      "eval_rewards/rejected": -1.877130150794983,
      "eval_runtime": 93.6728,
      "eval_samples_per_second": 19.408,
      "eval_steps_per_second": 0.31,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 27.450541462100524,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": 0.4360331594944,
      "logits/rejected": 1.4486221075057983,
      "logps/chosen": -380.5530700683594,
      "logps/rejected": -445.94012451171875,
      "loss": 0.5239,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.0258790254592896,
      "rewards/margins": 0.7690252065658569,
      "rewards/rejected": -1.794904351234436,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 28.27778056239839,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": 0.6628949642181396,
      "logits/rejected": 2.0110418796539307,
      "logps/chosen": -364.7244873046875,
      "logps/rejected": -434.63568115234375,
      "loss": 0.5013,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.9049250483512878,
      "rewards/margins": 1.0965297222137451,
      "rewards/rejected": -2.0014548301696777,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 29.667411994858202,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": 0.9564372897148132,
      "logits/rejected": 1.7951595783233643,
      "logps/chosen": -377.6117858886719,
      "logps/rejected": -421.7931213378906,
      "loss": 0.5021,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -1.030896544456482,
      "rewards/margins": 0.7160336375236511,
      "rewards/rejected": -1.7469301223754883,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 30.895742020543867,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": 0.8280539512634277,
      "logits/rejected": 1.6557743549346924,
      "logps/chosen": -388.16351318359375,
      "logps/rejected": -436.76531982421875,
      "loss": 0.4897,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -1.001708984375,
      "rewards/margins": 0.7661042809486389,
      "rewards/rejected": -1.7678134441375732,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 27.764567632511632,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": 0.9885643124580383,
      "logits/rejected": 2.0599663257598877,
      "logps/chosen": -376.7607727050781,
      "logps/rejected": -424.14910888671875,
      "loss": 0.5125,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -1.0226970911026,
      "rewards/margins": 0.8421795964241028,
      "rewards/rejected": -1.8648765087127686,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 25.930811449352582,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": 0.9005087614059448,
      "logits/rejected": 1.6520808935165405,
      "logps/chosen": -343.540771484375,
      "logps/rejected": -431.3505859375,
      "loss": 0.4866,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.912207305431366,
      "rewards/margins": 0.8159275054931641,
      "rewards/rejected": -1.7281348705291748,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 26.612138820793557,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": 0.6678069829940796,
      "logits/rejected": 1.7626101970672607,
      "logps/chosen": -394.8943176269531,
      "logps/rejected": -445.3042907714844,
      "loss": 0.498,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.9530981183052063,
      "rewards/margins": 0.8741987943649292,
      "rewards/rejected": -1.8272969722747803,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 33.570889750971375,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": 0.7491241693496704,
      "logits/rejected": 2.0040550231933594,
      "logps/chosen": -383.720703125,
      "logps/rejected": -444.7874450683594,
      "loss": 0.5038,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.8734270930290222,
      "rewards/margins": 0.9978850483894348,
      "rewards/rejected": -1.871312141418457,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 27.00226950820523,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": 1.115740180015564,
      "logits/rejected": 2.061350107192993,
      "logps/chosen": -346.07574462890625,
      "logps/rejected": -409.1957092285156,
      "loss": 0.5106,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.9411827325820923,
      "rewards/margins": 0.8577120900154114,
      "rewards/rejected": -1.7988946437835693,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 29.536262595417668,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": 1.2691802978515625,
      "logits/rejected": 2.047137975692749,
      "logps/chosen": -316.4430236816406,
      "logps/rejected": -404.92962646484375,
      "loss": 0.5037,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.91883385181427,
      "rewards/margins": 0.8383063077926636,
      "rewards/rejected": -1.7571399211883545,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": 0.9259076118469238,
      "eval_logits/rejected": 2.242485761642456,
      "eval_logps/chosen": -379.86492919921875,
      "eval_logps/rejected": -438.6234436035156,
      "eval_loss": 0.48977744579315186,
      "eval_rewards/accuracies": 0.7543103694915771,
      "eval_rewards/chosen": -0.9477482438087463,
      "eval_rewards/margins": 0.9732612371444702,
      "eval_rewards/rejected": -1.9210097789764404,
      "eval_runtime": 94.7635,
      "eval_samples_per_second": 19.185,
      "eval_steps_per_second": 0.306,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 34.10288961478374,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": 1.2957347631454468,
      "logits/rejected": 2.590200901031494,
      "logps/chosen": -376.28472900390625,
      "logps/rejected": -439.2782287597656,
      "loss": 0.4987,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.053328275680542,
      "rewards/margins": 0.9589872360229492,
      "rewards/rejected": -2.012315273284912,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 33.702853354259936,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": 1.3565906286239624,
      "logits/rejected": 2.3881638050079346,
      "logps/chosen": -364.3350830078125,
      "logps/rejected": -450.5853576660156,
      "loss": 0.4968,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.0489609241485596,
      "rewards/margins": 0.9981186985969543,
      "rewards/rejected": -2.047079563140869,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 31.154782672945057,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": 0.9770921468734741,
      "logits/rejected": 2.4192652702331543,
      "logps/chosen": -402.85589599609375,
      "logps/rejected": -433.6226501464844,
      "loss": 0.4895,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.9968045949935913,
      "rewards/margins": 0.8990565538406372,
      "rewards/rejected": -1.895861268043518,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.5438270043889317,
      "train_runtime": 11297.0501,
      "train_samples_per_second": 4.936,
      "train_steps_per_second": 0.039
    }
  ],
  "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
}