{
  "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-07,
      "logits/chosen": -1.7278180122375488,
      "logits/rejected": -1.7377450466156006,
      "logps/chosen": -29.553977966308594,
      "logps/rejected": -42.813133239746094,
      "loss": 1.5625,
      "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": -1.8663330078125,
      "logits/rejected": -1.870657205581665,
      "logps/chosen": -36.98442077636719,
      "logps/rejected": -33.6639289855957,
      "loss": 1.4725,
      "rewards/accuracies": 0.5416666865348816,
      "rewards/chosen": 0.008871171623468399,
      "rewards/margins": 0.02241421490907669,
      "rewards/rejected": -0.013543044216930866,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.564102564102564e-06,
      "logits/chosen": -1.9977548122406006,
      "logits/rejected": -2.0003960132598877,
      "logps/chosen": -29.64242172241211,
      "logps/rejected": -29.04966163635254,
      "loss": 1.6323,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": -9.128078818321228e-05,
      "rewards/margins": -0.005604482255876064,
      "rewards/rejected": 0.005513200536370277,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": -1.9204576015472412,
      "logits/rejected": -1.917781114578247,
      "logps/chosen": -31.413188934326172,
      "logps/rejected": -33.22998046875,
      "loss": 1.5666,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.0011628343490883708,
      "rewards/margins": 0.005770112853497267,
      "rewards/rejected": -0.0046072788536548615,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438434e-06,
      "logits/chosen": -2.0172924995422363,
      "logits/rejected": -2.0085625648498535,
      "logps/chosen": -32.583003997802734,
      "logps/rejected": -32.50304412841797,
      "loss": 1.6493,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.0024785413406789303,
      "rewards/margins": -0.006329345051199198,
      "rewards/rejected": 0.00385080324485898,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542187e-06,
      "logits/chosen": -1.8622093200683594,
      "logits/rejected": -1.851438283920288,
      "logps/chosen": -33.5463981628418,
      "logps/rejected": -35.46432113647461,
      "loss": 1.569,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.004430866800248623,
      "rewards/margins": 0.009221290238201618,
      "rewards/rejected": -0.004790422972291708,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941119e-06,
      "logits/chosen": -1.939429521560669,
      "logits/rejected": -1.941382646560669,
      "logps/chosen": -32.560157775878906,
      "logps/rejected": -33.21000671386719,
      "loss": 1.5224,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.016012197360396385,
      "rewards/margins": 0.03146379441022873,
      "rewards/rejected": -0.015451604500412941,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413549e-06,
      "logits/chosen": -2.0719385147094727,
      "logits/rejected": -2.0768914222717285,
      "logps/chosen": -33.97148513793945,
      "logps/rejected": -36.61540985107422,
      "loss": 1.5627,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.006800562143325806,
      "rewards/margins": 0.025414401665329933,
      "rewards/rejected": -0.018613843247294426,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-06,
      "logits/chosen": -1.9341154098510742,
      "logits/rejected": -1.9372339248657227,
      "logps/chosen": -34.31813049316406,
      "logps/rejected": -34.64340591430664,
      "loss": 1.3398,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.043871380388736725,
      "rewards/margins": 0.06426960229873657,
      "rewards/rejected": -0.020398225635290146,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.7367166013034295e-06,
      "logits/chosen": -1.9430510997772217,
      "logits/rejected": -1.9475826025009155,
      "logps/chosen": -32.405487060546875,
      "logps/rejected": -32.3362922668457,
      "loss": 1.6048,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.02477400377392769,
      "rewards/margins": 0.01572720892727375,
      "rewards/rejected": 0.009046795777976513,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.626245458345211e-06,
      "logits/chosen": -2.040834426879883,
      "logits/rejected": -2.0388479232788086,
      "logps/chosen": -32.189510345458984,
      "logps/rejected": -31.297454833984375,
      "loss": 1.4583,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.020529210567474365,
      "rewards/margins": 0.0330255962908268,
      "rewards/rejected": -0.012496383860707283,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.2356765270233154,
      "eval_logits/rejected": -2.2308189868927,
      "eval_logps/chosen": -34.03032302856445,
      "eval_logps/rejected": -37.523555755615234,
      "eval_loss": 1.6161867380142212,
      "eval_rewards/accuracies": 0.5245016813278198,
      "eval_rewards/chosen": 0.0016898562898859382,
      "eval_rewards/margins": 0.004464337136596441,
      "eval_rewards/rejected": -0.0027744807302951813,
      "eval_runtime": 145.7904,
      "eval_samples_per_second": 2.353,
      "eval_steps_per_second": 0.295,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.498257201263691e-06,
      "logits/chosen": -1.996921181678772,
      "logits/rejected": -1.9945523738861084,
      "logps/chosen": -33.126094818115234,
      "logps/rejected": -34.01633071899414,
      "loss": 1.6379,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.04708842188119888,
      "rewards/margins": 0.041870661079883575,
      "rewards/rejected": 0.005217757076025009,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777678e-06,
      "logits/chosen": -2.0087766647338867,
      "logits/rejected": -2.0004258155822754,
      "logps/chosen": -32.360713958740234,
      "logps/rejected": -32.135719299316406,
      "loss": 1.5576,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.03367554023861885,
      "rewards/margins": 0.023361390456557274,
      "rewards/rejected": 0.010314146056771278,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.1940827077152755e-06,
      "logits/chosen": -2.036501407623291,
      "logits/rejected": -2.0285134315490723,
      "logps/chosen": -30.350208282470703,
      "logps/rejected": -32.078697204589844,
      "loss": 1.4709,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.05020326375961304,
      "rewards/margins": 0.06047834828495979,
      "rewards/rejected": -0.010275078006088734,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.0204024186666215e-06,
      "logits/chosen": -1.966408371925354,
      "logits/rejected": -1.9766807556152344,
      "logps/chosen": -31.250171661376953,
      "logps/rejected": -32.605186462402344,
      "loss": 1.2633,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.06779567152261734,
      "rewards/margins": 0.09591736644506454,
      "rewards/rejected": -0.028121691197156906,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.834196265035119e-06,
      "logits/chosen": -1.8785550594329834,
      "logits/rejected": -1.8797214031219482,
      "logps/chosen": -33.973148345947266,
      "logps/rejected": -34.82178497314453,
      "loss": 1.271,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.08781836926937103,
      "rewards/margins": 0.11363118886947632,
      "rewards/rejected": -0.02581283450126648,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800573e-06,
      "logits/chosen": -1.9306548833847046,
      "logits/rejected": -1.927215337753296,
      "logps/chosen": -36.05388259887695,
      "logps/rejected": -32.733726501464844,
      "loss": 1.3903,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.04711749404668808,
      "rewards/margins": 0.051506586372852325,
      "rewards/rejected": -0.004389096051454544,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4304331721118078e-06,
      "logits/chosen": -2.0320611000061035,
      "logits/rejected": -2.0246951580047607,
      "logps/chosen": -33.545772552490234,
      "logps/rejected": -31.430078506469727,
      "loss": 1.1164,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.10083547979593277,
      "rewards/margins": 0.13212835788726807,
      "rewards/rejected": -0.0312928780913353,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.2162026428305436e-06,
      "logits/chosen": -2.0383529663085938,
      "logits/rejected": -2.0436065196990967,
      "logps/chosen": -32.30766677856445,
      "logps/rejected": -32.46817398071289,
      "loss": 1.2603,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.09484858065843582,
      "rewards/margins": 0.08806191384792328,
      "rewards/rejected": 0.006786657031625509,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.996071664294641e-06,
      "logits/chosen": -2.0393786430358887,
      "logits/rejected": -2.036611795425415,
      "logps/chosen": -31.301233291625977,
      "logps/rejected": -31.35910415649414,
      "loss": 1.2965,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.0750882551074028,
      "rewards/margins": 0.09144291281700134,
      "rewards/rejected": -0.016354653984308243,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.7718537898066833e-06,
      "logits/chosen": -1.9094291925430298,
      "logits/rejected": -1.9140870571136475,
      "logps/chosen": -31.389511108398438,
      "logps/rejected": -32.832645416259766,
      "loss": 1.279,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.08670923858880997,
      "rewards/margins": 0.10339467227458954,
      "rewards/rejected": -0.01668544113636017,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.2359395027160645,
      "eval_logits/rejected": -2.231079339981079,
      "eval_logps/chosen": -34.08246994018555,
      "eval_logps/rejected": -37.59228515625,
      "eval_loss": 1.6099865436553955,
      "eval_rewards/accuracies": 0.525747537612915,
      "eval_rewards/chosen": -0.019167287275195122,
      "eval_rewards/margins": 0.011099184863269329,
      "eval_rewards/rejected": -0.030266473069787025,
      "eval_runtime": 145.6348,
      "eval_samples_per_second": 2.355,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402006e-06,
      "logits/chosen": -2.021488666534424,
      "logits/rejected": -2.0321521759033203,
      "logps/chosen": -31.773427963256836,
      "logps/rejected": -33.9339485168457,
      "loss": 1.1843,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.07915131002664566,
      "rewards/margins": 0.107650026679039,
      "rewards/rejected": -0.02849872037768364,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.3185646976551794e-06,
      "logits/chosen": -1.9136203527450562,
      "logits/rejected": -1.9284076690673828,
      "logps/chosen": -29.9388370513916,
      "logps/rejected": -31.61223793029785,
      "loss": 1.2413,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.06908205896615982,
      "rewards/margins": 0.09070716798305511,
      "rewards/rejected": -0.021625112742185593,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.0932279108998323e-06,
      "logits/chosen": -1.9700790643692017,
      "logits/rejected": -1.9740597009658813,
      "logps/chosen": -33.1496467590332,
      "logps/rejected": -31.65804100036621,
      "loss": 1.1206,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.10712933540344238,
      "rewards/margins": 0.15062502026557922,
      "rewards/rejected": -0.043495677411556244,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279358e-06,
      "logits/chosen": -1.9689995050430298,
      "logits/rejected": -1.9470970630645752,
      "logps/chosen": -33.9182243347168,
      "logps/rejected": -35.09490966796875,
      "loss": 1.073,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.09421048313379288,
      "rewards/margins": 0.15185722708702087,
      "rewards/rejected": -0.0576467402279377,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.6544367689701824e-06,
      "logits/chosen": -2.010103702545166,
      "logits/rejected": -2.006786346435547,
      "logps/chosen": -32.75348663330078,
      "logps/rejected": -36.29840087890625,
      "loss": 1.2351,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.0686345100402832,
      "rewards/margins": 0.09802549332380295,
      "rewards/rejected": -0.029390990734100342,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.4445974030621963e-06,
      "logits/chosen": -1.877804160118103,
      "logits/rejected": -1.8753639459609985,
      "logps/chosen": -34.015769958496094,
      "logps/rejected": -35.56401824951172,
      "loss": 1.2236,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.06905192881822586,
      "rewards/margins": 0.09281176328659058,
      "rewards/rejected": -0.023759832605719566,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.243452991757889e-06,
      "logits/chosen": -1.8633472919464111,
      "logits/rejected": -1.860896348953247,
      "logps/chosen": -34.244041442871094,
      "logps/rejected": -31.82059669494629,
      "loss": 1.2837,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.06142274662852287,
      "rewards/margins": 0.08112844079732895,
      "rewards/rejected": -0.019705694168806076,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603523e-06,
      "logits/chosen": -1.9670900106430054,
      "logits/rejected": -1.9565184116363525,
      "logps/chosen": -35.00214767456055,
      "logps/rejected": -31.88429832458496,
      "loss": 0.9577,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.14179818332195282,
      "rewards/margins": 0.15912050008773804,
      "rewards/rejected": -0.017322316765785217,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071455e-07,
      "logits/chosen": -2.0620455741882324,
      "logits/rejected": -2.0470833778381348,
      "logps/chosen": -30.75039291381836,
      "logps/rejected": -32.6963996887207,
      "loss": 1.3545,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.06949276477098465,
      "rewards/margins": 0.08731748908758163,
      "rewards/rejected": -0.017824724316596985,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-07,
      "logits/chosen": -1.933719277381897,
      "logits/rejected": -1.931168556213379,
      "logps/chosen": -32.519874572753906,
      "logps/rejected": -30.880207061767578,
      "loss": 1.0043,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.1628943681716919,
      "rewards/margins": 0.1882532835006714,
      "rewards/rejected": -0.02535891905426979,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.2311501502990723,
      "eval_logits/rejected": -2.226292610168457,
      "eval_logps/chosen": -34.0844841003418,
      "eval_logps/rejected": -37.60036087036133,
      "eval_loss": 1.5962141752243042,
      "eval_rewards/accuracies": 0.5166113376617432,
      "eval_rewards/chosen": -0.019972510635852814,
      "eval_rewards/margins": 0.013524654321372509,
      "eval_rewards/rejected": -0.0334971658885479,
      "eval_runtime": 145.5939,
      "eval_samples_per_second": 2.356,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589035e-07,
      "logits/chosen": -1.9168885946273804,
      "logits/rejected": -1.9136085510253906,
      "logps/chosen": -31.36187744140625,
      "logps/rejected": -33.820152282714844,
      "loss": 1.1726,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.09658866375684738,
      "rewards/margins": 0.127788707613945,
      "rewards/rejected": -0.03120005689561367,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380913e-07,
      "logits/chosen": -1.9674562215805054,
      "logits/rejected": -1.955171823501587,
      "logps/chosen": -34.4316291809082,
      "logps/rejected": -33.64500045776367,
      "loss": 1.1609,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.06577816605567932,
      "rewards/margins": 0.10824279487133026,
      "rewards/rejected": -0.04246463626623154,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-07,
      "logits/chosen": -2.002774238586426,
      "logits/rejected": -2.0013155937194824,
      "logps/chosen": -33.31989288330078,
      "logps/rejected": -32.59081268310547,
      "loss": 1.119,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.06635875999927521,
      "rewards/margins": 0.1108517199754715,
      "rewards/rejected": -0.04449295997619629,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.0579377374915805e-07,
      "logits/chosen": -2.089189052581787,
      "logits/rejected": -2.0734314918518066,
      "logps/chosen": -33.87842559814453,
      "logps/rejected": -33.1202278137207,
      "loss": 1.1504,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.12038698047399521,
      "rewards/margins": 0.12126438319683075,
      "rewards/rejected": -0.0008773911977186799,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.2518018074041684e-07,
      "logits/chosen": -1.9613533020019531,
      "logits/rejected": -1.9604451656341553,
      "logps/chosen": -32.962162017822266,
      "logps/rejected": -32.54551315307617,
      "loss": 1.1702,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.12797386944293976,
      "rewards/margins": 0.16012001037597656,
      "rewards/rejected": -0.032146163284778595,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-08,
      "logits/chosen": -1.9175068140029907,
      "logits/rejected": -1.9278007745742798,
      "logps/chosen": -31.958675384521484,
      "logps/rejected": -35.354461669921875,
      "loss": 1.1577,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.09840553998947144,
      "rewards/margins": 0.12468685954809189,
      "rewards/rejected": -0.026281306520104408,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050325e-08,
      "logits/chosen": -2.056452751159668,
      "logits/rejected": -2.0499649047851562,
      "logps/chosen": -33.426971435546875,
      "logps/rejected": -29.26825523376465,
      "loss": 1.1458,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.0899999737739563,
      "rewards/margins": 0.10673253238201141,
      "rewards/rejected": -0.01673254929482937,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-09,
      "logits/chosen": -1.916009545326233,
      "logits/rejected": -1.9182236194610596,
      "logps/chosen": -33.94841384887695,
      "logps/rejected": -30.931049346923828,
      "loss": 1.1073,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.11625485122203827,
      "rewards/margins": 0.13962247967720032,
      "rewards/rejected": -0.02336762472987175,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 1.3100324989913346,
      "train_runtime": 3250.9757,
      "train_samples_per_second": 0.947,
      "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
}