{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9906542056074765,
  "eval_steps": 50,
  "global_step": 240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06230529595015576,
      "grad_norm": 61.7364860897157,
      "learning_rate": 5e-07,
      "logits/chosen": -2.726062059402466,
      "logits/rejected": -2.7418746948242188,
      "logps/chosen": -264.51171875,
      "logps/rejected": -202.7697296142578,
      "loss": 0.6899,
      "rewards/accuracies": 0.38749998807907104,
      "rewards/chosen": 0.020583948120474815,
      "rewards/margins": 0.009055671282112598,
      "rewards/rejected": 0.011528277769684792,
      "step": 5
    },
    {
      "epoch": 0.12461059190031153,
      "grad_norm": 49.11239347044956,
      "learning_rate": 1e-06,
      "logits/chosen": -2.6860156059265137,
      "logits/rejected": -2.666875123977661,
      "logps/chosen": -245.5350799560547,
      "logps/rejected": -233.3320770263672,
      "loss": 0.6338,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": 0.5292676687240601,
      "rewards/margins": 0.1557035893201828,
      "rewards/rejected": 0.37356406450271606,
      "step": 10
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 42.978157224616126,
      "learning_rate": 9.988343845952696e-07,
      "logits/chosen": -2.480203151702881,
      "logits/rejected": -2.471378803253174,
      "logps/chosen": -235.6896514892578,
      "logps/rejected": -183.0069122314453,
      "loss": 0.6089,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 1.5770940780639648,
      "rewards/margins": 0.9468550682067871,
      "rewards/rejected": 0.6302390098571777,
      "step": 15
    },
    {
      "epoch": 0.24922118380062305,
      "grad_norm": 44.40442856187882,
      "learning_rate": 9.953429730181652e-07,
      "logits/chosen": -2.3026254177093506,
      "logits/rejected": -2.2847046852111816,
      "logps/chosen": -223.083251953125,
      "logps/rejected": -198.93948364257812,
      "loss": 0.667,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": 1.5826599597930908,
      "rewards/margins": 1.0483930110931396,
      "rewards/rejected": 0.5342669486999512,
      "step": 20
    },
    {
      "epoch": 0.3115264797507788,
      "grad_norm": 41.20960146376919,
      "learning_rate": 9.895420438411615e-07,
      "logits/chosen": -2.3145434856414795,
      "logits/rejected": -2.3113181591033936,
      "logps/chosen": -252.1824951171875,
      "logps/rejected": -197.47164916992188,
      "loss": 0.6233,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.545325756072998,
      "rewards/margins": 0.8761142492294312,
      "rewards/rejected": 0.6692114472389221,
      "step": 25
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 38.93893290936411,
      "learning_rate": 9.814586436738997e-07,
      "logits/chosen": -2.3786375522613525,
      "logits/rejected": -2.3869917392730713,
      "logps/chosen": -237.798828125,
      "logps/rejected": -212.68075561523438,
      "loss": 0.6269,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.5185164213180542,
      "rewards/margins": 1.2036529779434204,
      "rewards/rejected": 0.3148634433746338,
      "step": 30
    },
    {
      "epoch": 0.43613707165109034,
      "grad_norm": 39.19127764379894,
      "learning_rate": 9.711304610594102e-07,
      "logits/chosen": -2.440690517425537,
      "logits/rejected": -2.4353909492492676,
      "logps/chosen": -243.35546875,
      "logps/rejected": -212.8701629638672,
      "loss": 0.5609,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.8680373430252075,
      "rewards/margins": 1.0491241216659546,
      "rewards/rejected": -0.1810867041349411,
      "step": 35
    },
    {
      "epoch": 0.4984423676012461,
      "grad_norm": 37.39468496881122,
      "learning_rate": 9.586056507527264e-07,
      "logits/chosen": -2.4579291343688965,
      "logits/rejected": -2.4394490718841553,
      "logps/chosen": -245.80770874023438,
      "logps/rejected": -226.79977416992188,
      "loss": 0.5379,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.9454948306083679,
      "rewards/margins": 1.2604901790618896,
      "rewards/rejected": -0.31499534845352173,
      "step": 40
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 36.57554162682888,
      "learning_rate": 9.439426092011875e-07,
      "logits/chosen": -2.4525513648986816,
      "logits/rejected": -2.426307201385498,
      "logps/chosen": -268.70843505859375,
      "logps/rejected": -200.7107391357422,
      "loss": 0.5435,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.9689223170280457,
      "rewards/margins": 1.6185741424560547,
      "rewards/rejected": -0.6496520042419434,
      "step": 45
    },
    {
      "epoch": 0.6230529595015576,
      "grad_norm": 35.85551989361158,
      "learning_rate": 9.272097022732443e-07,
      "logits/chosen": -2.3485758304595947,
      "logits/rejected": -2.3458869457244873,
      "logps/chosen": -254.47097778320312,
      "logps/rejected": -219.59909057617188,
      "loss": 0.5367,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.0948731899261475,
      "rewards/margins": 1.4575660228729248,
      "rewards/rejected": -0.3626927435398102,
      "step": 50
    },
    {
      "epoch": 0.6230529595015576,
      "eval_logits/chosen": -2.3203799724578857,
      "eval_logits/rejected": -2.3100695610046387,
      "eval_logps/chosen": -242.62884521484375,
      "eval_logps/rejected": -204.77874755859375,
      "eval_loss": 0.5754929780960083,
      "eval_rewards/accuracies": 0.7604166865348816,
      "eval_rewards/chosen": 1.1959108114242554,
      "eval_rewards/margins": 1.3574522733688354,
      "eval_rewards/rejected": -0.16154144704341888,
      "eval_runtime": 153.3612,
      "eval_samples_per_second": 14.867,
      "eval_steps_per_second": 0.235,
      "step": 50
    },
    {
      "epoch": 0.6853582554517134,
      "grad_norm": 38.90911338348836,
      "learning_rate": 9.084849465052209e-07,
      "logits/chosen": -2.343371629714966,
      "logits/rejected": -2.3140501976013184,
      "logps/chosen": -232.2624969482422,
      "logps/rejected": -216.8009033203125,
      "loss": 0.5809,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 1.4138634204864502,
      "rewards/margins": 1.6005926132202148,
      "rewards/rejected": -0.18672919273376465,
      "step": 55
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 47.023889326268446,
      "learning_rate": 8.878556453522099e-07,
      "logits/chosen": -2.3421199321746826,
      "logits/rejected": -2.3611013889312744,
      "logps/chosen": -240.03848266601562,
      "logps/rejected": -183.4457244873047,
      "loss": 0.5443,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": 1.0544302463531494,
      "rewards/margins": 1.1835196018218994,
      "rewards/rejected": -0.12908944487571716,
      "step": 60
    },
    {
      "epoch": 0.8099688473520249,
      "grad_norm": 34.095725802748696,
      "learning_rate": 8.654179821390621e-07,
      "logits/chosen": -2.3532650470733643,
      "logits/rejected": -2.328545570373535,
      "logps/chosen": -241.86349487304688,
      "logps/rejected": -211.8879852294922,
      "loss": 0.5031,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.9804493188858032,
      "rewards/margins": 1.4752452373504639,
      "rewards/rejected": -0.4947957992553711,
      "step": 65
    },
    {
      "epoch": 0.8722741433021807,
      "grad_norm": 38.629908145531026,
      "learning_rate": 8.41276571609327e-07,
      "logits/chosen": -2.37026309967041,
      "logits/rejected": -2.341160535812378,
      "logps/chosen": -264.19927978515625,
      "logps/rejected": -228.86752319335938,
      "loss": 0.5605,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.7270997762680054,
      "rewards/margins": 1.4355695247650146,
      "rewards/rejected": -0.708469808101654,
      "step": 70
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 30.048777436008372,
      "learning_rate": 8.155439721630264e-07,
      "logits/chosen": -2.3301854133605957,
      "logits/rejected": -2.333800792694092,
      "logps/chosen": -232.375732421875,
      "logps/rejected": -189.99459838867188,
      "loss": 0.4699,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 0.896081268787384,
      "rewards/margins": 1.9089847803115845,
      "rewards/rejected": -1.0129035711288452,
      "step": 75
    },
    {
      "epoch": 0.9968847352024922,
      "grad_norm": 33.228661608807776,
      "learning_rate": 7.883401610574336e-07,
      "logits/chosen": -2.417353391647339,
      "logits/rejected": -2.390876054763794,
      "logps/chosen": -240.1530303955078,
      "logps/rejected": -214.0542449951172,
      "loss": 0.4852,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.922646701335907,
      "rewards/margins": 1.8798067569732666,
      "rewards/rejected": -0.9571601748466492,
      "step": 80
    },
    {
      "epoch": 1.0591900311526479,
      "grad_norm": 18.48400971470129,
      "learning_rate": 7.597919750177168e-07,
      "logits/chosen": -2.5199341773986816,
      "logits/rejected": -2.4877076148986816,
      "logps/chosen": -234.12948608398438,
      "logps/rejected": -236.8078155517578,
      "loss": 0.2646,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.3459094762802124,
      "rewards/margins": 3.2457852363586426,
      "rewards/rejected": -1.8998762369155884,
      "step": 85
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 17.398245773283783,
      "learning_rate": 7.30032518865576e-07,
      "logits/chosen": -2.5517895221710205,
      "logits/rejected": -2.541144847869873,
      "logps/chosen": -249.17434692382812,
      "logps/rejected": -215.4199676513672,
      "loss": 0.2414,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.5442664623260498,
      "rewards/margins": 3.122053623199463,
      "rewards/rejected": -1.577787160873413,
      "step": 90
    },
    {
      "epoch": 1.1838006230529594,
      "grad_norm": 22.634622959038673,
      "learning_rate": 6.992005449231207e-07,
      "logits/chosen": -2.577732801437378,
      "logits/rejected": -2.5622453689575195,
      "logps/chosen": -236.0395050048828,
      "logps/rejected": -224.38516235351562,
      "loss": 0.2744,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.1463866233825684,
      "rewards/margins": 3.273930072784424,
      "rewards/rejected": -1.1275434494018555,
      "step": 95
    },
    {
      "epoch": 1.2461059190031152,
      "grad_norm": 22.411472968651037,
      "learning_rate": 6.67439806085493e-07,
      "logits/chosen": -2.611746311187744,
      "logits/rejected": -2.6113743782043457,
      "logps/chosen": -241.65982055664062,
      "logps/rejected": -233.66464233398438,
      "loss": 0.2278,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.5751986503601074,
      "rewards/margins": 3.778069019317627,
      "rewards/rejected": -1.2028696537017822,
      "step": 100
    },
    {
      "epoch": 1.2461059190031152,
      "eval_logits/chosen": -2.6071712970733643,
      "eval_logits/rejected": -2.605973958969116,
      "eval_logps/chosen": -238.49212646484375,
      "eval_logps/rejected": -209.84432983398438,
      "eval_loss": 0.5324757099151611,
      "eval_rewards/accuracies": 0.7986111044883728,
      "eval_rewards/chosen": 1.6095833778381348,
      "eval_rewards/margins": 2.2776832580566406,
      "eval_rewards/rejected": -0.6680999398231506,
      "eval_runtime": 153.2854,
      "eval_samples_per_second": 14.874,
      "eval_steps_per_second": 0.235,
      "step": 100
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 20.675405716719652,
      "learning_rate": 6.348983855785121e-07,
      "logits/chosen": -2.6054115295410156,
      "logits/rejected": -2.608341932296753,
      "logps/chosen": -237.2246551513672,
      "logps/rejected": -213.109619140625,
      "loss": 0.2368,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.0681982040405273,
      "rewards/margins": 3.7451348304748535,
      "rewards/rejected": -1.676936388015747,
      "step": 105
    },
    {
      "epoch": 1.3707165109034267,
      "grad_norm": 22.0776657882233,
      "learning_rate": 6.01728006526317e-07,
      "logits/chosen": -2.5973682403564453,
      "logits/rejected": -2.568312168121338,
      "logps/chosen": -211.4546356201172,
      "logps/rejected": -207.23056030273438,
      "loss": 0.2787,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.3856191635131836,
      "rewards/margins": 3.1070494651794434,
      "rewards/rejected": -1.7214300632476807,
      "step": 110
    },
    {
      "epoch": 1.4330218068535825,
      "grad_norm": 25.563491459257378,
      "learning_rate": 5.680833245481234e-07,
      "logits/chosen": -2.621310234069824,
      "logits/rejected": -2.581016778945923,
      "logps/chosen": -241.168701171875,
      "logps/rejected": -224.9405517578125,
      "loss": 0.2729,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.9591270685195923,
      "rewards/margins": 4.1248779296875,
      "rewards/rejected": -2.1657509803771973,
      "step": 115
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 21.35257864784403,
      "learning_rate": 5.341212066823355e-07,
      "logits/chosen": -2.6895904541015625,
      "logits/rejected": -2.684328556060791,
      "logps/chosen": -218.7779998779297,
      "logps/rejected": -225.8174285888672,
      "loss": 0.2398,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.9919631481170654,
      "rewards/margins": 3.9902572631835938,
      "rewards/rejected": -1.9982938766479492,
      "step": 120
    },
    {
      "epoch": 1.557632398753894,
      "grad_norm": 23.709989070497038,
      "learning_rate": 5e-07,
      "logits/chosen": -2.6939549446105957,
      "logits/rejected": -2.710253953933716,
      "logps/chosen": -231.3859405517578,
      "logps/rejected": -235.0880126953125,
      "loss": 0.2475,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.3965744972229004,
      "rewards/margins": 3.938891649246216,
      "rewards/rejected": -1.5423166751861572,
      "step": 125
    },
    {
      "epoch": 1.6199376947040498,
      "grad_norm": 27.658356589562743,
      "learning_rate": 4.6587879331766457e-07,
      "logits/chosen": -2.6186256408691406,
      "logits/rejected": -2.623016119003296,
      "logps/chosen": -221.36141967773438,
      "logps/rejected": -215.9159698486328,
      "loss": 0.2926,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.9702951908111572,
      "rewards/margins": 3.7315688133239746,
      "rewards/rejected": -1.761273980140686,
      "step": 130
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 19.693800201267894,
      "learning_rate": 4.3191667545187675e-07,
      "logits/chosen": -2.5465903282165527,
      "logits/rejected": -2.531930446624756,
      "logps/chosen": -225.39285278320312,
      "logps/rejected": -238.19900512695312,
      "loss": 0.2704,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 1.8806304931640625,
      "rewards/margins": 4.001955986022949,
      "rewards/rejected": -2.121325969696045,
      "step": 135
    },
    {
      "epoch": 1.7445482866043613,
      "grad_norm": 28.281038714692883,
      "learning_rate": 3.9827199347368317e-07,
      "logits/chosen": -2.4746885299682617,
      "logits/rejected": -2.454610824584961,
      "logps/chosen": -243.63681030273438,
      "logps/rejected": -211.3682098388672,
      "loss": 0.2934,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.9741780757904053,
      "rewards/margins": 3.636749744415283,
      "rewards/rejected": -1.6625711917877197,
      "step": 140
    },
    {
      "epoch": 1.8068535825545171,
      "grad_norm": 25.467568770581156,
      "learning_rate": 3.651016144214878e-07,
      "logits/chosen": -2.486293077468872,
      "logits/rejected": -2.4611892700195312,
      "logps/chosen": -249.49136352539062,
      "logps/rejected": -209.4505157470703,
      "loss": 0.317,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.520716905593872,
      "rewards/margins": 3.2239551544189453,
      "rewards/rejected": -1.7032381296157837,
      "step": 145
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 22.6157368057804,
      "learning_rate": 3.325601939145069e-07,
      "logits/chosen": -2.504690170288086,
      "logits/rejected": -2.483131170272827,
      "logps/chosen": -232.9539337158203,
      "logps/rejected": -225.02053833007812,
      "loss": 0.2926,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 1.4270459413528442,
      "rewards/margins": 3.224562168121338,
      "rewards/rejected": -1.7975162267684937,
      "step": 150
    },
    {
      "epoch": 1.8691588785046729,
      "eval_logits/chosen": -2.487774610519409,
      "eval_logits/rejected": -2.4769062995910645,
      "eval_logps/chosen": -244.09713745117188,
      "eval_logps/rejected": -217.30593872070312,
      "eval_loss": 0.5151265263557434,
      "eval_rewards/accuracies": 0.8194444179534912,
      "eval_rewards/chosen": 1.0490803718566895,
      "eval_rewards/margins": 2.463338613510132,
      "eval_rewards/rejected": -1.4142580032348633,
      "eval_runtime": 153.2856,
      "eval_samples_per_second": 14.874,
      "eval_steps_per_second": 0.235,
      "step": 150
    },
    {
      "epoch": 1.9314641744548287,
      "grad_norm": 20.708689437767955,
      "learning_rate": 3.007994550768793e-07,
      "logits/chosen": -2.499063014984131,
      "logits/rejected": -2.5030670166015625,
      "logps/chosen": -270.1714172363281,
      "logps/rejected": -220.50241088867188,
      "loss": 0.2647,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.8190641403198242,
      "rewards/margins": 3.6099021434783936,
      "rewards/rejected": -1.7908376455307007,
      "step": 155
    },
    {
      "epoch": 1.9937694704049844,
      "grad_norm": 21.602984524185192,
      "learning_rate": 2.699674811344239e-07,
      "logits/chosen": -2.497757911682129,
      "logits/rejected": -2.48551082611084,
      "logps/chosen": -232.5058135986328,
      "logps/rejected": -230.62548828125,
      "loss": 0.2871,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.4191097021102905,
      "rewards/margins": 3.648488998413086,
      "rewards/rejected": -2.229379653930664,
      "step": 160
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 12.705028614882366,
      "learning_rate": 2.4020802498228334e-07,
      "logits/chosen": -2.5586256980895996,
      "logits/rejected": -2.55572772026062,
      "logps/chosen": -233.24447631835938,
      "logps/rejected": -230.09194946289062,
      "loss": 0.1574,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.8690952062606812,
      "rewards/margins": 3.4824376106262207,
      "rewards/rejected": -1.613342046737671,
      "step": 165
    },
    {
      "epoch": 2.1183800623052957,
      "grad_norm": 15.577416747084582,
      "learning_rate": 2.1165983894256646e-07,
      "logits/chosen": -2.58217191696167,
      "logits/rejected": -2.5582995414733887,
      "logps/chosen": -236.4599609375,
      "logps/rejected": -222.7432861328125,
      "loss": 0.1544,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.19240403175354,
      "rewards/margins": 4.188112735748291,
      "rewards/rejected": -1.9957090616226196,
      "step": 170
    },
    {
      "epoch": 2.1806853582554515,
      "grad_norm": 14.738089914507803,
      "learning_rate": 1.8445602783697373e-07,
      "logits/chosen": -2.5714597702026367,
      "logits/rejected": -2.5561861991882324,
      "logps/chosen": -242.63815307617188,
      "logps/rejected": -217.72396850585938,
      "loss": 0.1464,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 2.3251590728759766,
      "rewards/margins": 4.059892177581787,
      "rewards/rejected": -1.7347338199615479,
      "step": 175
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 9.782876087080993,
      "learning_rate": 1.5872342839067304e-07,
      "logits/chosen": -2.572836399078369,
      "logits/rejected": -2.5783233642578125,
      "logps/chosen": -245.83883666992188,
      "logps/rejected": -240.05703735351562,
      "loss": 0.1449,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 2.478684902191162,
      "rewards/margins": 4.188434600830078,
      "rewards/rejected": -1.709749460220337,
      "step": 180
    },
    {
      "epoch": 2.305295950155763,
      "grad_norm": 13.185958498645032,
      "learning_rate": 1.3458201786093794e-07,
      "logits/chosen": -2.580310583114624,
      "logits/rejected": -2.5744805335998535,
      "logps/chosen": -241.9453887939453,
      "logps/rejected": -243.2576446533203,
      "loss": 0.1538,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 2.116957426071167,
      "rewards/margins": 4.271279335021973,
      "rewards/rejected": -2.1543219089508057,
      "step": 185
    },
    {
      "epoch": 2.367601246105919,
      "grad_norm": 16.171375163789843,
      "learning_rate": 1.1214435464779003e-07,
      "logits/chosen": -2.5760138034820557,
      "logits/rejected": -2.547816514968872,
      "logps/chosen": -241.6053009033203,
      "logps/rejected": -249.4426727294922,
      "loss": 0.1443,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.4218621253967285,
      "rewards/margins": 4.459113597869873,
      "rewards/rejected": -2.0372517108917236,
      "step": 190
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 16.016550245435063,
      "learning_rate": 9.1515053494779e-08,
      "logits/chosen": -2.560492753982544,
      "logits/rejected": -2.5511984825134277,
      "logps/chosen": -233.70095825195312,
      "logps/rejected": -239.994873046875,
      "loss": 0.1537,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.1391959190368652,
      "rewards/margins": 4.481522560119629,
      "rewards/rejected": -2.342327356338501,
      "step": 195
    },
    {
      "epoch": 2.4922118380062304,
      "grad_norm": 9.732992881226295,
      "learning_rate": 7.279029772675571e-08,
      "logits/chosen": -2.549442768096924,
      "logits/rejected": -2.5353286266326904,
      "logps/chosen": -228.6923370361328,
      "logps/rejected": -243.9005584716797,
      "loss": 0.1423,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.9665724039077759,
      "rewards/margins": 4.31165075302124,
      "rewards/rejected": -2.345078229904175,
      "step": 200
    },
    {
      "epoch": 2.4922118380062304,
      "eval_logits/chosen": -2.550607204437256,
      "eval_logits/rejected": -2.541142225265503,
      "eval_logps/chosen": -241.43997192382812,
      "eval_logps/rejected": -216.2832489013672,
      "eval_loss": 0.5126479864120483,
      "eval_rewards/accuracies": 0.8125,
      "eval_rewards/chosen": 1.314799189567566,
      "eval_rewards/margins": 2.626788377761841,
      "eval_rewards/rejected": -1.3119895458221436,
      "eval_runtime": 153.1578,
      "eval_samples_per_second": 14.887,
      "eval_steps_per_second": 0.235,
      "step": 200
    },
    {
      "epoch": 2.554517133956386,
      "grad_norm": 16.255837258696836,
      "learning_rate": 5.605739079881239e-08,
      "logits/chosen": -2.551528215408325,
      "logits/rejected": -2.5295708179473877,
      "logps/chosen": -217.1802978515625,
      "logps/rejected": -209.26034545898438,
      "loss": 0.1642,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.7498328685760498,
      "rewards/margins": 3.6112678050994873,
      "rewards/rejected": -1.8614346981048584,
      "step": 205
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 16.86066212768277,
      "learning_rate": 4.139434924727358e-08,
      "logits/chosen": -2.5366992950439453,
      "logits/rejected": -2.524761438369751,
      "logps/chosen": -224.43533325195312,
      "logps/rejected": -211.21841430664062,
      "loss": 0.1408,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.0774292945861816,
      "rewards/margins": 4.29391622543335,
      "rewards/rejected": -2.216486692428589,
      "step": 210
    },
    {
      "epoch": 2.6791277258566977,
      "grad_norm": 19.27319121585229,
      "learning_rate": 2.88695389405898e-08,
      "logits/chosen": -2.551309585571289,
      "logits/rejected": -2.5264077186584473,
      "logps/chosen": -236.3275604248047,
      "logps/rejected": -233.2661590576172,
      "loss": 0.1537,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.7357547283172607,
      "rewards/margins": 4.0864739418029785,
      "rewards/rejected": -2.3507192134857178,
      "step": 215
    },
    {
      "epoch": 2.7414330218068534,
      "grad_norm": 19.82591090420402,
      "learning_rate": 1.8541356326100433e-08,
      "logits/chosen": -2.554718494415283,
      "logits/rejected": -2.5463624000549316,
      "logps/chosen": -235.51748657226562,
      "logps/rejected": -237.22164916992188,
      "loss": 0.1667,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.1637890338897705,
      "rewards/margins": 4.466395854949951,
      "rewards/rejected": -2.3026070594787598,
      "step": 220
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 17.59777463927716,
      "learning_rate": 1.0457956158838544e-08,
      "logits/chosen": -2.552560329437256,
      "logits/rejected": -2.5126850605010986,
      "logps/chosen": -238.8792724609375,
      "logps/rejected": -211.9958038330078,
      "loss": 0.156,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.6842753887176514,
      "rewards/margins": 3.9238758087158203,
      "rewards/rejected": -2.239600419998169,
      "step": 225
    },
    {
      "epoch": 2.866043613707165,
      "grad_norm": 16.44733253120753,
      "learning_rate": 4.657026981834622e-09,
      "logits/chosen": -2.5144219398498535,
      "logits/rejected": -2.518833637237549,
      "logps/chosen": -225.4315643310547,
      "logps/rejected": -228.33837890625,
      "loss": 0.1364,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.924109697341919,
      "rewards/margins": 4.412649631500244,
      "rewards/rejected": -2.4885401725769043,
      "step": 230
    },
    {
      "epoch": 2.9283489096573208,
      "grad_norm": 14.374937463169573,
      "learning_rate": 1.165615404730369e-09,
      "logits/chosen": -2.5307087898254395,
      "logits/rejected": -2.544435739517212,
      "logps/chosen": -223.05960083007812,
      "logps/rejected": -214.98324584960938,
      "loss": 0.1399,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.9505195617675781,
      "rewards/margins": 4.171855926513672,
      "rewards/rejected": -2.2213361263275146,
      "step": 235
    },
    {
      "epoch": 2.9906542056074765,
      "grad_norm": 14.364094388233855,
      "learning_rate": 0.0,
      "logits/chosen": -2.5418922901153564,
      "logits/rejected": -2.5131523609161377,
      "logps/chosen": -233.7677764892578,
      "logps/rejected": -223.32827758789062,
      "loss": 0.155,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 2.059994697570801,
      "rewards/margins": 4.537837028503418,
      "rewards/rejected": -2.477842330932617,
      "step": 240
    }
  ],
  "logging_steps": 5,
  "max_steps": 240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2829829665718272.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}