{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.999345977763244,
  "eval_steps": 100,
  "global_step": 382,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.282051282051282e-07,
      "logits/chosen": -2.3458919525146484,
      "logits/rejected": -2.3298444747924805,
      "logps/chosen": -24.293487548828125,
      "logps/rejected": -56.14665603637695,
      "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": 1.282051282051282e-06,
      "logits/chosen": -2.5303263664245605,
      "logits/rejected": -2.469114065170288,
      "logps/chosen": -87.91757202148438,
      "logps/rejected": -76.600830078125,
      "loss": 0.6931,
      "rewards/accuracies": 0.2361111044883728,
      "rewards/chosen": 5.832286842633039e-05,
      "rewards/margins": -0.00032225216273218393,
      "rewards/rejected": 0.00038057498750276864,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.564102564102564e-06,
      "logits/chosen": -2.51778507232666,
      "logits/rejected": -2.4767203330993652,
      "logps/chosen": -82.92424011230469,
      "logps/rejected": -82.9404067993164,
      "loss": 0.6929,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": 0.001368564902804792,
      "rewards/margins": 0.0007934757741168141,
      "rewards/rejected": 0.0005750892451032996,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": -2.524854898452759,
      "logits/rejected": -2.492816925048828,
      "logps/chosen": -101.72441864013672,
      "logps/rejected": -104.42093658447266,
      "loss": 0.6928,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": 0.0047247810289263725,
      "rewards/margins": 0.0011784282978624105,
      "rewards/rejected": 0.003546352731063962,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999895137916554e-06,
      "logits/chosen": -2.5693459510803223,
      "logits/rejected": -2.5531678199768066,
      "logps/chosen": -65.36104583740234,
      "logps/rejected": -71.78102111816406,
      "loss": 0.6926,
      "rewards/accuracies": 0.32499998807907104,
      "rewards/chosen": 0.006629495415836573,
      "rewards/margins": 0.0005699460161849856,
      "rewards/rejected": 0.006059548817574978,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987322328603843e-06,
      "logits/chosen": -2.5499138832092285,
      "logits/rejected": -2.4812991619110107,
      "logps/chosen": -85.08619689941406,
      "logps/rejected": -91.72368621826172,
      "loss": 0.6921,
      "rewards/accuracies": 0.3375000059604645,
      "rewards/chosen": 0.011005833745002747,
      "rewards/margins": 0.00240146997384727,
      "rewards/rejected": 0.008604364469647408,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.953897892477664e-06,
      "logits/chosen": -2.3451428413391113,
      "logits/rejected": -2.380490303039551,
      "logps/chosen": -70.91415405273438,
      "logps/rejected": -81.67916870117188,
      "loss": 0.6917,
      "rewards/accuracies": 0.38749998807907104,
      "rewards/chosen": 0.015571129508316517,
      "rewards/margins": 0.0033102519810199738,
      "rewards/rejected": 0.012260876595973969,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.899902032011388e-06,
      "logits/chosen": -2.5166893005371094,
      "logits/rejected": -2.5204968452453613,
      "logps/chosen": -70.66954040527344,
      "logps/rejected": -66.42029571533203,
      "loss": 0.6907,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": 0.01822517439723015,
      "rewards/margins": 0.006782916374504566,
      "rewards/rejected": 0.011442260816693306,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.825787403267713e-06,
      "logits/chosen": -2.2986605167388916,
      "logits/rejected": -2.2916884422302246,
      "logps/chosen": -55.5330924987793,
      "logps/rejected": -64.45872497558594,
      "loss": 0.6899,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.020927606150507927,
      "rewards/margins": 0.007273596711456776,
      "rewards/rejected": 0.013654010370373726,
      "step": 80
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.732175321209399e-06,
      "logits/chosen": -2.451927900314331,
      "logits/rejected": -2.4550557136535645,
      "logps/chosen": -83.72239685058594,
      "logps/rejected": -77.17060852050781,
      "loss": 0.6892,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.028661560267210007,
      "rewards/margins": 0.011586638167500496,
      "rewards/rejected": 0.01707492023706436,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.619850551115105e-06,
      "logits/chosen": -2.462538003921509,
      "logits/rejected": -2.384904146194458,
      "logps/chosen": -45.45023727416992,
      "logps/rejected": -49.149192810058594,
      "loss": 0.69,
      "rewards/accuracies": 0.2750000059604645,
      "rewards/chosen": 0.016560647636651993,
      "rewards/margins": 0.0038987069856375456,
      "rewards/rejected": 0.012661941349506378,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.4911162853240967,
      "eval_logits/rejected": -2.4903504848480225,
      "eval_logps/chosen": -63.03419876098633,
      "eval_logps/rejected": -67.10006713867188,
      "eval_loss": 0.6897407174110413,
      "eval_rewards/accuracies": 0.2680000066757202,
      "eval_rewards/chosen": 0.023249642923474312,
      "eval_rewards/margins": 0.0064011020585894585,
      "eval_rewards/rejected": 0.016848541796207428,
      "eval_runtime": 376.1511,
      "eval_samples_per_second": 5.317,
      "eval_steps_per_second": 0.665,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.489754729764663e-06,
      "logits/chosen": -2.5341131687164307,
      "logits/rejected": -2.5058047771453857,
      "logps/chosen": -82.01190185546875,
      "logps/rejected": -82.14210510253906,
      "loss": 0.6892,
      "rewards/accuracies": 0.32499998807907104,
      "rewards/chosen": 0.026387155055999756,
      "rewards/margins": 0.006781895644962788,
      "rewards/rejected": 0.019605260342359543,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.342978471545118e-06,
      "logits/chosen": -2.406015157699585,
      "logits/rejected": -2.4332752227783203,
      "logps/chosen": -59.783790588378906,
      "logps/rejected": -47.791358947753906,
      "loss": 0.688,
      "rewards/accuracies": 0.3375000059604645,
      "rewards/chosen": 0.023528261110186577,
      "rewards/margins": 0.01194488350301981,
      "rewards/rejected": 0.011583376675844193,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.1807522256532925e-06,
      "logits/chosen": -2.484022617340088,
      "logits/rejected": -2.4554779529571533,
      "logps/chosen": -71.26405334472656,
      "logps/rejected": -76.42189025878906,
      "loss": 0.6887,
      "rewards/accuracies": 0.32499998807907104,
      "rewards/chosen": 0.027204791083931923,
      "rewards/margins": 0.01390073262155056,
      "rewards/rejected": 0.013304059393703938,
      "step": 130
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.0044359610404505e-06,
      "logits/chosen": -2.4449095726013184,
      "logits/rejected": -2.513423442840576,
      "logps/chosen": -73.46162414550781,
      "logps/rejected": -68.35511016845703,
      "loss": 0.6873,
      "rewards/accuracies": 0.26249998807907104,
      "rewards/chosen": 0.017034200951457024,
      "rewards/margins": 0.007175432052463293,
      "rewards/rejected": 0.009858767502009869,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.815507765571832e-06,
      "logits/chosen": -2.3777384757995605,
      "logits/rejected": -2.3317112922668457,
      "logps/chosen": -48.66705322265625,
      "logps/rejected": -60.531578063964844,
      "loss": 0.6871,
      "rewards/accuracies": 0.3125,
      "rewards/chosen": 0.013008219189941883,
      "rewards/margins": 0.010229717940092087,
      "rewards/rejected": 0.00277850148268044,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.6155514549761644e-06,
      "logits/chosen": -2.4121792316436768,
      "logits/rejected": -2.4337995052337646,
      "logps/chosen": -61.59517288208008,
      "logps/rejected": -63.75123977661133,
      "loss": 0.6884,
      "rewards/accuracies": 0.3125,
      "rewards/chosen": 0.021260840818285942,
      "rewards/margins": 0.00982903502881527,
      "rewards/rejected": 0.011431803926825523,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4062432954613256e-06,
      "logits/chosen": -2.4510605335235596,
      "logits/rejected": -2.4552080631256104,
      "logps/chosen": -78.11065673828125,
      "logps/rejected": -71.26216888427734,
      "loss": 0.6839,
      "rewards/accuracies": 0.3375000059604645,
      "rewards/chosen": 0.024866515770554543,
      "rewards/margins": 0.017677070572972298,
      "rewards/rejected": 0.007189445197582245,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.189337951302639e-06,
      "logits/chosen": -2.557910203933716,
      "logits/rejected": -2.5174174308776855,
      "logps/chosen": -63.83809280395508,
      "logps/rejected": -68.5108871459961,
      "loss": 0.686,
      "rewards/accuracies": 0.32499998807907104,
      "rewards/chosen": 0.022966545075178146,
      "rewards/margins": 0.017569351941347122,
      "rewards/rejected": 0.005397191736847162,
      "step": 180
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.9666537752074443e-06,
      "logits/chosen": -2.4893686771392822,
      "logits/rejected": -2.4778811931610107,
      "logps/chosen": -86.08170318603516,
      "logps/rejected": -79.71577453613281,
      "loss": 0.6841,
      "rewards/accuracies": 0.3375000059604645,
      "rewards/chosen": 0.007741170935332775,
      "rewards/margins": 0.013728922232985497,
      "rewards/rejected": -0.005987753160297871,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.7400575647692046e-06,
      "logits/chosen": -2.307661294937134,
      "logits/rejected": -2.314356565475464,
      "logps/chosen": -34.22106170654297,
      "logps/rejected": -51.29539108276367,
      "loss": 0.6869,
      "rewards/accuracies": 0.26249998807907104,
      "rewards/chosen": 0.006807768251746893,
      "rewards/margins": 0.01217351108789444,
      "rewards/rejected": -0.005365744233131409,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.457301139831543,
      "eval_logits/rejected": -2.4556241035461426,
      "eval_logps/chosen": -64.6949691772461,
      "eval_logps/rejected": -69.70600128173828,
      "eval_loss": 0.6848968267440796,
      "eval_rewards/accuracies": 0.3059999942779541,
      "eval_rewards/chosen": 0.00664189737290144,
      "eval_rewards/margins": 0.015852706506848335,
      "eval_rewards/rejected": -0.009210809133946896,
      "eval_runtime": 376.2332,
      "eval_samples_per_second": 5.316,
      "eval_steps_per_second": 0.664,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5114489128002655e-06,
      "logits/chosen": -2.2540907859802246,
      "logits/rejected": -2.2756266593933105,
      "logps/chosen": -68.72586822509766,
      "logps/rejected": -80.93492126464844,
      "loss": 0.683,
      "rewards/accuracies": 0.2874999940395355,
      "rewards/chosen": 0.014960775151848793,
      "rewards/margins": 0.020435906946659088,
      "rewards/rejected": -0.005475130397826433,
      "step": 210
    },
    {
      "epoch": 0.58,
      "learning_rate": 2.2827442827369776e-06,
      "logits/chosen": -2.36279296875,
      "logits/rejected": -2.3906493186950684,
      "logps/chosen": -66.4981689453125,
      "logps/rejected": -72.55201721191406,
      "loss": 0.6845,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.009893990121781826,
      "rewards/margins": 0.030444588512182236,
      "rewards/rejected": -0.020550597459077835,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.0558609426156346e-06,
      "logits/chosen": -2.5613975524902344,
      "logits/rejected": -2.5647411346435547,
      "logps/chosen": -91.46321868896484,
      "logps/rejected": -82.19490051269531,
      "loss": 0.6826,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": 0.009320605546236038,
      "rewards/margins": 0.036185480654239655,
      "rewards/rejected": -0.026864871382713318,
      "step": 230
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8327008923033224e-06,
      "logits/chosen": -2.4174304008483887,
      "logits/rejected": -2.395275115966797,
      "logps/chosen": -75.15461730957031,
      "logps/rejected": -72.01311492919922,
      "loss": 0.6828,
      "rewards/accuracies": 0.25,
      "rewards/chosen": 0.005219348706305027,
      "rewards/margins": 0.016013307496905327,
      "rewards/rejected": -0.010793961584568024,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.6151349187243066e-06,
      "logits/chosen": -2.4290623664855957,
      "logits/rejected": -2.4516799449920654,
      "logps/chosen": -92.50470733642578,
      "logps/rejected": -91.04740905761719,
      "loss": 0.6824,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": -0.011667473241686821,
      "rewards/margins": 0.016532858833670616,
      "rewards/rejected": -0.028200332075357437,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.4049869127495886e-06,
      "logits/chosen": -2.419191837310791,
      "logits/rejected": -2.3789072036743164,
      "logps/chosen": -55.09886932373047,
      "logps/rejected": -63.9772834777832,
      "loss": 0.6842,
      "rewards/accuracies": 0.22499999403953552,
      "rewards/chosen": -0.0026586714666336775,
      "rewards/margins": 0.007516457699239254,
      "rewards/rejected": -0.010175129398703575,
      "step": 260
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.2040185792236874e-06,
      "logits/chosen": -2.366826295852661,
      "logits/rejected": -2.387421131134033,
      "logps/chosen": -78.35263061523438,
      "logps/rejected": -85.33952331542969,
      "loss": 0.6812,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 0.0012182873906567693,
      "rewards/margins": 0.039223961532115936,
      "rewards/rejected": -0.038005679845809937,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0139146683069728e-06,
      "logits/chosen": -2.507392168045044,
      "logits/rejected": -2.514265775680542,
      "logps/chosen": -80.45622253417969,
      "logps/rejected": -72.48971557617188,
      "loss": 0.6785,
      "rewards/accuracies": 0.30000001192092896,
      "rewards/chosen": 0.01629328541457653,
      "rewards/margins": 0.02651815116405487,
      "rewards/rejected": -0.01022486761212349,
      "step": 280
    },
    {
      "epoch": 0.76,
      "learning_rate": 8.362688519416084e-07,
      "logits/chosen": -2.448162078857422,
      "logits/rejected": -2.432882070541382,
      "logps/chosen": -69.45552062988281,
      "logps/rejected": -75.71241760253906,
      "loss": 0.6867,
      "rewards/accuracies": 0.2750000059604645,
      "rewards/chosen": 0.0061899758875370026,
      "rewards/margins": 0.012647658586502075,
      "rewards/rejected": -0.006457681301981211,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 6.725703638409861e-07,
      "logits/chosen": -2.4520912170410156,
      "logits/rejected": -2.4525935649871826,
      "logps/chosen": -73.40272521972656,
      "logps/rejected": -72.8621826171875,
      "loss": 0.681,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.0016829551896080375,
      "rewards/margins": 0.024337466806173325,
      "rewards/rejected": -0.022654514759778976,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.4445886611938477,
      "eval_logits/rejected": -2.4429824352264404,
      "eval_logps/chosen": -65.6224136352539,
      "eval_logps/rejected": -71.427978515625,
      "eval_loss": 0.6814776062965393,
      "eval_rewards/accuracies": 0.2879999876022339,
      "eval_rewards/chosen": -0.002632532035931945,
      "eval_rewards/margins": 0.02379809506237507,
      "eval_rewards/rejected": -0.026430627331137657,
      "eval_runtime": 376.1246,
      "eval_samples_per_second": 5.317,
      "eval_steps_per_second": 0.665,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.24191515001774e-07,
      "logits/chosen": -2.3931050300598145,
      "logits/rejected": -2.394754409790039,
      "logps/chosen": -76.21150970458984,
      "logps/rejected": -73.65968322753906,
      "loss": 0.6818,
      "rewards/accuracies": 0.2874999940395355,
      "rewards/chosen": 0.005356580018997192,
      "rewards/margins": 0.02418547496199608,
      "rewards/rejected": -0.018828898668289185,
      "step": 310
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.9237618939808655e-07,
      "logits/chosen": -2.3786847591400146,
      "logits/rejected": -2.3697614669799805,
      "logps/chosen": -68.56962585449219,
      "logps/rejected": -69.29893493652344,
      "loss": 0.6827,
      "rewards/accuracies": 0.2874999940395355,
      "rewards/chosen": -0.002341556828469038,
      "rewards/margins": 0.014613455161452293,
      "rewards/rejected": -0.01695501245558262,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 2.782294163002261e-07,
      "logits/chosen": -2.3469510078430176,
      "logits/rejected": -2.324633836746216,
      "logps/chosen": -69.15370178222656,
      "logps/rejected": -79.9259033203125,
      "loss": 0.679,
      "rewards/accuracies": 0.30000001192092896,
      "rewards/chosen": 0.008440351113677025,
      "rewards/margins": 0.024770593270659447,
      "rewards/rejected": -0.01633024401962757,
      "step": 330
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.8270810663494591e-07,
      "logits/chosen": -2.4450182914733887,
      "logits/rejected": -2.427203416824341,
      "logps/chosen": -67.0771484375,
      "logps/rejected": -68.71158599853516,
      "loss": 0.6797,
      "rewards/accuracies": 0.2750000059604645,
      "rewards/chosen": 0.0001767903595464304,
      "rewards/margins": 0.018586840480566025,
      "rewards/rejected": -0.018410051241517067,
      "step": 340
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.0661303104582882e-07,
      "logits/chosen": -2.5075066089630127,
      "logits/rejected": -2.515806198120117,
      "logps/chosen": -100.72679138183594,
      "logps/rejected": -111.1237564086914,
      "loss": 0.683,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": 0.004834707826375961,
      "rewards/margins": 0.03868640214204788,
      "rewards/rejected": -0.03385169431567192,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 5.0582106903004636e-08,
      "logits/chosen": -2.3713040351867676,
      "logits/rejected": -2.476989507675171,
      "logps/chosen": -80.11368560791016,
      "logps/rejected": -83.44255065917969,
      "loss": 0.6804,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": 0.0026325061917304993,
      "rewards/margins": 0.0385906919836998,
      "rewards/rejected": -0.0359581895172596,
      "step": 360
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.508505053829046e-08,
      "logits/chosen": -2.4487078189849854,
      "logits/rejected": -2.4094619750976562,
      "logps/chosen": -64.8487548828125,
      "logps/rejected": -94.2999496459961,
      "loss": 0.6805,
      "rewards/accuracies": 0.32499998807907104,
      "rewards/chosen": 0.00822712853550911,
      "rewards/margins": 0.03635001927614212,
      "rewards/rejected": -0.02812289260327816,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 4.19439536940569e-10,
      "logits/chosen": -2.4872801303863525,
      "logits/rejected": -2.508479595184326,
      "logps/chosen": -71.79349517822266,
      "logps/rejected": -80.17054748535156,
      "loss": 0.6814,
      "rewards/accuracies": 0.3375000059604645,
      "rewards/chosen": -0.00255067041143775,
      "rewards/margins": 0.025686126202344894,
      "rewards/rejected": -0.02823679707944393,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 382,
      "total_flos": 0.0,
      "train_loss": 0.6857879155593393,
      "train_runtime": 5551.1855,
      "train_samples_per_second": 2.203,
      "train_steps_per_second": 0.069
    }
  ],
  "logging_steps": 10,
  "max_steps": 382,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}