{
  "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": 0.6931,
      "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": -1.86637282371521,
      "logits/rejected": -1.8706889152526855,
      "logps/chosen": -36.99605941772461,
      "logps/rejected": -33.66212844848633,
      "loss": 0.6929,
      "rewards/accuracies": 0.4861111044883728,
      "rewards/chosen": 0.00010540737275732681,
      "rewards/margins": 0.00042594841215759516,
      "rewards/rejected": -0.0003205410612281412,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.5641025641025636e-07,
      "logits/chosen": -1.9977174997329712,
      "logits/rejected": -2.0003600120544434,
      "logps/chosen": -29.638538360595703,
      "logps/rejected": -29.072032928466797,
      "loss": 0.6931,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 3.6563815228873864e-05,
      "rewards/margins": 0.00012243764649610966,
      "rewards/rejected": -8.587385673308745e-05,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.8461538461538463e-07,
      "logits/chosen": -1.91976797580719,
      "logits/rejected": -1.917059302330017,
      "logps/chosen": -31.427032470703125,
      "logps/rejected": -33.21014404296875,
      "loss": 0.6932,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.00010937436309177428,
      "rewards/margins": -0.000192504987353459,
      "rewards/rejected": 8.313067519338802e-05,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438433e-07,
      "logits/chosen": -2.0164756774902344,
      "logits/rejected": -2.0077216625213623,
      "logps/chosen": -32.58618927001953,
      "logps/rejected": -32.47720718383789,
      "loss": 0.6934,
      "rewards/accuracies": 0.36250001192092896,
      "rewards/chosen": -9.37797303777188e-05,
      "rewards/margins": -0.00044843609794043005,
      "rewards/rejected": 0.0003546563093550503,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542186e-07,
      "logits/chosen": -1.864516019821167,
      "logits/rejected": -1.8537566661834717,
      "logps/chosen": -33.572547912597656,
      "logps/rejected": -35.45746994018555,
      "loss": 0.6932,
      "rewards/accuracies": 0.42500001192092896,
      "rewards/chosen": -0.00015066245396155864,
      "rewards/margins": -9.940571908373386e-05,
      "rewards/rejected": -5.125668758410029e-05,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941118e-07,
      "logits/chosen": -1.9447959661483765,
      "logits/rejected": -1.946738839149475,
      "logps/chosen": -32.576927185058594,
      "logps/rejected": -33.18446350097656,
      "loss": 0.693,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.0002326248650206253,
      "rewards/margins": 0.0003634841996245086,
      "rewards/rejected": -0.00013085929094813764,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413548e-07,
      "logits/chosen": -2.079601764678955,
      "logits/rejected": -2.0845787525177,
      "logps/chosen": -33.98583221435547,
      "logps/rejected": -36.567405700683594,
      "loss": 0.6931,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 2.6532879928709008e-05,
      "rewards/margins": 1.1868818546645343e-05,
      "rewards/rejected": 1.4664057744084857e-05,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-07,
      "logits/chosen": -1.9421491622924805,
      "logits/rejected": -1.9453084468841553,
      "logps/chosen": -34.42374038696289,
      "logps/rejected": -34.57770538330078,
      "loss": 0.6932,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": 4.0700066165300086e-05,
      "rewards/margins": -0.0001062863739207387,
      "rewards/rejected": 0.0001469864509999752,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.736716601303429e-07,
      "logits/chosen": -1.9507558345794678,
      "logits/rejected": -1.9552711248397827,
      "logps/chosen": -32.4448356628418,
      "logps/rejected": -32.35200881958008,
      "loss": 0.6931,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 0.0002258731983602047,
      "rewards/margins": 0.00015688061830587685,
      "rewards/rejected": 6.899260188220069e-05,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.62624545834521e-07,
      "logits/chosen": -2.0490055084228516,
      "logits/rejected": -2.0470049381256104,
      "logps/chosen": -32.24321365356445,
      "logps/rejected": -31.265995025634766,
      "loss": 0.6932,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -2.38026914303191e-05,
      "rewards/margins": -2.5972036382881925e-05,
      "rewards/rejected": 2.169329718526569e-06,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.244168996810913,
      "eval_logits/rejected": -2.23928165435791,
      "eval_logps/chosen": -34.029258728027344,
      "eval_logps/rejected": -37.50376510620117,
      "eval_loss": 0.6931847929954529,
      "eval_rewards/accuracies": 0.46054819226264954,
      "eval_rewards/chosen": 5.296626841300167e-05,
      "eval_rewards/margins": -7.559578079963103e-05,
      "eval_rewards/rejected": 0.00012856203829869628,
      "eval_runtime": 146.0358,
      "eval_samples_per_second": 2.349,
      "eval_steps_per_second": 0.294,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.4982572012636904e-07,
      "logits/chosen": -2.005403518676758,
      "logits/rejected": -2.0029940605163574,
      "logps/chosen": -33.23992156982422,
      "logps/rejected": -34.01720428466797,
      "loss": 0.6932,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": 3.894489418598823e-05,
      "rewards/margins": -8.27050898806192e-05,
      "rewards/rejected": 0.00012164998042862862,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777677e-07,
      "logits/chosen": -2.0166852474212646,
      "logits/rejected": -2.008302688598633,
      "logps/chosen": -32.434364318847656,
      "logps/rejected": -32.170448303222656,
      "loss": 0.6931,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.00010542555537540466,
      "rewards/margins": 0.00019486629753373563,
      "rewards/rejected": -8.94407567102462e-05,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.194082707715275e-07,
      "logits/chosen": -2.0464212894439697,
      "logits/rejected": -2.0383687019348145,
      "logps/chosen": -30.498855590820312,
      "logps/rejected": -32.038177490234375,
      "loss": 0.6933,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": -0.0002313825534656644,
      "rewards/margins": -0.0003796928212977946,
      "rewards/rejected": 0.00014831028238404542,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.020402418666621e-07,
      "logits/chosen": -1.9772886037826538,
      "logits/rejected": -1.987566590309143,
      "logps/chosen": -31.40262794494629,
      "logps/rejected": -32.533485412597656,
      "loss": 0.6931,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.00017036919598467648,
      "rewards/margins": 0.00015645996609237045,
      "rewards/rejected": 1.3909203516959678e-05,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.8341962650351185e-07,
      "logits/chosen": -1.8902900218963623,
      "logits/rejected": -1.8913724422454834,
      "logps/chosen": -34.174896240234375,
      "logps/rejected": -34.764488220214844,
      "loss": 0.693,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.00017796576139517128,
      "rewards/margins": 0.00025028776144608855,
      "rewards/rejected": -7.232203643070534e-05,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800572e-07,
      "logits/chosen": -1.9422016143798828,
      "logits/rejected": -1.938711404800415,
      "logps/chosen": -36.14549255371094,
      "logps/rejected": -32.743194580078125,
      "loss": 0.6929,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.00026183511363342404,
      "rewards/margins": 0.0004662507271859795,
      "rewards/rejected": -0.00020441561355255544,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.430433172111807e-07,
      "logits/chosen": -2.041975498199463,
      "logits/rejected": -2.034566879272461,
      "logps/chosen": -33.78548049926758,
      "logps/rejected": -31.344890594482422,
      "loss": 0.6931,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0001237928809132427,
      "rewards/margins": 5.422577305580489e-05,
      "rewards/rejected": 6.956708239158615e-05,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.216202642830543e-07,
      "logits/chosen": -2.047132968902588,
      "logits/rejected": -2.0523972511291504,
      "logps/chosen": -32.526214599609375,
      "logps/rejected": -32.523216247558594,
      "loss": 0.6929,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.00018572784028947353,
      "rewards/margins": 0.000566480855923146,
      "rewards/rejected": -0.0003807530156336725,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.9960716642946403e-07,
      "logits/chosen": -2.048360824584961,
      "logits/rejected": -2.0455949306488037,
      "logps/chosen": -31.48598289489746,
      "logps/rejected": -31.33578109741211,
      "loss": 0.693,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 2.9682874810532667e-05,
      "rewards/margins": 0.00020535592921078205,
      "rewards/rejected": -0.00017567307804711163,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.771853789806683e-07,
      "logits/chosen": -1.918714165687561,
      "logits/rejected": -1.9233949184417725,
      "logps/chosen": -31.575008392333984,
      "logps/rejected": -32.78480911254883,
      "loss": 0.693,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.0003127219679299742,
      "rewards/margins": 0.00025148707209154963,
      "rewards/rejected": 6.123496132204309e-05,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.2438974380493164,
      "eval_logits/rejected": -2.2390170097351074,
      "eval_logps/chosen": -34.02705764770508,
      "eval_logps/rejected": -37.500091552734375,
      "eval_loss": 0.6931942701339722,
      "eval_rewards/accuracies": 0.4991694688796997,
      "eval_rewards/chosen": 7.496162288589403e-05,
      "eval_rewards/margins": -9.032993693836033e-05,
      "eval_rewards/rejected": 0.0001652915816521272,
      "eval_runtime": 145.6901,
      "eval_samples_per_second": 2.354,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402e-07,
      "logits/chosen": -2.0324437618255615,
      "logits/rejected": -2.0431275367736816,
      "logps/chosen": -31.958581924438477,
      "logps/rejected": -33.874977111816406,
      "loss": 0.693,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.00012724331463687122,
      "rewards/margins": 0.00025004256167449057,
      "rewards/rejected": -0.00012279924703761935,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.318564697655179e-07,
      "logits/chosen": -1.9262778759002686,
      "logits/rejected": -1.9411571025848389,
      "logps/chosen": -30.077739715576172,
      "logps/rejected": -31.5812931060791,
      "loss": 0.6929,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.0003380063863005489,
      "rewards/margins": 0.0005691683036275208,
      "rewards/rejected": -0.00023116180091165006,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.093227910899832e-07,
      "logits/chosen": -1.984165906906128,
      "logits/rejected": -1.9881302118301392,
      "logps/chosen": -33.39875030517578,
      "logps/rejected": -31.574642181396484,
      "loss": 0.6929,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.00018722013919614255,
      "rewards/margins": 0.0004406363586895168,
      "rewards/rejected": -0.00025341621949337423,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279356e-07,
      "logits/chosen": -1.9830011129379272,
      "logits/rejected": -1.9610159397125244,
      "logps/chosen": -34.161964416503906,
      "logps/rejected": -34.9656867980957,
      "loss": 0.6931,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -8.213297405745834e-05,
      "rewards/margins": 6.679827492916957e-05,
      "rewards/rejected": -0.0001489312417106703,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.654436768970182e-07,
      "logits/chosen": -2.024148464202881,
      "logits/rejected": -2.020843982696533,
      "logps/chosen": -32.924686431884766,
      "logps/rejected": -36.229591369628906,
      "loss": 0.6931,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 3.8046332520025317e-06,
      "rewards/margins": 5.0471979193389416e-05,
      "rewards/rejected": -4.666733366320841e-05,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.444597403062196e-07,
      "logits/chosen": -1.8910725116729736,
      "logits/rejected": -1.8886131048202515,
      "logps/chosen": -34.17741012573242,
      "logps/rejected": -35.50103759765625,
      "loss": 0.6931,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.00010990710143232718,
      "rewards/margins": 7.404615462291986e-05,
      "rewards/rejected": 3.586096136132255e-05,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.2434529917578887e-07,
      "logits/chosen": -1.8763153553009033,
      "logits/rejected": -1.873784065246582,
      "logps/chosen": -34.39301300048828,
      "logps/rejected": -31.753448486328125,
      "loss": 0.6932,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 4.5871554902987555e-05,
      "rewards/margins": -0.00013292724906932563,
      "rewards/rejected": 0.00017879880033433437,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603521e-07,
      "logits/chosen": -1.9792474508285522,
      "logits/rejected": -1.9685996770858765,
      "logps/chosen": -35.32111740112305,
      "logps/rejected": -31.84047508239746,
      "loss": 0.693,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.00035524697159416974,
      "rewards/margins": 0.0003500745224300772,
      "rewards/rejected": 5.172446435608435e-06,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071453e-08,
      "logits/chosen": -2.0755350589752197,
      "logits/rejected": -2.0605063438415527,
      "logps/chosen": -30.92108154296875,
      "logps/rejected": -32.65219497680664,
      "loss": 0.6931,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": 3.043042670469731e-05,
      "rewards/margins": 3.4023898479063064e-05,
      "rewards/rejected": -3.593444034777349e-06,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-08,
      "logits/chosen": -1.9464104175567627,
      "logits/rejected": -1.9438741207122803,
      "logps/chosen": -32.899200439453125,
      "logps/rejected": -30.818578720092773,
      "loss": 0.693,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.00027910774224437773,
      "rewards/margins": 0.00029683284810744226,
      "rewards/rejected": -1.772511677700095e-05,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.244286060333252,
      "eval_logits/rejected": -2.2394051551818848,
      "eval_logps/chosen": -34.025508880615234,
      "eval_logps/rejected": -37.50877380371094,
      "eval_loss": 0.6931424140930176,
      "eval_rewards/accuracies": 0.4958471655845642,
      "eval_rewards/chosen": 9.041713929036632e-05,
      "eval_rewards/margins": 1.1897234799107537e-05,
      "eval_rewards/rejected": 7.851990085327998e-05,
      "eval_runtime": 145.6886,
      "eval_samples_per_second": 2.354,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589034e-08,
      "logits/chosen": -1.9285697937011719,
      "logits/rejected": -1.9253038167953491,
      "logps/chosen": -31.567317962646484,
      "logps/rejected": -33.728389739990234,
      "loss": 0.693,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.00036028382601216435,
      "rewards/margins": 0.00022268423344939947,
      "rewards/rejected": 0.00013759959256276488,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380912e-08,
      "logits/chosen": -1.9806677103042603,
      "logits/rejected": -1.9683544635772705,
      "logps/chosen": -34.57938766479492,
      "logps/rejected": -33.576698303222656,
      "loss": 0.6929,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.0001669025223236531,
      "rewards/margins": 0.0005454975180327892,
      "rewards/rejected": -0.0003785950248129666,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-08,
      "logits/chosen": -2.0156922340393066,
      "logits/rejected": -2.014228105545044,
      "logps/chosen": -33.48464584350586,
      "logps/rejected": -32.47287368774414,
      "loss": 0.6932,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 1.1468652701296378e-05,
      "rewards/margins": -5.562617297982797e-05,
      "rewards/rejected": 6.709482113365084e-05,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.05793773749158e-08,
      "logits/chosen": -2.1031441688537598,
      "logits/rejected": -2.0873398780822754,
      "logps/chosen": -34.18089294433594,
      "logps/rejected": -33.10489273071289,
      "loss": 0.6932,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": -1.5041228834888898e-05,
      "rewards/margins": -0.000146475198562257,
      "rewards/rejected": 0.0001314339751843363,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.251801807404168e-08,
      "logits/chosen": -1.9747949838638306,
      "logits/rejected": -1.9738445281982422,
      "logps/chosen": -33.250999450683594,
      "logps/rejected": -32.463138580322266,
      "loss": 0.693,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.0003109553945250809,
      "rewards/margins": 0.00029085634741932154,
      "rewards/rejected": 2.009904164879117e-05,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-09,
      "logits/chosen": -1.9307000637054443,
      "logits/rejected": -1.9410593509674072,
      "logps/chosen": -32.20253372192383,
      "logps/rejected": -35.28581619262695,
      "loss": 0.6932,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 2.157145536330063e-05,
      "rewards/margins": -7.86231794336345e-06,
      "rewards/rejected": 2.9433791496558115e-05,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050324e-09,
      "logits/chosen": -2.069913148880005,
      "logits/rejected": -2.0633533000946045,
      "logps/chosen": -33.65105438232422,
      "logps/rejected": -29.224279403686523,
      "loss": 0.6932,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 9.127380508289207e-06,
      "rewards/margins": -1.2333152881183196e-05,
      "rewards/rejected": 2.1460538846440613e-05,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-10,
      "logits/chosen": -1.9291718006134033,
      "logits/rejected": -1.9313350915908813,
      "logps/chosen": -34.24486541748047,
      "logps/rejected": -30.897415161132812,
      "loss": 0.6931,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -5.8152021665591747e-05,
      "rewards/margins": 0.00018974675913341343,
      "rewards/rejected": -0.000247898802626878,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.1530164842481737,
      "train_runtime": 627.5816,
      "train_samples_per_second": 4.906,
      "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
}