{
  "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": -1.866068720817566,
      "logits/rejected": -1.8703863620758057,
      "logps/chosen": -36.99105453491211,
      "logps/rejected": -33.65992736816406,
      "loss": 0.9728,
      "rewards/accuracies": 0.5138888955116272,
      "rewards/chosen": 0.009325359016656876,
      "rewards/margins": 0.027240369468927383,
      "rewards/rejected": -0.017915012314915657,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.5641025641025636e-07,
      "logits/chosen": -1.9970887899398804,
      "logits/rejected": -1.9997413158416748,
      "logps/chosen": -29.63595199584961,
      "logps/rejected": -29.070571899414062,
      "loss": 0.992,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.003744622226804495,
      "rewards/margins": 0.008021654561161995,
      "rewards/rejected": -0.004277031868696213,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.8461538461538463e-07,
      "logits/chosen": -1.9195963144302368,
      "logits/rejected": -1.9168994426727295,
      "logps/chosen": -31.414175033569336,
      "logps/rejected": -33.25041961669922,
      "loss": 0.9797,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.0011531396303325891,
      "rewards/margins": 0.02032935619354248,
      "rewards/rejected": -0.019176214933395386,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438433e-07,
      "logits/chosen": -2.01688814163208,
      "logits/rejected": -2.0081212520599365,
      "logps/chosen": -32.583614349365234,
      "logps/rejected": -32.48966979980469,
      "loss": 1.0179,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.004085154738277197,
      "rewards/margins": -0.017886672168970108,
      "rewards/rejected": 0.013801517896354198,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542186e-07,
      "logits/chosen": -1.8644500970840454,
      "logits/rejected": -1.853655219078064,
      "logps/chosen": -33.54804229736328,
      "logps/rejected": -35.43767166137695,
      "loss": 1.0031,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 0.005664288066327572,
      "rewards/margins": -0.0031387731432914734,
      "rewards/rejected": 0.00880306214094162,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941118e-07,
      "logits/chosen": -1.9449504613876343,
      "logits/rejected": -1.946890115737915,
      "logps/chosen": -32.58049774169922,
      "logps/rejected": -33.199378967285156,
      "loss": 0.9714,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.011811850592494011,
      "rewards/margins": 0.02860717847943306,
      "rewards/rejected": -0.016795331612229347,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413548e-07,
      "logits/chosen": -2.0795464515686035,
      "logits/rejected": -2.0845184326171875,
      "logps/chosen": -33.98798370361328,
      "logps/rejected": -36.569732666015625,
      "loss": 0.9992,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.0003018675488419831,
      "rewards/margins": 0.0008206713828258216,
      "rewards/rejected": -0.0005188033101148903,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-07,
      "logits/chosen": -1.9425585269927979,
      "logits/rejected": -1.945704698562622,
      "logps/chosen": -34.408721923828125,
      "logps/rejected": -34.581504821777344,
      "loss": 0.9951,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.011452903971076012,
      "rewards/margins": 0.0049109989777207375,
      "rewards/rejected": 0.006541903130710125,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.736716601303429e-07,
      "logits/chosen": -1.9505088329315186,
      "logits/rejected": -1.9550243616104126,
      "logps/chosen": -32.451255798339844,
      "logps/rejected": -32.366790771484375,
      "loss": 0.9856,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.009697502478957176,
      "rewards/margins": 0.014425436034798622,
      "rewards/rejected": -0.0047279344871640205,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.62624545834521e-07,
      "logits/chosen": -2.048947334289551,
      "logits/rejected": -2.046935796737671,
      "logps/chosen": -32.25363540649414,
      "logps/rejected": -31.277917861938477,
      "loss": 1.0007,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.007680465932935476,
      "rewards/margins": -0.0006578800384886563,
      "rewards/rejected": -0.007022587116807699,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.243957042694092,
      "eval_logits/rejected": -2.2390689849853516,
      "eval_logps/chosen": -34.019203186035156,
      "eval_logps/rejected": -37.50347900390625,
      "eval_loss": 0.998858630657196,
      "eval_rewards/accuracies": 0.5402824282646179,
      "eval_rewards/chosen": 0.009209612384438515,
      "eval_rewards/margins": 0.0013267018366605043,
      "eval_rewards/rejected": 0.00788290984928608,
      "eval_runtime": 145.9923,
      "eval_samples_per_second": 2.349,
      "eval_steps_per_second": 0.295,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.4982572012636904e-07,
      "logits/chosen": -2.005302667617798,
      "logits/rejected": -2.002894878387451,
      "logps/chosen": -33.23944854736328,
      "logps/rejected": -34.005104064941406,
      "loss": 1.0119,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 0.002622986678034067,
      "rewards/margins": -0.01194014959037304,
      "rewards/rejected": 0.014563137665390968,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777677e-07,
      "logits/chosen": -2.0167059898376465,
      "logits/rejected": -2.0083398818969727,
      "logps/chosen": -32.45713424682617,
      "logps/rejected": -32.17597198486328,
      "loss": 0.9987,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.0073361145332455635,
      "rewards/margins": 0.0013455990701913834,
      "rewards/rejected": -0.008681714534759521,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.194082707715275e-07,
      "logits/chosen": -2.0464558601379395,
      "logits/rejected": -2.038412094116211,
      "logps/chosen": -30.482311248779297,
      "logps/rejected": -32.04933166503906,
      "loss": 1.0062,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.003956255037337542,
      "rewards/margins": -0.006162940990179777,
      "rewards/rejected": 0.002206685720011592,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.020402418666621e-07,
      "logits/chosen": -1.9769861698150635,
      "logits/rejected": -1.9872652292251587,
      "logps/chosen": -31.415592193603516,
      "logps/rejected": -32.5562858581543,
      "loss": 0.9847,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.0024404728319495916,
      "rewards/margins": 0.015285758301615715,
      "rewards/rejected": -0.01284528523683548,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.8341962650351185e-07,
      "logits/chosen": -1.890750527381897,
      "logits/rejected": -1.8918195962905884,
      "logps/chosen": -34.196632385253906,
      "logps/rejected": -34.74388885498047,
      "loss": 1.0104,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": -0.002364098560065031,
      "rewards/margins": -0.0103833619505167,
      "rewards/rejected": 0.008019264787435532,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800572e-07,
      "logits/chosen": -1.9424158334732056,
      "logits/rejected": -1.9389365911483765,
      "logps/chosen": -36.13062286376953,
      "logps/rejected": -32.720314025878906,
      "loss": 0.9768,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.02463117241859436,
      "rewards/margins": 0.023165332153439522,
      "rewards/rejected": 0.0014658428262919188,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.430433172111807e-07,
      "logits/chosen": -2.04191517829895,
      "logits/rejected": -2.034518003463745,
      "logps/chosen": -33.780067443847656,
      "logps/rejected": -31.35748863220215,
      "loss": 0.9859,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.010676576755940914,
      "rewards/margins": 0.014060018584132195,
      "rewards/rejected": -0.0033834422938525677,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.216202642830543e-07,
      "logits/chosen": -2.0474770069122314,
      "logits/rejected": -2.0527572631835938,
      "logps/chosen": -32.51685333251953,
      "logps/rejected": -32.48542022705078,
      "loss": 0.9831,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.01676076650619507,
      "rewards/margins": 0.016928378492593765,
      "rewards/rejected": -0.00016761067672632635,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.9960716642946403e-07,
      "logits/chosen": -2.0486884117126465,
      "logits/rejected": -2.045897960662842,
      "logps/chosen": -31.490026473999023,
      "logps/rejected": -31.336299896240234,
      "loss": 0.9898,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.0006444025784730911,
      "rewards/margins": 0.010206506587564945,
      "rewards/rejected": -0.010850909166038036,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.771853789806683e-07,
      "logits/chosen": -1.9185329675674438,
      "logits/rejected": -1.9231984615325928,
      "logps/chosen": -31.579875946044922,
      "logps/rejected": -32.781455993652344,
      "loss": 0.9898,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.015846019610762596,
      "rewards/margins": 0.010161412879824638,
      "rewards/rejected": 0.005684606730937958,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.2439582347869873,
      "eval_logits/rejected": -2.239088296890259,
      "eval_logps/chosen": -34.01380920410156,
      "eval_logps/rejected": -37.500911712646484,
      "eval_loss": 0.9969704151153564,
      "eval_rewards/accuracies": 0.5307309031486511,
      "eval_rewards/chosen": 0.01244510430842638,
      "eval_rewards/margins": 0.0030194728169590235,
      "eval_rewards/rejected": 0.009425631724298,
      "eval_runtime": 145.8589,
      "eval_samples_per_second": 2.352,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402e-07,
      "logits/chosen": -2.0318081378936768,
      "logits/rejected": -2.0424768924713135,
      "logps/chosen": -31.95905113220215,
      "logps/rejected": -33.87731170654297,
      "loss": 0.9839,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.007351105101406574,
      "rewards/margins": 0.01611880213022232,
      "rewards/rejected": -0.008767696097493172,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.318564697655179e-07,
      "logits/chosen": -1.9253685474395752,
      "logits/rejected": -1.940233826637268,
      "logps/chosen": -30.08770179748535,
      "logps/rejected": -31.577301025390625,
      "loss": 0.9742,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.014303619973361492,
      "rewards/margins": 0.025777745991945267,
      "rewards/rejected": -0.011474122293293476,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.093227910899832e-07,
      "logits/chosen": -1.9832963943481445,
      "logits/rejected": -1.9872655868530273,
      "logps/chosen": -33.39131546020508,
      "logps/rejected": -31.548086166381836,
      "loss": 0.985,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.015694385394454002,
      "rewards/margins": 0.014964587986469269,
      "rewards/rejected": 0.0007297966512851417,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279356e-07,
      "logits/chosen": -1.9828859567642212,
      "logits/rejected": -1.9609191417694092,
      "logps/chosen": -34.16459655761719,
      "logps/rejected": -34.95134735107422,
      "loss": 1.0062,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": -0.0065087564289569855,
      "rewards/margins": -0.006178082898259163,
      "rewards/rejected": -0.0003306727739982307,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.654436768970182e-07,
      "logits/chosen": -2.0246593952178955,
      "logits/rejected": -2.0213539600372314,
      "logps/chosen": -32.9222412109375,
      "logps/rejected": -36.228416442871094,
      "loss": 0.9962,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": 0.0016961356159299612,
      "rewards/margins": 0.0037924889475107193,
      "rewards/rejected": -0.0020963551942259073,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.444597403062196e-07,
      "logits/chosen": -1.8918907642364502,
      "logits/rejected": -1.8894577026367188,
      "logps/chosen": -34.189327239990234,
      "logps/rejected": -35.505496978759766,
      "loss": 1.0,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.0005569729837588966,
      "rewards/margins": -2.999706157424953e-05,
      "rewards/rejected": -0.0005269756657071412,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.2434529917578887e-07,
      "logits/chosen": -1.8758213520050049,
      "logits/rejected": -1.8732786178588867,
      "logps/chosen": -34.38149642944336,
      "logps/rejected": -31.752059936523438,
      "loss": 1.0019,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.009660609066486359,
      "rewards/margins": -0.0019017171580344439,
      "rewards/rejected": 0.011562327854335308,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603521e-07,
      "logits/chosen": -1.9798027276992798,
      "logits/rejected": -1.9691746234893799,
      "logps/chosen": -35.328853607177734,
      "logps/rejected": -31.868383407592773,
      "loss": 0.9669,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.016671547666192055,
      "rewards/margins": 0.03310702368617058,
      "rewards/rejected": -0.016435474157333374,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071453e-08,
      "logits/chosen": -2.075648546218872,
      "logits/rejected": -2.0605978965759277,
      "logps/chosen": -30.926809310913086,
      "logps/rejected": -32.621070861816406,
      "loss": 1.0201,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": -0.0016108205309137702,
      "rewards/margins": -0.020070534199476242,
      "rewards/rejected": 0.018459713086485863,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-08,
      "logits/chosen": -1.9466993808746338,
      "logits/rejected": -1.9441711902618408,
      "logps/chosen": -32.90770721435547,
      "logps/rejected": -30.82305335998535,
      "loss": 0.9846,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.011643724516034126,
      "rewards/margins": 0.015391023829579353,
      "rewards/rejected": -0.0037472969852387905,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.243670701980591,
      "eval_logits/rejected": -2.2387943267822266,
      "eval_logps/chosen": -34.02626419067383,
      "eval_logps/rejected": -37.50523376464844,
      "eval_loss": 1.001994013786316,
      "eval_rewards/accuracies": 0.4634551703929901,
      "eval_rewards/chosen": 0.0049731116741895676,
      "eval_rewards/margins": -0.0018590801628306508,
      "eval_rewards/rejected": 0.006832191254943609,
      "eval_runtime": 145.8336,
      "eval_samples_per_second": 2.352,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589034e-08,
      "logits/chosen": -1.9287983179092407,
      "logits/rejected": -1.9255239963531494,
      "logps/chosen": -31.573007583618164,
      "logps/rejected": -33.742637634277344,
      "loss": 0.9815,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.018206708133220673,
      "rewards/margins": 0.018496429547667503,
      "rewards/rejected": -0.0002897246740758419,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380912e-08,
      "logits/chosen": -1.9805524349212646,
      "logits/rejected": -1.9682422876358032,
      "logps/chosen": -34.57670211791992,
      "logps/rejected": -33.561363220214844,
      "loss": 0.9749,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.011624744161963463,
      "rewards/margins": 0.0251374039798975,
      "rewards/rejected": -0.013512656092643738,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-08,
      "logits/chosen": -2.016091823577881,
      "logits/rejected": -2.0146238803863525,
      "logps/chosen": -33.47222137451172,
      "logps/rejected": -32.46577835083008,
      "loss": 1.0001,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": 0.008143280632793903,
      "rewards/margins": -0.0001393534184899181,
      "rewards/rejected": 0.008282631635665894,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.05793773749158e-08,
      "logits/chosen": -2.1032156944274902,
      "logits/rejected": -2.0874197483062744,
      "logps/chosen": -34.174861907958984,
      "logps/rejected": -33.088890075683594,
      "loss": 1.0148,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.002716648392379284,
      "rewards/margins": -0.014771336689591408,
      "rewards/rejected": 0.017487986013293266,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.251801807404168e-08,
      "logits/chosen": -1.9747329950332642,
      "logits/rejected": -1.9737837314605713,
      "logps/chosen": -33.25817108154297,
      "logps/rejected": -32.459861755371094,
      "loss": 0.9888,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 0.014356844127178192,
      "rewards/margins": 0.01118617132306099,
      "rewards/rejected": 0.003170671407133341,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-09,
      "logits/chosen": -1.9303538799285889,
      "logits/rejected": -1.94071364402771,
      "logps/chosen": -32.20769500732422,
      "logps/rejected": -35.273780822753906,
      "loss": 1.0108,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.0018019669223576784,
      "rewards/margins": -0.010786842554807663,
      "rewards/rejected": 0.008984875865280628,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050324e-09,
      "logits/chosen": -2.069509744644165,
      "logits/rejected": -2.0629658699035645,
      "logps/chosen": -33.647117614746094,
      "logps/rejected": -29.217060089111328,
      "loss": 1.0027,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.002911838237196207,
      "rewards/margins": -0.0027081891894340515,
      "rewards/rejected": 0.005620025563985109,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-10,
      "logits/chosen": -1.929449439048767,
      "logits/rejected": -1.9315989017486572,
      "logps/chosen": -34.25099563598633,
      "logps/rejected": -30.906116485595703,
      "loss": 0.9871,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -0.007165629416704178,
      "rewards/margins": 0.012927901931107044,
      "rewards/rejected": -0.020093530416488647,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.9932638019710393,
      "train_runtime": 3254.1177,
      "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
}