{
  "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": 0.9889,
      "rewards/accuracies": 0.5694444179534912,
      "rewards/chosen": 0.0036439618561416864,
      "rewards/margins": 0.01108560897409916,
      "rewards/rejected": -0.007441645488142967,
      "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": 1.0023,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 0.0032594085205346346,
      "rewards/margins": -0.002268626820296049,
      "rewards/rejected": 0.00552803510800004,
      "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": 1.0025,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.0010770887602120638,
      "rewards/margins": -0.0024619889445602894,
      "rewards/rejected": 0.001384900533594191,
      "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": 1.0077,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.0035086136776953936,
      "rewards/margins": -0.007655883673578501,
      "rewards/rejected": 0.004147270228713751,
      "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": 1.0075,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.0016377497231587768,
      "rewards/margins": -0.007458710577338934,
      "rewards/rejected": 0.005820960737764835,
      "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": 0.9976,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.00012556914589367807,
      "rewards/margins": 0.0024188074748963118,
      "rewards/rejected": -0.0022932388819754124,
      "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": 0.9989,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -6.058474536985159e-05,
      "rewards/margins": 0.0010579143417999148,
      "rewards/rejected": -0.0011184990871697664,
      "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": 0.9978,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.005425452254712582,
      "rewards/margins": 0.0022025699727237225,
      "rewards/rejected": 0.0032228827476501465,
      "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": 0.9995,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0014136198442429304,
      "rewards/margins": 0.0005185690824873745,
      "rewards/rejected": 0.0008950509363785386,
      "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": 1.0006,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.0005457091028802097,
      "rewards/margins": -0.0006409892230294645,
      "rewards/rejected": 0.0011866979766637087,
      "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": 1.0016428232192993,
      "eval_rewards/accuracies": 0.49833887815475464,
      "eval_rewards/chosen": 0.0022677299566566944,
      "eval_rewards/margins": -0.0016096002655103803,
      "eval_rewards/rejected": 0.0038773303385823965,
      "eval_runtime": 145.8128,
      "eval_samples_per_second": 2.352,
      "eval_steps_per_second": 0.295,
      "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": 1.0005,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.0014460014645010233,
      "rewards/margins": -0.0005222518229857087,
      "rewards/rejected": 0.0019682529382407665,
      "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": 0.9982,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.0025884758215397596,
      "rewards/margins": 0.0018243432277813554,
      "rewards/rejected": -0.004412819631397724,
      "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": 1.0012,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.00013765673793386668,
      "rewards/margins": -0.001203052932396531,
      "rewards/rejected": 0.0013407098595052958,
      "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": 0.9958,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.0024768461007624865,
      "rewards/margins": 0.004160420503467321,
      "rewards/rejected": -0.0016835747519508004,
      "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": 0.9994,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.002465262543410063,
      "rewards/margins": 0.0005540539277717471,
      "rewards/rejected": -0.003019316354766488,
      "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": 0.9934,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.005514757242053747,
      "rewards/margins": 0.006609287112951279,
      "rewards/rejected": -0.001094528939574957,
      "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": 0.991,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.005186290480196476,
      "rewards/margins": 0.009045111015439034,
      "rewards/rejected": -0.0038588200695812702,
      "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": 0.9909,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.004038581624627113,
      "rewards/margins": 0.009138843044638634,
      "rewards/rejected": -0.005100260488688946,
      "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": 1.0005,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.0007584737613797188,
      "rewards/margins": -0.0005425609415397048,
      "rewards/rejected": -0.0002159134455723688,
      "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": 0.9981,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.00272659445181489,
      "rewards/margins": 0.0019350949442014098,
      "rewards/rejected": 0.0007914996822364628,
      "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": 0.9966108798980713,
      "eval_rewards/accuracies": 0.5328072905540466,
      "eval_rewards/chosen": 0.0031679810490459204,
      "eval_rewards/margins": 0.0033402685075998306,
      "eval_rewards/rejected": -0.00017228761862497777,
      "eval_runtime": 145.8204,
      "eval_samples_per_second": 2.352,
      "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": 0.994,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.004580962937325239,
      "rewards/margins": 0.0060088313184678555,
      "rewards/rejected": -0.0014278689632192254,
      "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": 0.9985,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.0023371793795377016,
      "rewards/margins": 0.001521837548352778,
      "rewards/rejected": 0.0008153414237312973,
      "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": 0.9929,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.0059667350724339485,
      "rewards/margins": 0.007074916269630194,
      "rewards/rejected": -0.0011081816628575325,
      "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": 0.9983,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -0.0009392915526404977,
      "rewards/margins": 0.0016581962117925286,
      "rewards/rejected": -0.0025974875316023827,
      "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": 0.9947,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -6.657392077613622e-05,
      "rewards/margins": 0.005290796514600515,
      "rewards/rejected": -0.005357370711863041,
      "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": 0.9993,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.0012316217180341482,
      "rewards/margins": 0.0007347877835854888,
      "rewards/rejected": -0.001966409618034959,
      "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": 1.0054,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.0015969609376043081,
      "rewards/margins": -0.005393522325903177,
      "rewards/rejected": 0.0037965611554682255,
      "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": 0.9942,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.004867845680564642,
      "rewards/margins": 0.005807613953948021,
      "rewards/rejected": -0.0009397673420608044,
      "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": 0.9989,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.00334669416770339,
      "rewards/margins": 0.0010900094639509916,
      "rewards/rejected": 0.002256684470921755,
      "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": 0.9944,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.006510418839752674,
      "rewards/margins": 0.005625545047223568,
      "rewards/rejected": 0.0008848730358295143,
      "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": 0.9999422430992126,
      "eval_rewards/accuracies": 0.490448534488678,
      "eval_rewards/chosen": 0.003376962151378393,
      "eval_rewards/margins": 8.94198747118935e-05,
      "eval_rewards/rejected": 0.00328754261136055,
      "eval_runtime": 145.8777,
      "eval_samples_per_second": 2.351,
      "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": 1.0017,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": -2.9836966859875247e-05,
      "rewards/margins": -0.001651174039579928,
      "rewards/rejected": 0.0016213370254263282,
      "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": 0.9914,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.00215008738450706,
      "rewards/margins": 0.008550785481929779,
      "rewards/rejected": -0.006400698330253363,
      "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": 1.001,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.0010741351870819926,
      "rewards/margins": -0.0010322926100343466,
      "rewards/rejected": -4.1842577047646046e-05,
      "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": 1.0081,
      "rewards/accuracies": 0.42500001192092896,
      "rewards/chosen": -0.001107137417420745,
      "rewards/margins": -0.00807467382401228,
      "rewards/rejected": 0.006967535708099604,
      "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": 0.9878,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.00987558625638485,
      "rewards/margins": 0.012156149372458458,
      "rewards/rejected": -0.002280563348904252,
      "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": 1.0034,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.003144216490909457,
      "rewards/margins": -0.003439632710069418,
      "rewards/rejected": 0.00029541627736762166,
      "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": 0.997,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.0030023655854165554,
      "rewards/margins": 0.0030115083791315556,
      "rewards/rejected": -9.143399438471533e-06,
      "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": 0.9969,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -0.0009016336989589036,
      "rewards/margins": 0.0031216249335557222,
      "rewards/rejected": -0.004023258574306965,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.998242255619594,
      "train_runtime": 3253.3829,
      "train_samples_per_second": 0.946,
      "train_steps_per_second": 0.118
    }
  ],
  "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
}