{
  "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": 0.5,
      "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": -1.8662651777267456,
      "logits/rejected": -1.8705823421478271,
      "logps/chosen": -37.01448440551758,
      "logps/rejected": -33.66516876220703,
      "loss": 0.5,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -7.885291415732354e-05,
      "rewards/margins": 0.0002720804768614471,
      "rewards/rejected": -0.00035093340557068586,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.5641025641025636e-07,
      "logits/chosen": -1.9972978830337524,
      "logits/rejected": -1.9999454021453857,
      "logps/chosen": -29.626657485961914,
      "logps/rejected": -29.065475463867188,
      "loss": 0.5,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.00015535543207079172,
      "rewards/margins": 0.00017565980670042336,
      "rewards/rejected": -2.0304325516917743e-05,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.8461538461538463e-07,
      "logits/chosen": -1.9197067022323608,
      "logits/rejected": -1.917022466659546,
      "logps/chosen": -31.42661476135254,
      "logps/rejected": -33.22706985473633,
      "loss": 0.5,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.0001051858052960597,
      "rewards/margins": -1.9103928934782743e-05,
      "rewards/rejected": -8.608187636127695e-05,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438433e-07,
      "logits/chosen": -2.0163283348083496,
      "logits/rejected": -2.0075771808624268,
      "logps/chosen": -32.59123992919922,
      "logps/rejected": -32.48072052001953,
      "loss": 0.5001,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.00014431352610699832,
      "rewards/margins": -0.00046383924200199544,
      "rewards/rejected": 0.0003195256576873362,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542186e-07,
      "logits/chosen": -1.8642585277557373,
      "logits/rejected": -1.8534786701202393,
      "logps/chosen": -33.55145263671875,
      "logps/rejected": -35.415096282958984,
      "loss": 0.5001,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 6.026435949024744e-05,
      "rewards/margins": -0.0003122163761872798,
      "rewards/rejected": 0.00037248071748763323,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941118e-07,
      "logits/chosen": -1.9458900690078735,
      "logits/rejected": -1.9478263854980469,
      "logps/chosen": -32.57926559448242,
      "logps/rejected": -33.194175720214844,
      "loss": 0.4999,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.00020916396169923246,
      "rewards/margins": 0.00043710897443816066,
      "rewards/rejected": -0.00022794504184275866,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413548e-07,
      "logits/chosen": -2.0795435905456543,
      "logits/rejected": -2.084535837173462,
      "logps/chosen": -33.986812591552734,
      "logps/rejected": -36.573219299316406,
      "loss": 0.5,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 1.6729489288991317e-05,
      "rewards/margins": 6.016326369717717e-05,
      "rewards/rejected": -4.343378896010108e-05,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-07,
      "logits/chosen": -1.9420961141586304,
      "logits/rejected": -1.9452673196792603,
      "logps/chosen": -34.41278839111328,
      "logps/rejected": -34.570655822753906,
      "loss": 0.5,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.00015023796004243195,
      "rewards/margins": -6.727596337441355e-05,
      "rewards/rejected": 0.0002175139234168455,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.736716601303429e-07,
      "logits/chosen": -1.9506244659423828,
      "logits/rejected": -1.9551340341567993,
      "logps/chosen": -32.45224380493164,
      "logps/rejected": -32.344200134277344,
      "loss": 0.5,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0001517904020147398,
      "rewards/margins": 4.694242761615897e-06,
      "rewards/rejected": 0.00014709614333696663,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.62624545834521e-07,
      "logits/chosen": -2.048786163330078,
      "logits/rejected": -2.046797275543213,
      "logps/chosen": -32.23818588256836,
      "logps/rejected": -31.283069610595703,
      "loss": 0.5,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 2.6492711185710505e-05,
      "rewards/margins": 0.0001950610603671521,
      "rewards/rejected": -0.00016856836737133563,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.243879556655884,
      "eval_logits/rejected": -2.2390010356903076,
      "eval_logps/chosen": -34.026790618896484,
      "eval_logps/rejected": -37.49333572387695,
      "eval_loss": 0.5000349879264832,
      "eval_rewards/accuracies": 0.5137043595314026,
      "eval_rewards/chosen": 7.762637687847018e-05,
      "eval_rewards/margins": -0.00015523387992288917,
      "eval_rewards/rejected": 0.00023286025680135936,
      "eval_runtime": 146.2199,
      "eval_samples_per_second": 2.346,
      "eval_steps_per_second": 0.294,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.4982572012636904e-07,
      "logits/chosen": -2.005380392074585,
      "logits/rejected": -2.0029690265655518,
      "logps/chosen": -33.23069381713867,
      "logps/rejected": -34.01119613647461,
      "loss": 0.5,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.0001312425301875919,
      "rewards/margins": -5.051253901910968e-05,
      "rewards/rejected": 0.00018175505101680756,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777677e-07,
      "logits/chosen": -2.016681432723999,
      "logits/rejected": -2.0083086490631104,
      "logps/chosen": -32.43777847290039,
      "logps/rejected": -32.1855354309082,
      "loss": 0.5,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 7.132566679501906e-05,
      "rewards/margins": 0.00031160234357230365,
      "rewards/rejected": -0.0002402766840532422,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.194082707715275e-07,
      "logits/chosen": -2.04573392868042,
      "logits/rejected": -2.0376956462860107,
      "logps/chosen": -30.474706649780273,
      "logps/rejected": -32.02512741088867,
      "loss": 0.5001,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 1.0108102287631482e-05,
      "rewards/margins": -0.0002687241358216852,
      "rewards/rejected": 0.00027883221628144383,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.020402418666621e-07,
      "logits/chosen": -1.9768989086151123,
      "logits/rejected": -1.9871835708618164,
      "logps/chosen": -31.406789779663086,
      "logps/rejected": -32.54368209838867,
      "loss": 0.5,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.00012870995851699263,
      "rewards/margins": 0.00021670905698556453,
      "rewards/rejected": -8.799910574452952e-05,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.8341962650351185e-07,
      "logits/chosen": -1.8907657861709595,
      "logits/rejected": -1.8918464183807373,
      "logps/chosen": -34.17363739013672,
      "logps/rejected": -34.758506774902344,
      "loss": 0.5,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.00019054378208238631,
      "rewards/margins": 0.00020306630176492035,
      "rewards/rejected": -1.2522534234449267e-05,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800572e-07,
      "logits/chosen": -1.9421974420547485,
      "logits/rejected": -1.9387152194976807,
      "logps/chosen": -36.14430618286133,
      "logps/rejected": -32.699371337890625,
      "loss": 0.5,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.0002737110189627856,
      "rewards/margins": 3.994655708083883e-05,
      "rewards/rejected": 0.00023376448370981961,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.430433172111807e-07,
      "logits/chosen": -2.042232036590576,
      "logits/rejected": -2.034842014312744,
      "logps/chosen": -33.76560974121094,
      "logps/rejected": -31.359146118164062,
      "loss": 0.4999,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.0003225207910872996,
      "rewards/margins": 0.00039547347114421427,
      "rewards/rejected": -7.295267278095707e-05,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.216202642830543e-07,
      "logits/chosen": -2.047433376312256,
      "logits/rejected": -2.052696704864502,
      "logps/chosen": -32.51101303100586,
      "logps/rejected": -32.49535369873047,
      "loss": 0.4999,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.00033775376505218446,
      "rewards/margins": 0.00043991307029500604,
      "rewards/rejected": -0.0001021592179313302,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.9960716642946403e-07,
      "logits/chosen": -2.048191785812378,
      "logits/rejected": -2.0454134941101074,
      "logps/chosen": -31.475040435791016,
      "logps/rejected": -31.315017700195312,
      "loss": 0.5,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.00013913081784266979,
      "rewards/margins": 0.00010714305244619027,
      "rewards/rejected": 3.1987779948394746e-05,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.771853789806683e-07,
      "logits/chosen": -1.9189927577972412,
      "logits/rejected": -1.9236657619476318,
      "logps/chosen": -31.597900390625,
      "logps/rejected": -32.774513244628906,
      "loss": 0.5,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 8.383009844692424e-05,
      "rewards/margins": -8.035473001655191e-05,
      "rewards/rejected": 0.0001641848066356033,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.244427442550659,
      "eval_logits/rejected": -2.2395591735839844,
      "eval_logps/chosen": -34.002986907958984,
      "eval_logps/rejected": -37.49692916870117,
      "eval_loss": 0.4999743700027466,
      "eval_rewards/accuracies": 0.529900312423706,
      "eval_rewards/chosen": 0.00031563788070343435,
      "eval_rewards/margins": 0.0001187350062537007,
      "eval_rewards/rejected": 0.00019690288172569126,
      "eval_runtime": 145.9131,
      "eval_samples_per_second": 2.351,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402e-07,
      "logits/chosen": -2.031794786453247,
      "logits/rejected": -2.0424611568450928,
      "logps/chosen": -31.942611694335938,
      "logps/rejected": -33.87322998046875,
      "loss": 0.4999,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0002869277377612889,
      "rewards/margins": 0.0003922684118151665,
      "rewards/rejected": -0.00010534075408941135,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.318564697655179e-07,
      "logits/chosen": -1.925797462463379,
      "logits/rejected": -1.9406683444976807,
      "logps/chosen": -30.079837799072266,
      "logps/rejected": -31.56089210510254,
      "loss": 0.4999,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.00031700366525910795,
      "rewards/margins": 0.0003441544540692121,
      "rewards/rejected": -2.715078335313592e-05,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.093227910899832e-07,
      "logits/chosen": -1.9836803674697876,
      "logits/rejected": -1.9876391887664795,
      "logps/chosen": -33.396968841552734,
      "logps/rejected": -31.5497989654541,
      "loss": 0.5,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.00020496857177931815,
      "rewards/margins": 0.0002099033008562401,
      "rewards/rejected": -4.934729986416642e-06,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279356e-07,
      "logits/chosen": -1.9834476709365845,
      "logits/rejected": -1.9614884853363037,
      "logps/chosen": -34.1593017578125,
      "logps/rejected": -34.974971771240234,
      "loss": 0.5,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -5.55333899683319e-05,
      "rewards/margins": 0.00018626711971592158,
      "rewards/rejected": -0.00024180053151212633,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.654436768970182e-07,
      "logits/chosen": -2.0252513885498047,
      "logits/rejected": -2.021942615509033,
      "logps/chosen": -32.891014099121094,
      "logps/rejected": -36.209808349609375,
      "loss": 0.4999,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": 0.00034049636451527476,
      "rewards/margins": 0.00018933510000351816,
      "rewards/rejected": 0.00015116124995984137,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.444597403062196e-07,
      "logits/chosen": -1.891709327697754,
      "logits/rejected": -1.889259934425354,
      "logps/chosen": -34.162227630615234,
      "logps/rejected": -35.52161407470703,
      "loss": 0.4999,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.0002616822312120348,
      "rewards/margins": 0.0004316047707106918,
      "rewards/rejected": -0.0001699226413620636,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.2434529917578887e-07,
      "logits/chosen": -1.8765720129013062,
      "logits/rejected": -1.8740289211273193,
      "logps/chosen": -34.38764190673828,
      "logps/rejected": -31.73568344116211,
      "loss": 0.5,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 9.958025475498289e-05,
      "rewards/margins": -0.00025685896980576217,
      "rewards/rejected": 0.00035643921000882983,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603521e-07,
      "logits/chosen": -1.980292558670044,
      "logits/rejected": -1.9696547985076904,
      "logps/chosen": -35.3267936706543,
      "logps/rejected": -31.829299926757812,
      "loss": 0.5,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": 0.00029847928090021014,
      "rewards/margins": 0.00018156312580686063,
      "rewards/rejected": 0.00011691616236930713,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071453e-08,
      "logits/chosen": -2.075955390930176,
      "logits/rejected": -2.0609400272369385,
      "logps/chosen": -30.922338485717773,
      "logps/rejected": -32.651031494140625,
      "loss": 0.5,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 1.7866250345832668e-05,
      "rewards/margins": 9.798279279493727e-06,
      "rewards/rejected": 8.067952876444906e-06,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-08,
      "logits/chosen": -1.9476518630981445,
      "logits/rejected": -1.945123314857483,
      "logps/chosen": -32.88804626464844,
      "logps/rejected": -30.81081771850586,
      "loss": 0.4999,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.00039066909812390804,
      "rewards/margins": 0.0003307634615339339,
      "rewards/rejected": 5.9905636589974165e-05,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.2444517612457275,
      "eval_logits/rejected": -2.239577054977417,
      "eval_logps/chosen": -34.00169372558594,
      "eval_logps/rejected": -37.48704528808594,
      "eval_loss": 0.49999141693115234,
      "eval_rewards/accuracies": 0.5365448594093323,
      "eval_rewards/chosen": 0.0003285549464635551,
      "eval_rewards/margins": 3.2781601476017386e-05,
      "eval_rewards/rejected": 0.00029577332315966487,
      "eval_runtime": 145.927,
      "eval_samples_per_second": 2.35,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589034e-08,
      "logits/chosen": -1.929610013961792,
      "logits/rejected": -1.9263417720794678,
      "logps/chosen": -31.55474281311035,
      "logps/rejected": -33.74074935913086,
      "loss": 0.4999,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.0004860997141804546,
      "rewards/margins": 0.0004720650613307953,
      "rewards/rejected": 1.4034636478754692e-05,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380912e-08,
      "logits/chosen": -1.980975866317749,
      "logits/rejected": -1.9686771631240845,
      "logps/chosen": -34.552978515625,
      "logps/rejected": -33.54259490966797,
      "loss": 0.4999,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.0004309539217501879,
      "rewards/margins": 0.00046852690866217017,
      "rewards/rejected": -3.757303420570679e-05,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-08,
      "logits/chosen": -2.016475200653076,
      "logits/rejected": -2.015024185180664,
      "logps/chosen": -33.44532012939453,
      "logps/rejected": -32.47471237182617,
      "loss": 0.4999,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.00040476821595802903,
      "rewards/margins": 0.0003560716868378222,
      "rewards/rejected": 4.869652912020683e-05,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.05793773749158e-08,
      "logits/chosen": -2.10372257232666,
      "logits/rejected": -2.0879228115081787,
      "logps/chosen": -34.1622200012207,
      "logps/rejected": -33.08952331542969,
      "loss": 0.5,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": 0.00017170429055113345,
      "rewards/margins": -0.00011341056961100549,
      "rewards/rejected": 0.00028511486016213894,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.251801807404168e-08,
      "logits/chosen": -1.9753808975219727,
      "logits/rejected": -1.9744256734848022,
      "logps/chosen": -33.2431755065918,
      "logps/rejected": -32.457496643066406,
      "loss": 0.4999,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 0.0003892013046424836,
      "rewards/margins": 0.00031266576843336225,
      "rewards/rejected": 7.653549255337566e-05,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-09,
      "logits/chosen": -1.9311378002166748,
      "logits/rejected": -1.9414989948272705,
      "logps/chosen": -32.21795654296875,
      "logps/rejected": -35.26204299926758,
      "loss": 0.5001,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.00013266984024085104,
      "rewards/margins": -0.0003998771426267922,
      "rewards/rejected": 0.00026720724417828023,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050324e-09,
      "logits/chosen": -2.0701680183410645,
      "logits/rejected": -2.063605546951294,
      "logps/chosen": -33.66097640991211,
      "logps/rejected": -29.19500732421875,
      "loss": 0.5001,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": -9.009057976072654e-05,
      "rewards/margins": -0.0004042644868604839,
      "rewards/rejected": 0.0003141739289276302,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-10,
      "logits/chosen": -1.9299176931381226,
      "logits/rejected": -1.9320815801620483,
      "logps/chosen": -34.23712921142578,
      "logps/rejected": -30.907180786132812,
      "loss": 0.4999,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 1.9172584870830178e-05,
      "rewards/margins": 0.00036473083309829235,
      "rewards/rejected": -0.0003455582191236317,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.4999783658362054,
      "train_runtime": 3254.8055,
      "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
}