{
  "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": 8.800843666556295,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.6664202213287354,
      "logits/rejected": -2.5855507850646973,
      "logps/chosen": -290.6496276855469,
      "logps/rejected": -275.44122314453125,
      "loss": 0.6931,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": 0.00033013062784448266,
      "rewards/margins": 0.00021790717437397689,
      "rewards/rejected": 0.00011222347529837862,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 8.175869900678,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.6562132835388184,
      "logits/rejected": -2.5835225582122803,
      "logps/chosen": -250.1300048828125,
      "logps/rejected": -232.3563995361328,
      "loss": 0.6921,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": 0.0010141785023733974,
      "rewards/margins": 0.002109699649736285,
      "rewards/rejected": -0.001095521030947566,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 7.970196521590508,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6696975231170654,
      "logits/rejected": -2.6478660106658936,
      "logps/chosen": -275.9632263183594,
      "logps/rejected": -266.75457763671875,
      "loss": 0.6876,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": 0.00607306556776166,
      "rewards/margins": 0.009838912636041641,
      "rewards/rejected": -0.003765846835449338,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 8.67000085322498,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.595798969268799,
      "logits/rejected": -2.5203123092651367,
      "logps/chosen": -286.19683837890625,
      "logps/rejected": -249.7381591796875,
      "loss": 0.6763,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.028836142271757126,
      "rewards/margins": 0.03639785572886467,
      "rewards/rejected": -0.007561707403510809,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 13.337487209496269,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.536128520965576,
      "logits/rejected": -2.4935240745544434,
      "logps/chosen": -288.22052001953125,
      "logps/rejected": -272.15167236328125,
      "loss": 0.6558,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.006760952528566122,
      "rewards/margins": 0.0824705958366394,
      "rewards/rejected": -0.07570964097976685,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.5882346630096436,
      "eval_logits/rejected": -2.4863200187683105,
      "eval_logps/chosen": -283.5689697265625,
      "eval_logps/rejected": -255.30152893066406,
      "eval_loss": 0.6454581022262573,
      "eval_rewards/accuracies": 0.681034505367279,
      "eval_rewards/chosen": 0.00436991173774004,
      "eval_rewards/margins": 0.10091028362512589,
      "eval_rewards/rejected": -0.09654037654399872,
      "eval_runtime": 90.5984,
      "eval_samples_per_second": 20.067,
      "eval_steps_per_second": 0.32,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 11.707830850446049,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.5649328231811523,
      "logits/rejected": -2.482309341430664,
      "logps/chosen": -296.80865478515625,
      "logps/rejected": -277.1481628417969,
      "loss": 0.635,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.026310335844755173,
      "rewards/margins": 0.16292977333068848,
      "rewards/rejected": -0.18924009799957275,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 15.087943087489696,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.5654821395874023,
      "logits/rejected": -2.473752498626709,
      "logps/chosen": -302.06201171875,
      "logps/rejected": -300.6982421875,
      "loss": 0.6205,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.1509528011083603,
      "rewards/margins": 0.2617207169532776,
      "rewards/rejected": -0.4126734733581543,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 12.495701165097614,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.513960361480713,
      "logits/rejected": -2.4418704509735107,
      "logps/chosen": -310.827392578125,
      "logps/rejected": -304.6408996582031,
      "loss": 0.6186,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": -0.3509863018989563,
      "rewards/margins": 0.16184711456298828,
      "rewards/rejected": -0.5128334760665894,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 18.19180173568921,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -2.5777266025543213,
      "logits/rejected": -2.5101935863494873,
      "logps/chosen": -290.23126220703125,
      "logps/rejected": -271.0259704589844,
      "loss": 0.5969,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.19850726425647736,
      "rewards/margins": 0.29140809178352356,
      "rewards/rejected": -0.4899153709411621,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 20.909839179480045,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -2.548067808151245,
      "logits/rejected": -2.48946475982666,
      "logps/chosen": -313.51727294921875,
      "logps/rejected": -314.1280212402344,
      "loss": 0.5907,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.2900985777378082,
      "rewards/margins": 0.3498932421207428,
      "rewards/rejected": -0.6399917602539062,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -2.590954065322876,
      "eval_logits/rejected": -2.465471029281616,
      "eval_logps/chosen": -307.22003173828125,
      "eval_logps/rejected": -299.41168212890625,
      "eval_loss": 0.5893865823745728,
      "eval_rewards/accuracies": 0.7068965435028076,
      "eval_rewards/chosen": -0.2321406602859497,
      "eval_rewards/margins": 0.30550122261047363,
      "eval_rewards/rejected": -0.5376418828964233,
      "eval_runtime": 91.8942,
      "eval_samples_per_second": 19.784,
      "eval_steps_per_second": 0.316,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 21.128492078002974,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -2.2998366355895996,
      "logits/rejected": -2.1539642810821533,
      "logps/chosen": -310.5582580566406,
      "logps/rejected": -343.615234375,
      "loss": 0.58,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.4958949089050293,
      "rewards/margins": 0.41762369871139526,
      "rewards/rejected": -0.9135186076164246,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 22.76420803543292,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": -1.5430848598480225,
      "logits/rejected": -1.318800926208496,
      "logps/chosen": -358.74884033203125,
      "logps/rejected": -403.6099548339844,
      "loss": 0.576,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.6421557664871216,
      "rewards/margins": 0.5780781507492065,
      "rewards/rejected": -1.2202337980270386,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 18.25021654810524,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": -1.3487799167633057,
      "logits/rejected": -1.0795713663101196,
      "logps/chosen": -370.9684143066406,
      "logps/rejected": -413.46026611328125,
      "loss": 0.5503,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.7266704440116882,
      "rewards/margins": 0.6452904939651489,
      "rewards/rejected": -1.371960997581482,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 26.911406799310377,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": -1.0087560415267944,
      "logits/rejected": -0.8604623079299927,
      "logps/chosen": -331.37322998046875,
      "logps/rejected": -340.6692810058594,
      "loss": 0.5608,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.7133186459541321,
      "rewards/margins": 0.42135268449783325,
      "rewards/rejected": -1.1346712112426758,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 21.7192249273613,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": -1.0606236457824707,
      "logits/rejected": -0.8972805738449097,
      "logps/chosen": -345.69580078125,
      "logps/rejected": -376.9483947753906,
      "loss": 0.5657,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.6985127329826355,
      "rewards/margins": 0.4048144221305847,
      "rewards/rejected": -1.1033271551132202,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": -1.0350327491760254,
      "eval_logits/rejected": -0.6546276807785034,
      "eval_logps/chosen": -335.6878662109375,
      "eval_logps/rejected": -348.57501220703125,
      "eval_loss": 0.5474189519882202,
      "eval_rewards/accuracies": 0.7198275923728943,
      "eval_rewards/chosen": -0.5168190598487854,
      "eval_rewards/margins": 0.5124561190605164,
      "eval_rewards/rejected": -1.0292751789093018,
      "eval_runtime": 90.811,
      "eval_samples_per_second": 20.02,
      "eval_steps_per_second": 0.319,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 20.42265728106707,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": -0.5147255063056946,
      "logits/rejected": -0.3840788006782532,
      "logps/chosen": -278.23956298828125,
      "logps/rejected": -349.45599365234375,
      "loss": 0.557,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.5052815675735474,
      "rewards/margins": 0.5687552690505981,
      "rewards/rejected": -1.0740368366241455,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 27.546124980926876,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": -0.2255803644657135,
      "logits/rejected": 0.25452661514282227,
      "logps/chosen": -356.0308837890625,
      "logps/rejected": -378.4934997558594,
      "loss": 0.5436,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.6629082560539246,
      "rewards/margins": 0.6058169007301331,
      "rewards/rejected": -1.268725037574768,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 26.686484538643338,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": -0.08900181949138641,
      "logits/rejected": 0.5574623942375183,
      "logps/chosen": -388.6288146972656,
      "logps/rejected": -392.36724853515625,
      "loss": 0.5357,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.8669939041137695,
      "rewards/margins": 0.6011493802070618,
      "rewards/rejected": -1.4681432247161865,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 26.88225056627165,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": -0.923051655292511,
      "logits/rejected": -0.5179102420806885,
      "logps/chosen": -328.10980224609375,
      "logps/rejected": -390.62261962890625,
      "loss": 0.5355,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.5608564615249634,
      "rewards/margins": 0.6695758104324341,
      "rewards/rejected": -1.2304322719573975,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 28.83910063760549,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": 0.1071801632642746,
      "logits/rejected": 0.8112713098526001,
      "logps/chosen": -398.0206604003906,
      "logps/rejected": -412.6197814941406,
      "loss": 0.5303,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -1.0966007709503174,
      "rewards/margins": 0.6390595436096191,
      "rewards/rejected": -1.735660195350647,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": 0.07066183537244797,
      "eval_logits/rejected": 0.7245904207229614,
      "eval_logps/chosen": -390.5937194824219,
      "eval_logps/rejected": -417.4531555175781,
      "eval_loss": 0.5413529872894287,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": -1.0658774375915527,
      "eval_rewards/margins": 0.6521791815757751,
      "eval_rewards/rejected": -1.7180566787719727,
      "eval_runtime": 91.7218,
      "eval_samples_per_second": 19.821,
      "eval_steps_per_second": 0.316,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 22.60643364808249,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": -0.1810363531112671,
      "logits/rejected": 0.5873329043388367,
      "logps/chosen": -370.56695556640625,
      "logps/rejected": -393.52349853515625,
      "loss": 0.52,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.848964512348175,
      "rewards/margins": 0.7796515226364136,
      "rewards/rejected": -1.6286159753799438,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 25.261348780053293,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": 0.12911781668663025,
      "logits/rejected": 0.8453273773193359,
      "logps/chosen": -356.42694091796875,
      "logps/rejected": -401.22625732421875,
      "loss": 0.5152,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.7875608801841736,
      "rewards/margins": 0.8164966702461243,
      "rewards/rejected": -1.6040576696395874,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 27.60361008184946,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": 0.1694536954164505,
      "logits/rejected": 0.745651125907898,
      "logps/chosen": -374.00701904296875,
      "logps/rejected": -404.31109619140625,
      "loss": 0.5324,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.8075408935546875,
      "rewards/margins": 0.6526550054550171,
      "rewards/rejected": -1.4601958990097046,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 26.900360490730417,
      "learning_rate": 2.5e-07,
      "logits/chosen": 0.5097543001174927,
      "logits/rejected": 1.0124820470809937,
      "logps/chosen": -356.8577575683594,
      "logps/rejected": -414.31085205078125,
      "loss": 0.5326,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.8185293078422546,
      "rewards/margins": 0.7143917083740234,
      "rewards/rejected": -1.5329210758209229,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 21.852049902948178,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": 0.41708582639694214,
      "logits/rejected": 1.3239974975585938,
      "logps/chosen": -396.1962890625,
      "logps/rejected": -424.10107421875,
      "loss": 0.5472,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.8911197781562805,
      "rewards/margins": 0.6929961442947388,
      "rewards/rejected": -1.5841158628463745,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": 0.42131370306015015,
      "eval_logits/rejected": 1.265686273574829,
      "eval_logps/chosen": -364.9606018066406,
      "eval_logps/rejected": -392.8293762207031,
      "eval_loss": 0.5268200635910034,
      "eval_rewards/accuracies": 0.7155172228813171,
      "eval_rewards/chosen": -0.8095463514328003,
      "eval_rewards/margins": 0.6622725129127502,
      "eval_rewards/rejected": -1.4718190431594849,
      "eval_runtime": 91.4885,
      "eval_samples_per_second": 19.871,
      "eval_steps_per_second": 0.317,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 23.156708624000842,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": 0.08591889590024948,
      "logits/rejected": 0.6472758054733276,
      "logps/chosen": -366.85699462890625,
      "logps/rejected": -412.9854431152344,
      "loss": 0.5217,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.7147234678268433,
      "rewards/margins": 0.6098002195358276,
      "rewards/rejected": -1.324523687362671,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 26.15521858032112,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": 0.2586960792541504,
      "logits/rejected": 1.1518855094909668,
      "logps/chosen": -362.0467224121094,
      "logps/rejected": -416.77716064453125,
      "loss": 0.5074,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.6773337125778198,
      "rewards/margins": 0.9160529375076294,
      "rewards/rejected": -1.5933866500854492,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 22.25288149573626,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": 0.54390949010849,
      "logits/rejected": 1.2391692399978638,
      "logps/chosen": -429.69769287109375,
      "logps/rejected": -436.3062438964844,
      "loss": 0.5485,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -1.0754292011260986,
      "rewards/margins": 0.619503915309906,
      "rewards/rejected": -1.6949331760406494,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 23.40120757809907,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": 0.3535264730453491,
      "logits/rejected": 1.0907752513885498,
      "logps/chosen": -380.4679260253906,
      "logps/rejected": -438.02178955078125,
      "loss": 0.5205,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.9396567344665527,
      "rewards/margins": 0.6416970491409302,
      "rewards/rejected": -1.581353783607483,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 28.751131516456226,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": 0.7159294486045837,
      "logits/rejected": 1.3497685194015503,
      "logps/chosen": -370.821533203125,
      "logps/rejected": -406.04010009765625,
      "loss": 0.5517,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -1.1309765577316284,
      "rewards/margins": 0.6069232821464539,
      "rewards/rejected": -1.7378997802734375,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": 0.29943859577178955,
      "eval_logits/rejected": 1.3137139081954956,
      "eval_logps/chosen": -373.143798828125,
      "eval_logps/rejected": -407.0939636230469,
      "eval_loss": 0.5283679962158203,
      "eval_rewards/accuracies": 0.7112069129943848,
      "eval_rewards/chosen": -0.8913781642913818,
      "eval_rewards/margins": 0.723086416721344,
      "eval_rewards/rejected": -1.6144647598266602,
      "eval_runtime": 91.561,
      "eval_samples_per_second": 19.856,
      "eval_steps_per_second": 0.317,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 31.58697998681577,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": 0.399633526802063,
      "logits/rejected": 1.2564690113067627,
      "logps/chosen": -335.18218994140625,
      "logps/rejected": -429.2320861816406,
      "loss": 0.5096,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.8466620445251465,
      "rewards/margins": 0.8323869705200195,
      "rewards/rejected": -1.6790491342544556,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 25.23384166425554,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": 0.13118520379066467,
      "logits/rejected": 0.912071704864502,
      "logps/chosen": -374.3765869140625,
      "logps/rejected": -386.69384765625,
      "loss": 0.5351,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.8689004182815552,
      "rewards/margins": 0.6572145223617554,
      "rewards/rejected": -1.5261149406433105,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 23.043674301052096,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": 0.2135738581418991,
      "logits/rejected": 1.1428295373916626,
      "logps/chosen": -332.0506286621094,
      "logps/rejected": -378.5418701171875,
      "loss": 0.5351,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.7963577508926392,
      "rewards/margins": 0.6999517679214478,
      "rewards/rejected": -1.4963096380233765,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 22.897014093425614,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": 0.228415846824646,
      "logits/rejected": 1.0023858547210693,
      "logps/chosen": -356.47955322265625,
      "logps/rejected": -407.20928955078125,
      "loss": 0.5292,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.906414806842804,
      "rewards/margins": 0.5302497744560242,
      "rewards/rejected": -1.4366645812988281,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 24.081238898664065,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": 0.30141282081604004,
      "logits/rejected": 1.139885663986206,
      "logps/chosen": -382.5127868652344,
      "logps/rejected": -426.60723876953125,
      "loss": 0.4943,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.8278465270996094,
      "rewards/margins": 0.7722162008285522,
      "rewards/rejected": -1.6000627279281616,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": 0.40436258912086487,
      "eval_logits/rejected": 1.4252060651779175,
      "eval_logps/chosen": -367.28948974609375,
      "eval_logps/rejected": -402.3227233886719,
      "eval_loss": 0.5236544609069824,
      "eval_rewards/accuracies": 0.7112069129943848,
      "eval_rewards/chosen": -0.8328355550765991,
      "eval_rewards/margins": 0.7339165806770325,
      "eval_rewards/rejected": -1.5667520761489868,
      "eval_runtime": 91.5336,
      "eval_samples_per_second": 19.862,
      "eval_steps_per_second": 0.317,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 25.9302171157121,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": 0.17887906730175018,
      "logits/rejected": 1.2261936664581299,
      "logps/chosen": -403.9292297363281,
      "logps/rejected": -456.3939514160156,
      "loss": 0.5292,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.9248501062393188,
      "rewards/margins": 0.827883243560791,
      "rewards/rejected": -1.7527334690093994,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 25.758246129198,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": 0.7486616373062134,
      "logits/rejected": 1.5815865993499756,
      "logps/chosen": -384.26458740234375,
      "logps/rejected": -433.817626953125,
      "loss": 0.5355,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.9148642420768738,
      "rewards/margins": 0.8338562846183777,
      "rewards/rejected": -1.7487205266952515,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 27.707849331024978,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": 0.8750450015068054,
      "logits/rejected": 1.3391921520233154,
      "logps/chosen": -381.8013610839844,
      "logps/rejected": -423.6830139160156,
      "loss": 0.5201,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.9749946594238281,
      "rewards/margins": 0.6640049815177917,
      "rewards/rejected": -1.638999581336975,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 26.946369171438246,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": 0.8233477473258972,
      "logits/rejected": 1.701456069946289,
      "logps/chosen": -368.91595458984375,
      "logps/rejected": -411.06951904296875,
      "loss": 0.5176,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.964392364025116,
      "rewards/margins": 0.733443021774292,
      "rewards/rejected": -1.6978353261947632,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 28.888256646435288,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": 0.8198953866958618,
      "logits/rejected": 1.1675937175750732,
      "logps/chosen": -357.7641296386719,
      "logps/rejected": -417.57720947265625,
      "loss": 0.5335,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.8717674016952515,
      "rewards/margins": 0.6356672644615173,
      "rewards/rejected": -1.5074348449707031,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": 0.3891567885875702,
      "eval_logits/rejected": 1.407207727432251,
      "eval_logps/chosen": -365.5233154296875,
      "eval_logps/rejected": -399.57244873046875,
      "eval_loss": 0.5228918790817261,
      "eval_rewards/accuracies": 0.7068965435028076,
      "eval_rewards/chosen": -0.8151733875274658,
      "eval_rewards/margins": 0.7240758538246155,
      "eval_rewards/rejected": -1.5392491817474365,
      "eval_runtime": 90.7936,
      "eval_samples_per_second": 20.023,
      "eval_steps_per_second": 0.319,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 26.17469084665507,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": 0.33993715047836304,
      "logits/rejected": 1.2628642320632935,
      "logps/chosen": -358.8694152832031,
      "logps/rejected": -421.8302307128906,
      "loss": 0.536,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.8047143220901489,
      "rewards/margins": 0.8802136182785034,
      "rewards/rejected": -1.6849279403686523,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 21.435784540966043,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": 0.5830007791519165,
      "logits/rejected": 1.3784797191619873,
      "logps/chosen": -336.01025390625,
      "logps/rejected": -399.4895324707031,
      "loss": 0.5098,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.856100857257843,
      "rewards/margins": 0.8189223408699036,
      "rewards/rejected": -1.675023078918457,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 33.824936700213314,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": 0.7273231744766235,
      "logits/rejected": 1.6111886501312256,
      "logps/chosen": -351.79278564453125,
      "logps/rejected": -395.43890380859375,
      "loss": 0.5241,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.8372980356216431,
      "rewards/margins": 0.6832476258277893,
      "rewards/rejected": -1.5205457210540771,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.5600318115785581,
      "train_runtime": 11378.3731,
      "train_samples_per_second": 4.9,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 10,
  "max_steps": 436,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}