{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9989528795811519,
  "eval_steps": 200,
  "global_step": 477,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0020942408376963353,
      "grad_norm": 6692.48095703125,
      "learning_rate": 0.0,
      "logits/chosen": -0.5995081663131714,
      "logits/rejected": -0.6144353747367859,
      "logps/chosen": -1.1853606700897217,
      "logps/rejected": -1.4816904067993164,
      "loss": 20000.3594,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": 1.3230741387815215e-05,
      "rewards/margins": 7.880535122239962e-06,
      "rewards/rejected": 5.350205356080551e-06,
      "step": 1
    },
    {
      "epoch": 0.020942408376963352,
      "grad_norm": 5734.94091796875,
      "learning_rate": 9.375e-08,
      "logits/chosen": -0.6315346956253052,
      "logits/rejected": -0.6363880634307861,
      "logps/chosen": -1.1410378217697144,
      "logps/rejected": -1.3229482173919678,
      "loss": 19999.9896,
      "rewards/accuracies": 0.5243055820465088,
      "rewards/chosen": 4.113825980311958e-06,
      "rewards/margins": 5.023256562708411e-06,
      "rewards/rejected": -9.094306392398721e-07,
      "step": 10
    },
    {
      "epoch": 0.041884816753926704,
      "grad_norm": 8863.994140625,
      "learning_rate": 1.9791666666666664e-07,
      "logits/chosen": -0.6001753807067871,
      "logits/rejected": -0.6314616799354553,
      "logps/chosen": -1.164839506149292,
      "logps/rejected": -1.3119820356369019,
      "loss": 19999.1,
      "rewards/accuracies": 0.5062500238418579,
      "rewards/chosen": 1.3984734323457815e-05,
      "rewards/margins": 1.4522911442327313e-05,
      "rewards/rejected": -5.381787673286453e-07,
      "step": 20
    },
    {
      "epoch": 0.06282722513089005,
      "grad_norm": 6735.06884765625,
      "learning_rate": 3.020833333333333e-07,
      "logits/chosen": -0.6162771582603455,
      "logits/rejected": -0.6086477041244507,
      "logps/chosen": -1.1174322366714478,
      "logps/rejected": -1.2005237340927124,
      "loss": 19997.9531,
      "rewards/accuracies": 0.5718749761581421,
      "rewards/chosen": 6.728776497766376e-05,
      "rewards/margins": 3.4343527659075335e-05,
      "rewards/rejected": 3.2944233680609614e-05,
      "step": 30
    },
    {
      "epoch": 0.08376963350785341,
      "grad_norm": 8399.2373046875,
      "learning_rate": 4.0625e-07,
      "logits/chosen": -0.6147913336753845,
      "logits/rejected": -0.6396074295043945,
      "logps/chosen": -1.1159758567810059,
      "logps/rejected": -1.2775156497955322,
      "loss": 19990.5094,
      "rewards/accuracies": 0.609375,
      "rewards/chosen": 0.000209792546229437,
      "rewards/margins": 8.913524652598426e-05,
      "rewards/rejected": 0.00012065729970345274,
      "step": 40
    },
    {
      "epoch": 0.10471204188481675,
      "grad_norm": 6641.185546875,
      "learning_rate": 4.999932966293553e-07,
      "logits/chosen": -0.6261532306671143,
      "logits/rejected": -0.6666221022605896,
      "logps/chosen": -1.0477269887924194,
      "logps/rejected": -1.2515939474105835,
      "loss": 19977.9938,
      "rewards/accuracies": 0.6468750238418579,
      "rewards/chosen": 0.0006367088062688708,
      "rewards/margins": 0.00035896195913664997,
      "rewards/rejected": 0.00027774684713222086,
      "step": 50
    },
    {
      "epoch": 0.1256544502617801,
      "grad_norm": 9381.8876953125,
      "learning_rate": 4.991893270335525e-07,
      "logits/chosen": -0.6432925462722778,
      "logits/rejected": -0.6579610109329224,
      "logps/chosen": -1.1118011474609375,
      "logps/rejected": -1.2953232526779175,
      "loss": 19958.8234,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": 0.0007649611798115075,
      "rewards/margins": 0.0004104266408830881,
      "rewards/rejected": 0.000354534451616928,
      "step": 60
    },
    {
      "epoch": 0.14659685863874344,
      "grad_norm": 9766.4326171875,
      "learning_rate": 4.970496218214204e-07,
      "logits/chosen": -0.7111603021621704,
      "logits/rejected": -0.7171494364738464,
      "logps/chosen": -1.1345610618591309,
      "logps/rejected": -1.3564822673797607,
      "loss": 19938.8937,
      "rewards/accuracies": 0.653124988079071,
      "rewards/chosen": 0.0003920519375242293,
      "rewards/margins": 0.0006278123473748565,
      "rewards/rejected": -0.0002357604244025424,
      "step": 70
    },
    {
      "epoch": 0.16753926701570682,
      "grad_norm": 16313.8701171875,
      "learning_rate": 4.935856505068998e-07,
      "logits/chosen": -0.6876430511474609,
      "logits/rejected": -0.6847123503684998,
      "logps/chosen": -1.186903953552246,
      "logps/rejected": -1.5136207342147827,
      "loss": 19887.5406,
      "rewards/accuracies": 0.671875,
      "rewards/chosen": -0.0005124356830492616,
      "rewards/margins": 0.0014533853391185403,
      "rewards/rejected": -0.001965821022167802,
      "step": 80
    },
    {
      "epoch": 0.18848167539267016,
      "grad_norm": 15382.6884765625,
      "learning_rate": 4.8881598109976e-07,
      "logits/chosen": -0.6924091577529907,
      "logits/rejected": -0.6979761123657227,
      "logps/chosen": -1.2851016521453857,
      "logps/rejected": -1.6851418018341064,
      "loss": 19835.3094,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.0014557663816958666,
      "rewards/margins": 0.0021598092280328274,
      "rewards/rejected": -0.003615575609728694,
      "step": 90
    },
    {
      "epoch": 0.2094240837696335,
      "grad_norm": 48282.9921875,
      "learning_rate": 4.827661805750437e-07,
      "logits/chosen": -0.6641468405723572,
      "logits/rejected": -0.6816408634185791,
      "logps/chosen": -1.3208650350570679,
      "logps/rejected": -1.6884527206420898,
      "loss": 19789.4344,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.0019086531829088926,
      "rewards/margins": 0.0021642055362462997,
      "rewards/rejected": -0.004072858951985836,
      "step": 100
    },
    {
      "epoch": 0.23036649214659685,
      "grad_norm": 46078.9921875,
      "learning_rate": 4.75468677825789e-07,
      "logits/chosen": -0.6582318544387817,
      "logits/rejected": -0.680435299873352,
      "logps/chosen": -1.3881313800811768,
      "logps/rejected": -2.1179776191711426,
      "loss": 19730.4641,
      "rewards/accuracies": 0.684374988079071,
      "rewards/chosen": -0.00284607894718647,
      "rewards/margins": 0.00455140694975853,
      "rewards/rejected": -0.007397486362606287,
      "step": 110
    },
    {
      "epoch": 0.2513089005235602,
      "grad_norm": 37169.12109375,
      "learning_rate": 4.669625898336438e-07,
      "logits/chosen": -0.64549720287323,
      "logits/rejected": -0.646693229675293,
      "logps/chosen": -1.5555226802825928,
      "logps/rejected": -1.9665334224700928,
      "loss": 19735.8312,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.004125476814806461,
      "rewards/margins": 0.0026077167131006718,
      "rewards/rejected": -0.006733193062245846,
      "step": 120
    },
    {
      "epoch": 0.27225130890052357,
      "grad_norm": 43399.15234375,
      "learning_rate": 4.5729351198915705e-07,
      "logits/chosen": -0.6375892758369446,
      "logits/rejected": -0.6280880570411682,
      "logps/chosen": -1.7876651287078857,
      "logps/rejected": -2.316862106323242,
      "loss": 19654.7406,
      "rewards/accuracies": 0.6156250238418579,
      "rewards/chosen": -0.005961687304079533,
      "rewards/margins": 0.003864200320094824,
      "rewards/rejected": -0.00982588715851307,
      "step": 130
    },
    {
      "epoch": 0.2931937172774869,
      "grad_norm": 91015.2734375,
      "learning_rate": 4.4651327368569684e-07,
      "logits/chosen": -0.627485990524292,
      "logits/rejected": -0.633907675743103,
      "logps/chosen": -2.3138880729675293,
      "logps/rejected": -2.7282049655914307,
      "loss": 19759.4781,
      "rewards/accuracies": 0.565625011920929,
      "rewards/chosen": -0.011143961921334267,
      "rewards/margins": 0.002580726984888315,
      "rewards/rejected": -0.01372468937188387,
      "step": 140
    },
    {
      "epoch": 0.31413612565445026,
      "grad_norm": 22526.0390625,
      "learning_rate": 4.346796604970912e-07,
      "logits/chosen": -0.6385868787765503,
      "logits/rejected": -0.6374993920326233,
      "logps/chosen": -2.103062152862549,
      "logps/rejected": -3.090477705001831,
      "loss": 19572.65,
      "rewards/accuracies": 0.6468750238418579,
      "rewards/chosen": -0.009195936843752861,
      "rewards/margins": 0.008161690086126328,
      "rewards/rejected": -0.017357628792524338,
      "step": 150
    },
    {
      "epoch": 0.33507853403141363,
      "grad_norm": 22023.46875,
      "learning_rate": 4.218561044282098e-07,
      "logits/chosen": -0.6221202611923218,
      "logits/rejected": -0.624545693397522,
      "logps/chosen": -2.033363103866577,
      "logps/rejected": -2.851893663406372,
      "loss": 19593.8156,
      "rewards/accuracies": 0.6656249761581421,
      "rewards/chosen": -0.00869058445096016,
      "rewards/margins": 0.006042894907295704,
      "rewards/rejected": -0.014733481220901012,
      "step": 160
    },
    {
      "epoch": 0.35602094240837695,
      "grad_norm": 49565.4296875,
      "learning_rate": 4.081113438988443e-07,
      "logits/chosen": -0.6099945306777954,
      "logits/rejected": -0.6049014329910278,
      "logps/chosen": -1.93422532081604,
      "logps/rejected": -2.8657894134521484,
      "loss": 19481.3359,
      "rewards/accuracies": 0.703125,
      "rewards/chosen": -0.008151264861226082,
      "rewards/margins": 0.007727779448032379,
      "rewards/rejected": -0.01587904430925846,
      "step": 170
    },
    {
      "epoch": 0.3769633507853403,
      "grad_norm": 68415.34375,
      "learning_rate": 3.935190552834828e-07,
      "logits/chosen": -0.5995316505432129,
      "logits/rejected": -0.613779604434967,
      "logps/chosen": -2.2348642349243164,
      "logps/rejected": -3.100151538848877,
      "loss": 19438.5375,
      "rewards/accuracies": 0.6156250238418579,
      "rewards/chosen": -0.01036677323281765,
      "rewards/margins": 0.007535932119935751,
      "rewards/rejected": -0.017902705818414688,
      "step": 180
    },
    {
      "epoch": 0.39790575916230364,
      "grad_norm": 131100.765625,
      "learning_rate": 3.781574579820464e-07,
      "logits/chosen": -0.6112013459205627,
      "logits/rejected": -0.6285637617111206,
      "logps/chosen": -2.523817300796509,
      "logps/rejected": -3.400487184524536,
      "loss": 19433.8438,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.013927498832345009,
      "rewards/margins": 0.006916465703397989,
      "rewards/rejected": -0.02084396593272686,
      "step": 190
    },
    {
      "epoch": 0.418848167539267,
      "grad_norm": 43200.578125,
      "learning_rate": 3.621088951385353e-07,
      "logits/chosen": -0.6157111525535583,
      "logits/rejected": -0.6312779784202576,
      "logps/chosen": -2.5808610916137695,
      "logps/rejected": -3.8715834617614746,
      "loss": 19252.4328,
      "rewards/accuracies": 0.609375,
      "rewards/chosen": -0.014415362849831581,
      "rewards/margins": 0.011067177169024944,
      "rewards/rejected": -0.0254825409501791,
      "step": 200
    },
    {
      "epoch": 0.418848167539267,
      "eval_logits/chosen": -0.635429322719574,
      "eval_logits/rejected": -0.6320876479148865,
      "eval_logps/chosen": -2.7498109340667725,
      "eval_logps/rejected": -3.7316019535064697,
      "eval_loss": 2421.5791015625,
      "eval_rewards/accuracies": 0.6359999775886536,
      "eval_rewards/chosen": -0.016332482919096947,
      "eval_rewards/margins": 0.008009255863726139,
      "eval_rewards/rejected": -0.02434173971414566,
      "eval_runtime": 78.6488,
      "eval_samples_per_second": 25.43,
      "eval_steps_per_second": 1.589,
      "step": 200
    },
    {
      "epoch": 0.4397905759162304,
      "grad_norm": 85786.203125,
      "learning_rate": 3.454593922550693e-07,
      "logits/chosen": -0.5917660593986511,
      "logits/rejected": -0.612362265586853,
      "logps/chosen": -2.5843453407287598,
      "logps/rejected": -4.024473190307617,
      "loss": 19338.9469,
      "rewards/accuracies": 0.6343749761581421,
      "rewards/chosen": -0.014349165372550488,
      "rewards/margins": 0.012513542547821999,
      "rewards/rejected": -0.02686270698904991,
      "step": 210
    },
    {
      "epoch": 0.4607329842931937,
      "grad_norm": 47775.9296875,
      "learning_rate": 3.2829819606729477e-07,
      "logits/chosen": -0.6420773267745972,
      "logits/rejected": -0.6413572430610657,
      "logps/chosen": -2.5743677616119385,
      "logps/rejected": -3.8509559631347656,
      "loss": 19310.3422,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.013864854350686073,
      "rewards/margins": 0.011251034215092659,
      "rewards/rejected": -0.02511589229106903,
      "step": 220
    },
    {
      "epoch": 0.4816753926701571,
      "grad_norm": 98925.953125,
      "learning_rate": 3.1071729615293424e-07,
      "logits/chosen": -0.6178671717643738,
      "logits/rejected": -0.6074029207229614,
      "logps/chosen": -2.9529340267181396,
      "logps/rejected": -4.022775173187256,
      "loss": 19296.9719,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.0175835732370615,
      "rewards/margins": 0.009152532555162907,
      "rewards/rejected": -0.026736104860901833,
      "step": 230
    },
    {
      "epoch": 0.5026178010471204,
      "grad_norm": 52807.45703125,
      "learning_rate": 2.9281093183781403e-07,
      "logits/chosen": -0.5948532223701477,
      "logits/rejected": -0.6133206486701965,
      "logps/chosen": -2.746427059173584,
      "logps/rejected": -3.9745686054229736,
      "loss": 19241.5422,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.015993554145097733,
      "rewards/margins": 0.010695805773139,
      "rewards/rejected": -0.026689359918236732,
      "step": 240
    },
    {
      "epoch": 0.5235602094240838,
      "grad_norm": 105720.5390625,
      "learning_rate": 2.7467508704251135e-07,
      "logits/chosen": -0.6323638558387756,
      "logits/rejected": -0.623803973197937,
      "logps/chosen": -2.7065515518188477,
      "logps/rejected": -4.675766944885254,
      "loss": 18968.025,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.015767419710755348,
      "rewards/margins": 0.016725491732358932,
      "rewards/rejected": -0.03249291330575943,
      "step": 250
    },
    {
      "epoch": 0.5445026178010471,
      "grad_norm": 62916.01953125,
      "learning_rate": 2.5640697577740815e-07,
      "logits/chosen": -0.6184276342391968,
      "logits/rejected": -0.6059120893478394,
      "logps/chosen": -3.357696056365967,
      "logps/rejected": -5.134308815002441,
      "loss": 19040.1156,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.021724792197346687,
      "rewards/margins": 0.016448460519313812,
      "rewards/rejected": -0.03817325085401535,
      "step": 260
    },
    {
      "epoch": 0.5654450261780105,
      "grad_norm": 335123.875,
      "learning_rate": 2.381045210440644e-07,
      "logits/chosen": -0.5789995193481445,
      "logits/rejected": -0.5803414583206177,
      "logps/chosen": -3.6282057762145996,
      "logps/rejected": -5.462268829345703,
      "loss": 18926.1906,
      "rewards/accuracies": 0.640625,
      "rewards/chosen": -0.02483251690864563,
      "rewards/margins": 0.0165585707873106,
      "rewards/rejected": -0.04139108583331108,
      "step": 270
    },
    {
      "epoch": 0.5863874345549738,
      "grad_norm": 69821.7578125,
      "learning_rate": 2.1986582993616925e-07,
      "logits/chosen": -0.5919681787490845,
      "logits/rejected": -0.5826513171195984,
      "logps/chosen": -3.1396141052246094,
      "logps/rejected": -5.983620643615723,
      "loss": 18751.1609,
      "rewards/accuracies": 0.715624988079071,
      "rewards/chosen": -0.019451040774583817,
      "rewards/margins": 0.02658805251121521,
      "rewards/rejected": -0.046039093285799026,
      "step": 280
    },
    {
      "epoch": 0.6073298429319371,
      "grad_norm": 113937.4453125,
      "learning_rate": 2.0178866775369774e-07,
      "logits/chosen": -0.5762513279914856,
      "logits/rejected": -0.5498114824295044,
      "logps/chosen": -3.843903064727783,
      "logps/rejected": -5.456618785858154,
      "loss": 18873.35,
      "rewards/accuracies": 0.621874988079071,
      "rewards/chosen": -0.026880085468292236,
      "rewards/margins": 0.014487823471426964,
      "rewards/rejected": -0.04136790707707405,
      "step": 290
    },
    {
      "epoch": 0.6282722513089005,
      "grad_norm": 252053.953125,
      "learning_rate": 1.839699339491937e-07,
      "logits/chosen": -0.5649229884147644,
      "logits/rejected": -0.5445654988288879,
      "logps/chosen": -4.1847944259643555,
      "logps/rejected": -5.4407172203063965,
      "loss": 19027.2969,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.029785338789224625,
      "rewards/margins": 0.012018920853734016,
      "rewards/rejected": -0.04180425778031349,
      "step": 300
    },
    {
      "epoch": 0.6492146596858639,
      "grad_norm": 176141.390625,
      "learning_rate": 1.6650514271527465e-07,
      "logits/chosen": -0.5240882635116577,
      "logits/rejected": -0.5253477096557617,
      "logps/chosen": -3.756901264190674,
      "logps/rejected": -5.1446757316589355,
      "loss": 18883.9656,
      "rewards/accuracies": 0.671875,
      "rewards/chosen": -0.026183191686868668,
      "rewards/margins": 0.012593801133334637,
      "rewards/rejected": -0.03877699002623558,
      "step": 310
    },
    {
      "epoch": 0.6701570680628273,
      "grad_norm": 89414.140625,
      "learning_rate": 1.4948791099758052e-07,
      "logits/chosen": -0.528454065322876,
      "logits/rejected": -0.5345233082771301,
      "logps/chosen": -3.773169755935669,
      "logps/rejected": -5.818202972412109,
      "loss": 18683.375,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.026144877076148987,
      "rewards/margins": 0.019181720912456512,
      "rewards/rejected": -0.0453265979886055,
      "step": 320
    },
    {
      "epoch": 0.6910994764397905,
      "grad_norm": 139503.84375,
      "learning_rate": 1.3300945667758012e-07,
      "logits/chosen": -0.5055021047592163,
      "logits/rejected": -0.5068212747573853,
      "logps/chosen": -3.9517123699188232,
      "logps/rejected": -6.562854766845703,
      "loss": 18805.9875,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.027165522798895836,
      "rewards/margins": 0.024907425045967102,
      "rewards/rejected": -0.052072953432798386,
      "step": 330
    },
    {
      "epoch": 0.7120418848167539,
      "grad_norm": 540849.0625,
      "learning_rate": 1.1715810961514072e-07,
      "logits/chosen": -0.4772934019565582,
      "logits/rejected": -0.4713551104068756,
      "logps/chosen": -4.6316022872924805,
      "logps/rejected": -6.692805290222168,
      "loss": 18882.6688,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.034036919474601746,
      "rewards/margins": 0.019201481714844704,
      "rewards/rejected": -0.0532384030520916,
      "step": 340
    },
    {
      "epoch": 0.7329842931937173,
      "grad_norm": 97645.3125,
      "learning_rate": 1.0201883817182949e-07,
      "logits/chosen": -0.4823254644870758,
      "logits/rejected": -0.48729053139686584,
      "logps/chosen": -4.076716423034668,
      "logps/rejected": -6.566468715667725,
      "loss": 18605.0328,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.028939301148056984,
      "rewards/margins": 0.02290305308997631,
      "rewards/rejected": -0.051842354238033295,
      "step": 350
    },
    {
      "epoch": 0.7539267015706806,
      "grad_norm": 109121.7265625,
      "learning_rate": 8.76727937529367e-08,
      "logits/chosen": -0.4419859051704407,
      "logits/rejected": -0.4463041424751282,
      "logps/chosen": -4.230574131011963,
      "logps/rejected": -6.5960187911987305,
      "loss": 18824.3844,
      "rewards/accuracies": 0.6343749761581421,
      "rewards/chosen": -0.03069915808737278,
      "rewards/margins": 0.02159249410033226,
      "rewards/rejected": -0.05229165032505989,
      "step": 360
    },
    {
      "epoch": 0.774869109947644,
      "grad_norm": 132723.203125,
      "learning_rate": 7.419687580962222e-08,
      "logits/chosen": -0.4869758188724518,
      "logits/rejected": -0.4819525182247162,
      "logps/chosen": -4.50799036026001,
      "logps/rejected": -6.79030704498291,
      "loss": 18469.2266,
      "rewards/accuracies": 0.653124988079071,
      "rewards/chosen": -0.03301721438765526,
      "rewards/margins": 0.020858043804764748,
      "rewards/rejected": -0.053875260055065155,
      "step": 370
    },
    {
      "epoch": 0.7958115183246073,
      "grad_norm": 166290.984375,
      "learning_rate": 6.166331963291519e-08,
      "logits/chosen": -0.42744749784469604,
      "logits/rejected": -0.43046921491622925,
      "logps/chosen": -4.028041362762451,
      "logps/rejected": -6.584651947021484,
      "loss": 18593.2734,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.028549715876579285,
      "rewards/margins": 0.024071592837572098,
      "rewards/rejected": -0.052621304988861084,
      "step": 380
    },
    {
      "epoch": 0.8167539267015707,
      "grad_norm": 124524.2734375,
      "learning_rate": 5.013930914912476e-08,
      "logits/chosen": -0.4235350489616394,
      "logits/rejected": -0.42626747488975525,
      "logps/chosen": -4.570054054260254,
      "logps/rejected": -6.840919494628906,
      "loss": 18766.2375,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.034251369535923004,
      "rewards/margins": 0.021041443571448326,
      "rewards/rejected": -0.05529281497001648,
      "step": 390
    },
    {
      "epoch": 0.837696335078534,
      "grad_norm": 273308.375,
      "learning_rate": 3.968661679220467e-08,
      "logits/chosen": -0.4133988320827484,
      "logits/rejected": -0.4258216917514801,
      "logps/chosen": -4.682262897491455,
      "logps/rejected": -6.322149753570557,
      "loss": 18622.8812,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -0.03509267792105675,
      "rewards/margins": 0.015564685687422752,
      "rewards/rejected": -0.05065736174583435,
      "step": 400
    },
    {
      "epoch": 0.837696335078534,
      "eval_logits/chosen": -0.42436543107032776,
      "eval_logits/rejected": -0.4255370497703552,
      "eval_logps/chosen": -4.56782341003418,
      "eval_logps/rejected": -7.104716777801514,
      "eval_loss": 2331.077392578125,
      "eval_rewards/accuracies": 0.6880000233650208,
      "eval_rewards/chosen": -0.034512605518102646,
      "eval_rewards/margins": 0.023560285568237305,
      "eval_rewards/rejected": -0.058072883635759354,
      "eval_runtime": 78.7071,
      "eval_samples_per_second": 25.411,
      "eval_steps_per_second": 1.588,
      "step": 400
    },
    {
      "epoch": 0.8586387434554974,
      "grad_norm": 107734.2890625,
      "learning_rate": 3.036127238347164e-08,
      "logits/chosen": -0.40318989753723145,
      "logits/rejected": -0.40800532698631287,
      "logps/chosen": -4.920913219451904,
      "logps/rejected": -7.074383735656738,
      "loss": 18670.1188,
      "rewards/accuracies": 0.628125011920929,
      "rewards/chosen": -0.03762999176979065,
      "rewards/margins": 0.020070334896445274,
      "rewards/rejected": -0.057700324803590775,
      "step": 410
    },
    {
      "epoch": 0.8795811518324608,
      "grad_norm": 239613.921875,
      "learning_rate": 2.2213262793589482e-08,
      "logits/chosen": -0.41440367698669434,
      "logits/rejected": -0.41112542152404785,
      "logps/chosen": -4.344244480133057,
      "logps/rejected": -6.621962070465088,
      "loss": 18558.8672,
      "rewards/accuracies": 0.671875,
      "rewards/chosen": -0.03219941258430481,
      "rewards/margins": 0.021176358684897423,
      "rewards/rejected": -0.05337576940655708,
      "step": 420
    },
    {
      "epoch": 0.900523560209424,
      "grad_norm": 171414.421875,
      "learning_rate": 1.5286263996730026e-08,
      "logits/chosen": -0.40855735540390015,
      "logits/rejected": -0.42862606048583984,
      "logps/chosen": -4.206170558929443,
      "logps/rejected": -7.679307460784912,
      "loss": 18518.95,
      "rewards/accuracies": 0.684374988079071,
      "rewards/chosen": -0.03075767122209072,
      "rewards/margins": 0.031888239085674286,
      "rewards/rejected": -0.06264591217041016,
      "step": 430
    },
    {
      "epoch": 0.9214659685863874,
      "grad_norm": 85183.390625,
      "learning_rate": 9.617406953185136e-09,
      "logits/chosen": -0.40596073865890503,
      "logits/rejected": -0.4101337492465973,
      "logps/chosen": -4.539244174957275,
      "logps/rejected": -6.813804626464844,
      "loss": 18567.3672,
      "rewards/accuracies": 0.6468750238418579,
      "rewards/chosen": -0.034016359597444534,
      "rewards/margins": 0.021200403571128845,
      "rewards/rejected": -0.05521675944328308,
      "step": 440
    },
    {
      "epoch": 0.9424083769633508,
      "grad_norm": 239505.046875,
      "learning_rate": 5.2370785753763356e-09,
      "logits/chosen": -0.40401849150657654,
      "logits/rejected": -0.4095350205898285,
      "logps/chosen": -4.571015357971191,
      "logps/rejected": -6.951676368713379,
      "loss": 18628.2578,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.03406498581171036,
      "rewards/margins": 0.022128159180283546,
      "rewards/rejected": -0.05619315057992935,
      "step": 450
    },
    {
      "epoch": 0.9633507853403142,
      "grad_norm": 132241.640625,
      "learning_rate": 2.168758844148272e-09,
      "logits/chosen": -0.4069870412349701,
      "logits/rejected": -0.4249423146247864,
      "logps/chosen": -4.348424434661865,
      "logps/rejected": -6.827984809875488,
      "loss": 18699.7781,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.031758736819028854,
      "rewards/margins": 0.02328142523765564,
      "rewards/rejected": -0.055040162056684494,
      "step": 460
    },
    {
      "epoch": 0.9842931937172775,
      "grad_norm": 100444.328125,
      "learning_rate": 4.288949484559934e-10,
      "logits/chosen": -0.36268186569213867,
      "logits/rejected": -0.3581438660621643,
      "logps/chosen": -4.572885990142822,
      "logps/rejected": -7.3913421630859375,
      "loss": 18543.4844,
      "rewards/accuracies": 0.690625011920929,
      "rewards/chosen": -0.033623624593019485,
      "rewards/margins": 0.026994019746780396,
      "rewards/rejected": -0.06061764806509018,
      "step": 470
    },
    {
      "epoch": 0.9989528795811519,
      "step": 477,
      "total_flos": 0.0,
      "train_loss": 19203.485685272535,
      "train_runtime": 5961.2859,
      "train_samples_per_second": 10.255,
      "train_steps_per_second": 0.08
    }
  ],
  "logging_steps": 10,
  "max_steps": 477,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}