{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 385,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.282051282051282e-08,
      "logits/chosen": -1.7278180122375488,
      "logits/rejected": -1.7377450466156006,
      "logps/chosen": -29.553977966308594,
      "logps/rejected": -42.813133239746094,
      "loss": 1.0,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.2820512820512818e-07,
      "logits/chosen": -1.866065502166748,
      "logits/rejected": -1.8703795671463013,
      "logps/chosen": -36.988380432128906,
      "logps/rejected": -33.66728210449219,
      "loss": 0.9557,
      "rewards/accuracies": 0.5694444179534912,
      "rewards/chosen": 0.014575847424566746,
      "rewards/margins": 0.04434243589639664,
      "rewards/rejected": -0.02976658195257187,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.5641025641025636e-07,
      "logits/chosen": -1.997332215309143,
      "logits/rejected": -1.999983549118042,
      "logps/chosen": -29.625896453857422,
      "logps/rejected": -29.035802841186523,
      "loss": 1.0091,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 0.013037634082138538,
      "rewards/margins": -0.009074507281184196,
      "rewards/rejected": 0.02211214043200016,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.8461538461538463e-07,
      "logits/chosen": -1.9199495315551758,
      "logits/rejected": -1.917249321937561,
      "logps/chosen": -31.421478271484375,
      "logps/rejected": -33.2115364074707,
      "loss": 1.0098,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.004308355040848255,
      "rewards/margins": -0.009847955778241158,
      "rewards/rejected": 0.005539602134376764,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438433e-07,
      "logits/chosen": -2.0169284343719482,
      "logits/rejected": -2.008178949356079,
      "logps/chosen": -32.59435272216797,
      "logps/rejected": -32.49193572998047,
      "loss": 1.0306,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.014034454710781574,
      "rewards/margins": -0.030623534694314003,
      "rewards/rejected": 0.016589080914855003,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542186e-07,
      "logits/chosen": -1.86457097530365,
      "logits/rejected": -1.8537908792495728,
      "logps/chosen": -33.56566619873047,
      "logps/rejected": -35.423240661621094,
      "loss": 1.0298,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.006550998892635107,
      "rewards/margins": -0.029834842309355736,
      "rewards/rejected": 0.02328384295105934,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941118e-07,
      "logits/chosen": -1.9449050426483154,
      "logits/rejected": -1.9468472003936768,
      "logps/chosen": -32.59955596923828,
      "logps/rejected": -33.1828498840332,
      "loss": 0.9903,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0005022765835747123,
      "rewards/margins": 0.009675229899585247,
      "rewards/rejected": -0.00917295552790165,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413548e-07,
      "logits/chosen": -2.079878330230713,
      "logits/rejected": -2.084862232208252,
      "logps/chosen": -33.98878860473633,
      "logps/rejected": -36.574462890625,
      "loss": 0.9958,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -0.00024233898147940636,
      "rewards/margins": 0.004231657367199659,
      "rewards/rejected": -0.004473996348679066,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-07,
      "logits/chosen": -1.9425569772720337,
      "logits/rejected": -1.9457191228866577,
      "logps/chosen": -34.40068054199219,
      "logps/rejected": -34.5762939453125,
      "loss": 0.9912,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.021701809018850327,
      "rewards/margins": 0.00881027989089489,
      "rewards/rejected": 0.012891530990600586,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.736716601303429e-07,
      "logits/chosen": -1.9507396221160889,
      "logits/rejected": -1.9552500247955322,
      "logps/chosen": -32.460357666015625,
      "logps/rejected": -32.354434967041016,
      "loss": 0.9979,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.005654479376971722,
      "rewards/margins": 0.002074276329949498,
      "rewards/rejected": 0.0035802037455141544,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.62624545834521e-07,
      "logits/chosen": -2.0492873191833496,
      "logits/rejected": -2.0472888946533203,
      "logps/chosen": -32.23810958862305,
      "logps/rejected": -31.260278701782227,
      "loss": 1.0026,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.0021828364115208387,
      "rewards/margins": -0.002563956892117858,
      "rewards/rejected": 0.004746791906654835,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.243312358856201,
      "eval_logits/rejected": -2.238436222076416,
      "eval_logps/chosen": -34.023216247558594,
      "eval_logps/rejected": -37.49723434448242,
      "eval_loss": 1.0065712928771973,
      "eval_rewards/accuracies": 0.49833887815475464,
      "eval_rewards/chosen": 0.009070919826626778,
      "eval_rewards/margins": -0.006438401062041521,
      "eval_rewards/rejected": 0.015509321354329586,
      "eval_runtime": 146.1739,
      "eval_samples_per_second": 2.347,
      "eval_steps_per_second": 0.294,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.4982572012636904e-07,
      "logits/chosen": -2.005169153213501,
      "logits/rejected": -2.0027499198913574,
      "logps/chosen": -33.2365837097168,
      "logps/rejected": -34.01953125,
      "loss": 1.0021,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.005784005858004093,
      "rewards/margins": -0.002089007291942835,
      "rewards/rejected": 0.007873011752963066,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777677e-07,
      "logits/chosen": -2.0166728496551514,
      "logits/rejected": -2.0083022117614746,
      "logps/chosen": -32.457847595214844,
      "logps/rejected": -32.18357467651367,
      "loss": 0.9927,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.010353903286159039,
      "rewards/margins": 0.0072973729111254215,
      "rewards/rejected": -0.017651278525590897,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.194082707715275e-07,
      "logits/chosen": -2.0462448596954346,
      "logits/rejected": -2.038203001022339,
      "logps/chosen": -30.475027084350586,
      "logps/rejected": -32.046302795410156,
      "loss": 1.0048,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.0005506269517354667,
      "rewards/margins": -0.004812211729586124,
      "rewards/rejected": 0.005362839438021183,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.020402418666621e-07,
      "logits/chosen": -1.976986289024353,
      "logits/rejected": -1.9872560501098633,
      "logps/chosen": -31.407278060913086,
      "logps/rejected": -32.543296813964844,
      "loss": 0.9834,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.009907384403049946,
      "rewards/margins": 0.016641682013869286,
      "rewards/rejected": -0.006734299007803202,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.8341962650351185e-07,
      "logits/chosen": -1.8905508518218994,
      "logits/rejected": -1.891632080078125,
      "logps/chosen": -34.20501708984375,
      "logps/rejected": -34.77235412597656,
      "loss": 0.9978,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.009861050173640251,
      "rewards/margins": 0.0022162157110869884,
      "rewards/rejected": -0.012077265419065952,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800572e-07,
      "logits/chosen": -1.94281005859375,
      "logits/rejected": -1.939327597618103,
      "logps/chosen": -36.144107818603516,
      "logps/rejected": -32.72822570800781,
      "loss": 0.9736,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.02205902896821499,
      "rewards/margins": 0.026437148451805115,
      "rewards/rejected": -0.004378115758299828,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.430433172111807e-07,
      "logits/chosen": -2.0420708656311035,
      "logits/rejected": -2.0346803665161133,
      "logps/chosen": -33.771934509277344,
      "logps/rejected": -31.371145248413086,
      "loss": 0.9638,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.020745161920785904,
      "rewards/margins": 0.036180444061756134,
      "rewards/rejected": -0.015435280278325081,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.216202642830543e-07,
      "logits/chosen": -2.0475738048553467,
      "logits/rejected": -2.0528526306152344,
      "logps/chosen": -32.524593353271484,
      "logps/rejected": -32.510643005371094,
      "loss": 0.9634,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.016154326498508453,
      "rewards/margins": 0.036555372178554535,
      "rewards/rejected": -0.020401041954755783,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.9960716642946403e-07,
      "logits/chosen": -2.048490524291992,
      "logits/rejected": -2.0457024574279785,
      "logps/chosen": -31.492746353149414,
      "logps/rejected": -31.319293975830078,
      "loss": 1.0022,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.003033895045518875,
      "rewards/margins": -0.002170243766158819,
      "rewards/rejected": -0.0008636537822894752,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.771853789806683e-07,
      "logits/chosen": -1.9185683727264404,
      "logits/rejected": -1.9232347011566162,
      "logps/chosen": -31.5926513671875,
      "logps/rejected": -32.78697204589844,
      "loss": 0.9923,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.01090637780725956,
      "rewards/margins": 0.007740379776805639,
      "rewards/rejected": 0.0031659987289458513,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.243807315826416,
      "eval_logits/rejected": -2.2389235496520996,
      "eval_logps/chosen": -34.018714904785156,
      "eval_logps/rejected": -37.517478942871094,
      "eval_loss": 0.9864435195922852,
      "eval_rewards/accuracies": 0.5328072905540466,
      "eval_rewards/chosen": 0.012671924196183681,
      "eval_rewards/margins": 0.013361074030399323,
      "eval_rewards/rejected": -0.0006891504744999111,
      "eval_runtime": 145.9847,
      "eval_samples_per_second": 2.35,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402e-07,
      "logits/chosen": -2.0318503379821777,
      "logits/rejected": -2.042539596557617,
      "logps/chosen": -31.948400497436523,
      "logps/rejected": -33.86983871459961,
      "loss": 0.976,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.018323851749300957,
      "rewards/margins": 0.024035325273871422,
      "rewards/rejected": -0.005711475852876902,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.318564697655179e-07,
      "logits/chosen": -1.9251388311386108,
      "logits/rejected": -1.9399843215942383,
      "logps/chosen": -30.099853515625,
      "logps/rejected": -31.55409812927246,
      "loss": 0.9939,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.009348717518150806,
      "rewards/margins": 0.006087350193411112,
      "rewards/rejected": 0.003261365694925189,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.093227910899832e-07,
      "logits/chosen": -1.9835479259490967,
      "logits/rejected": -1.9875112771987915,
      "logps/chosen": -33.387638092041016,
      "logps/rejected": -31.554845809936523,
      "loss": 0.9717,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.023866940289735794,
      "rewards/margins": 0.028299665078520775,
      "rewards/rejected": -0.00443272665143013,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279356e-07,
      "logits/chosen": -1.9826898574829102,
      "logits/rejected": -1.9607274532318115,
      "logps/chosen": -34.158443450927734,
      "logps/rejected": -34.963783264160156,
      "loss": 0.9934,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -0.0037571662105619907,
      "rewards/margins": 0.0066327848471701145,
      "rewards/rejected": -0.01038995012640953,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.654436768970182e-07,
      "logits/chosen": -2.024381160736084,
      "logits/rejected": -2.0210862159729004,
      "logps/chosen": -32.9254035949707,
      "logps/rejected": -36.251712799072266,
      "loss": 0.9788,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.0002662956831045449,
      "rewards/margins": 0.02116318605840206,
      "rewards/rejected": -0.021429482847452164,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.444597403062196e-07,
      "logits/chosen": -1.8911311626434326,
      "logits/rejected": -1.8886839151382446,
      "logps/chosen": -34.194557189941406,
      "logps/rejected": -35.51445770263672,
      "loss": 0.9971,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.004926486872136593,
      "rewards/margins": 0.002939151134341955,
      "rewards/rejected": -0.007865638472139835,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.2434529917578887e-07,
      "logits/chosen": -1.8759450912475586,
      "logits/rejected": -1.8734045028686523,
      "logps/chosen": -34.40558624267578,
      "logps/rejected": -31.752349853515625,
      "loss": 1.0216,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.0063878437504172325,
      "rewards/margins": -0.02157408930361271,
      "rewards/rejected": 0.015186244621872902,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603521e-07,
      "logits/chosen": -1.980015754699707,
      "logits/rejected": -1.9693737030029297,
      "logps/chosen": -35.33230209350586,
      "logps/rejected": -31.845691680908203,
      "loss": 0.9768,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.019471382722258568,
      "rewards/margins": 0.023230455815792084,
      "rewards/rejected": -0.0037590693682432175,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071453e-08,
      "logits/chosen": -2.0756278038024902,
      "logits/rejected": -2.060606002807617,
      "logps/chosen": -30.907390594482422,
      "logps/rejected": -32.64055252075195,
      "loss": 0.9956,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.01338677667081356,
      "rewards/margins": 0.0043600378558039665,
      "rewards/rejected": 0.00902673788368702,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-08,
      "logits/chosen": -1.946616768836975,
      "logits/rejected": -1.9440828561782837,
      "logps/chosen": -32.894561767578125,
      "logps/rejected": -30.812387466430664,
      "loss": 0.9775,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.026041675359010696,
      "rewards/margins": 0.022502180188894272,
      "rewards/rejected": 0.003539492143318057,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.2434821128845215,
      "eval_logits/rejected": -2.2386035919189453,
      "eval_logps/chosen": -34.017669677734375,
      "eval_logps/rejected": -37.50018310546875,
      "eval_loss": 0.9997690916061401,
      "eval_rewards/accuracies": 0.490448534488678,
      "eval_rewards/chosen": 0.013507848605513573,
      "eval_rewards/margins": 0.000357679498847574,
      "eval_rewards/rejected": 0.0131501704454422,
      "eval_runtime": 145.7344,
      "eval_samples_per_second": 2.354,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589034e-08,
      "logits/chosen": -1.9287067651748657,
      "logits/rejected": -1.925451636314392,
      "logps/chosen": -31.603496551513672,
      "logps/rejected": -33.734046936035156,
      "loss": 1.0066,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": -0.00011934786743950099,
      "rewards/margins": -0.006604696158319712,
      "rewards/rejected": 0.006485348101705313,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380912e-08,
      "logits/chosen": -1.9802377223968506,
      "logits/rejected": -1.9679291248321533,
      "logps/chosen": -34.585323333740234,
      "logps/rejected": -33.57084274291992,
      "loss": 0.9658,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.00860034953802824,
      "rewards/margins": 0.034203141927719116,
      "rewards/rejected": -0.02560279332101345,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-08,
      "logits/chosen": -2.015906810760498,
      "logits/rejected": -2.014427661895752,
      "logps/chosen": -33.49116516113281,
      "logps/rejected": -32.47978973388672,
      "loss": 1.0041,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.0042965407483279705,
      "rewards/margins": -0.004129170440137386,
      "rewards/rejected": -0.00016737030819058418,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.05793773749158e-08,
      "logits/chosen": -2.1030631065368652,
      "logits/rejected": -2.0872654914855957,
      "logps/chosen": -34.18492889404297,
      "logps/rejected": -33.08319854736328,
      "loss": 1.0323,
      "rewards/accuracies": 0.42500001192092896,
      "rewards/chosen": -0.00442854966968298,
      "rewards/margins": -0.03229869529604912,
      "rewards/rejected": 0.027870142832398415,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.251801807404168e-08,
      "logits/chosen": -1.9745471477508545,
      "logits/rejected": -1.9736032485961914,
      "logps/chosen": -33.23271942138672,
      "logps/rejected": -32.4765510559082,
      "loss": 0.9514,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.0395023450255394,
      "rewards/margins": 0.04862459748983383,
      "rewards/rejected": -0.009122253395617008,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-09,
      "logits/chosen": -1.9305438995361328,
      "logits/rejected": -1.940913438796997,
      "logps/chosen": -32.22040939331055,
      "logps/rejected": -35.28728103637695,
      "loss": 1.0138,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.012576865963637829,
      "rewards/margins": -0.013758530840277672,
      "rewards/rejected": 0.0011816651094704866,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050324e-09,
      "logits/chosen": -2.069648265838623,
      "logits/rejected": -2.0630898475646973,
      "logps/chosen": -33.63695526123047,
      "logps/rejected": -29.226470947265625,
      "loss": 0.988,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.012009462341666222,
      "rewards/margins": 0.012046033516526222,
      "rewards/rejected": -3.6573597753886133e-05,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-10,
      "logits/chosen": -1.928865671157837,
      "logits/rejected": -1.9310123920440674,
      "logps/chosen": -34.243560791015625,
      "logps/rejected": -30.892742156982422,
      "loss": 0.9875,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -0.003606534795835614,
      "rewards/margins": 0.012486499734222889,
      "rewards/rejected": -0.01609303429722786,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.21957750196580764,
      "train_runtime": 627.8531,
      "train_samples_per_second": 4.904,
      "train_steps_per_second": 0.613
    }
  ],
  "logging_steps": 10,
  "max_steps": 385,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}