{
  "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": 0.6944,
      "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": -1.8664441108703613,
      "logits/rejected": -1.8707667589187622,
      "logps/chosen": -36.99871826171875,
      "logps/rejected": -33.66032028198242,
      "loss": 0.6698,
      "rewards/accuracies": 0.5138888955116272,
      "rewards/chosen": 0.0047301375307142735,
      "rewards/margins": 0.022878725081682205,
      "rewards/rejected": -0.018148591741919518,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.564102564102564e-06,
      "logits/chosen": -1.9972795248031616,
      "logits/rejected": -1.999922513961792,
      "logps/chosen": -29.64887046813965,
      "logps/rejected": -29.054330825805664,
      "loss": 0.756,
      "rewards/accuracies": 0.38749998807907104,
      "rewards/chosen": -0.0040070666000247,
      "rewards/margins": -0.009474089369177818,
      "rewards/rejected": 0.0054670218378305435,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": -1.9204769134521484,
      "logits/rejected": -1.9177768230438232,
      "logps/chosen": -31.404077529907227,
      "logps/rejected": -33.22195053100586,
      "loss": 0.7112,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.007212291471660137,
      "rewards/margins": 0.009304212406277657,
      "rewards/rejected": -0.0020919169764965773,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438434e-06,
      "logits/chosen": -2.0172927379608154,
      "logits/rejected": -2.008545398712158,
      "logps/chosen": -32.57653045654297,
      "logps/rejected": -32.520172119140625,
      "loss": 0.7334,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.00016416943981312215,
      "rewards/margins": 0.004665405489504337,
      "rewards/rejected": -0.004501235671341419,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542187e-06,
      "logits/chosen": -1.8620569705963135,
      "logits/rejected": -1.8512712717056274,
      "logps/chosen": -33.55234146118164,
      "logps/rejected": -35.454689025878906,
      "loss": 0.7458,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": 0.003082751529291272,
      "rewards/margins": 0.004490534774959087,
      "rewards/rejected": -0.0014077809173613787,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941119e-06,
      "logits/chosen": -1.940212607383728,
      "logits/rejected": -1.9421584606170654,
      "logps/chosen": -32.56269454956055,
      "logps/rejected": -33.21684265136719,
      "loss": 0.6785,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.022492902353405952,
      "rewards/margins": 0.04977109283208847,
      "rewards/rejected": -0.027278194203972816,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413549e-06,
      "logits/chosen": -2.0728747844696045,
      "logits/rejected": -2.0778398513793945,
      "logps/chosen": -33.98863983154297,
      "logps/rejected": -36.62779235839844,
      "loss": 0.7641,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -9.316801879322156e-05,
      "rewards/margins": 0.035260457545518875,
      "rewards/rejected": -0.035353630781173706,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-06,
      "logits/chosen": -1.935132384300232,
      "logits/rejected": -1.9382661581039429,
      "logps/chosen": -34.321903228759766,
      "logps/rejected": -34.638282775878906,
      "loss": 0.6095,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.06354494392871857,
      "rewards/margins": 0.09106875956058502,
      "rewards/rejected": -0.027523810043931007,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.7367166013034295e-06,
      "logits/chosen": -1.9444854259490967,
      "logits/rejected": -1.9490203857421875,
      "logps/chosen": -32.41173553466797,
      "logps/rejected": -32.337852478027344,
      "loss": 0.7499,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.033411428332328796,
      "rewards/margins": 0.020780622959136963,
      "rewards/rejected": 0.012630810961127281,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.626245458345211e-06,
      "logits/chosen": -2.042165994644165,
      "logits/rejected": -2.0401768684387207,
      "logps/chosen": -32.13690185546875,
      "logps/rejected": -31.2872314453125,
      "loss": 0.5903,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.06236031651496887,
      "rewards/margins": 0.0749734491109848,
      "rewards/rejected": -0.012613129802048206,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.2366247177124023,
      "eval_logits/rejected": -2.231769323348999,
      "eval_logps/chosen": -34.01411819458008,
      "eval_logps/rejected": -37.50873947143555,
      "eval_loss": 0.7449910640716553,
      "eval_rewards/accuracies": 0.5240863561630249,
      "eval_rewards/chosen": 0.01226096786558628,
      "eval_rewards/margins": 0.007533104624599218,
      "eval_rewards/rejected": 0.0047278618440032005,
      "eval_runtime": 145.783,
      "eval_samples_per_second": 2.353,
      "eval_steps_per_second": 0.295,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.498257201263691e-06,
      "logits/chosen": -1.9979403018951416,
      "logits/rejected": -1.9955459833145142,
      "logps/chosen": -33.12257766723633,
      "logps/rejected": -33.99717712402344,
      "loss": 0.8547,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.07274619489908218,
      "rewards/margins": 0.05342693254351616,
      "rewards/rejected": 0.019319258630275726,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777678e-06,
      "logits/chosen": -2.009084463119507,
      "logits/rejected": -2.0007264614105225,
      "logps/chosen": -32.33968734741211,
      "logps/rejected": -32.13513946533203,
      "loss": 0.7104,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.06312941759824753,
      "rewards/margins": 0.047309644520282745,
      "rewards/rejected": 0.015819773077964783,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.1940827077152755e-06,
      "logits/chosen": -2.0365753173828125,
      "logits/rejected": -2.028564453125,
      "logps/chosen": -30.363454818725586,
      "logps/rejected": -32.06770706176758,
      "loss": 0.71,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.06735612452030182,
      "rewards/margins": 0.07617705315351486,
      "rewards/rejected": -0.008820934221148491,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.0204024186666215e-06,
      "logits/chosen": -1.9657623767852783,
      "logits/rejected": -1.976011872291565,
      "logps/chosen": -31.213130950927734,
      "logps/rejected": -32.592655181884766,
      "loss": 0.6309,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.12391922622919083,
      "rewards/margins": 0.1585826873779297,
      "rewards/rejected": -0.03466347977519035,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.834196265035119e-06,
      "logits/chosen": -1.8778575658798218,
      "logits/rejected": -1.879020094871521,
      "logps/chosen": -33.998287200927734,
      "logps/rejected": -34.83546829223633,
      "loss": 0.566,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.11664185672998428,
      "rewards/margins": 0.1635705530643463,
      "rewards/rejected": -0.04692870378494263,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800573e-06,
      "logits/chosen": -1.9302040338516235,
      "logits/rejected": -1.9267680644989014,
      "logps/chosen": -36.0306510925293,
      "logps/rejected": -32.73199462890625,
      "loss": 0.5795,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.08461670577526093,
      "rewards/margins": 0.09016285836696625,
      "rewards/rejected": -0.005546141881495714,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4304331721118078e-06,
      "logits/chosen": -2.031188726425171,
      "logits/rejected": -2.0238261222839355,
      "logps/chosen": -33.55583953857422,
      "logps/rejected": -31.381237030029297,
      "loss": 0.5041,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.14521491527557373,
      "rewards/margins": 0.16284644603729248,
      "rewards/rejected": -0.017631549388170242,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.2162026428305436e-06,
      "logits/chosen": -2.037135362625122,
      "logits/rejected": -2.0424118041992188,
      "logps/chosen": -32.33055114746094,
      "logps/rejected": -32.500144958496094,
      "loss": 0.5257,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.12854143977165222,
      "rewards/margins": 0.13754449784755707,
      "rewards/rejected": -0.009003056213259697,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.996071664294641e-06,
      "logits/chosen": -2.0384888648986816,
      "logits/rejected": -2.0357367992401123,
      "logps/chosen": -31.326950073242188,
      "logps/rejected": -31.338672637939453,
      "loss": 0.6504,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.09720095247030258,
      "rewards/margins": 0.10947537422180176,
      "rewards/rejected": -0.012274429202079773,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.7718537898066833e-06,
      "logits/chosen": -1.908755898475647,
      "logits/rejected": -1.9134235382080078,
      "logps/chosen": -31.36275863647461,
      "logps/rejected": -32.82594680786133,
      "loss": 0.59,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.14611336588859558,
      "rewards/margins": 0.16712167859077454,
      "rewards/rejected": -0.02100830152630806,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.233200788497925,
      "eval_logits/rejected": -2.228351354598999,
      "eval_logps/chosen": -34.069149017333984,
      "eval_logps/rejected": -37.57521057128906,
      "eval_loss": 0.750046968460083,
      "eval_rewards/accuracies": 0.5365448594093323,
      "eval_rewards/chosen": -0.020758595317602158,
      "eval_rewards/margins": 0.014396170154213905,
      "eval_rewards/rejected": -0.035154763609170914,
      "eval_runtime": 145.6956,
      "eval_samples_per_second": 2.354,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402006e-06,
      "logits/chosen": -2.0207877159118652,
      "logits/rejected": -2.0314865112304688,
      "logps/chosen": -31.7946834564209,
      "logps/rejected": -33.94416046142578,
      "loss": 0.544,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.10597337782382965,
      "rewards/margins": 0.1548517644405365,
      "rewards/rejected": -0.04887838289141655,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.3185646976551794e-06,
      "logits/chosen": -1.9139292240142822,
      "logits/rejected": -1.9287306070327759,
      "logps/chosen": -29.91892433166504,
      "logps/rejected": -31.5737247467041,
      "loss": 0.5694,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.11556863784790039,
      "rewards/margins": 0.12489677965641022,
      "rewards/rejected": -0.00932813435792923,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.0932279108998323e-06,
      "logits/chosen": -1.9701712131500244,
      "logits/rejected": -1.974172592163086,
      "logps/chosen": -33.17791748046875,
      "logps/rejected": -31.603836059570312,
      "loss": 0.4933,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.1437317281961441,
      "rewards/margins": 0.1764504611492157,
      "rewards/rejected": -0.03271873667836189,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279358e-06,
      "logits/chosen": -1.9691722393035889,
      "logits/rejected": -1.9472358226776123,
      "logps/chosen": -33.92795181274414,
      "logps/rejected": -35.06544876098633,
      "loss": 0.4966,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.13547900319099426,
      "rewards/margins": 0.20427203178405762,
      "rewards/rejected": -0.06879300624132156,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.6544367689701824e-06,
      "logits/chosen": -2.010617971420288,
      "logits/rejected": -2.0072851181030273,
      "logps/chosen": -32.785011291503906,
      "logps/rejected": -36.283348083496094,
      "loss": 0.5823,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.08403309434652328,
      "rewards/margins": 0.11908762156963348,
      "rewards/rejected": -0.035054538398981094,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.4445974030621963e-06,
      "logits/chosen": -1.8787453174591064,
      "logits/rejected": -1.8763244152069092,
      "logps/chosen": -34.000633239746094,
      "logps/rejected": -35.53338623046875,
      "loss": 0.5277,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.11265683174133301,
      "rewards/margins": 0.1299169510602951,
      "rewards/rejected": -0.017260130494832993,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.243452991757889e-06,
      "logits/chosen": -1.864057183265686,
      "logits/rejected": -1.861581802368164,
      "logps/chosen": -34.23812484741211,
      "logps/rejected": -31.777530670166016,
      "loss": 0.6307,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.09568403661251068,
      "rewards/margins": 0.09940369427204132,
      "rewards/rejected": -0.003719669533893466,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603523e-06,
      "logits/chosen": -1.9680726528167725,
      "logits/rejected": -1.957493782043457,
      "logps/chosen": -35.05353546142578,
      "logps/rejected": -31.8691463470459,
      "loss": 0.4252,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.18186399340629578,
      "rewards/margins": 0.19875726103782654,
      "rewards/rejected": -0.01689324900507927,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071455e-07,
      "logits/chosen": -2.0636587142944336,
      "logits/rejected": -2.048659563064575,
      "logps/chosen": -30.76055908203125,
      "logps/rejected": -32.65589141845703,
      "loss": 0.6667,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.09814020991325378,
      "rewards/margins": 0.10057313740253448,
      "rewards/rejected": -0.0024329207371920347,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-07,
      "logits/chosen": -1.9355440139770508,
      "logits/rejected": -1.9329853057861328,
      "logps/chosen": -32.57345199584961,
      "logps/rejected": -30.86557388305664,
      "loss": 0.3871,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.21219508349895477,
      "rewards/margins": 0.24145305156707764,
      "rewards/rejected": -0.029257934540510178,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.233344554901123,
      "eval_logits/rejected": -2.2284841537475586,
      "eval_logps/chosen": -34.06221008300781,
      "eval_logps/rejected": -37.56574249267578,
      "eval_loss": 0.7655012607574463,
      "eval_rewards/accuracies": 0.5070598125457764,
      "eval_rewards/chosen": -0.016592998057603836,
      "eval_rewards/margins": 0.012880506925284863,
      "eval_rewards/rejected": -0.029473504051566124,
      "eval_runtime": 145.6769,
      "eval_samples_per_second": 2.355,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589035e-07,
      "logits/chosen": -1.9186370372772217,
      "logits/rejected": -1.9153635501861572,
      "logps/chosen": -31.363632202148438,
      "logps/rejected": -33.790061950683594,
      "loss": 0.5232,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.1438317447900772,
      "rewards/margins": 0.17257718741893768,
      "rewards/rejected": -0.02874544821679592,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380913e-07,
      "logits/chosen": -1.9695978164672852,
      "logits/rejected": -1.9573185443878174,
      "logps/chosen": -34.42280960083008,
      "logps/rejected": -33.63968276977539,
      "loss": 0.4864,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.103958860039711,
      "rewards/margins": 0.1644643247127533,
      "rewards/rejected": -0.06050548702478409,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-07,
      "logits/chosen": -2.0046210289001465,
      "logits/rejected": -2.003176689147949,
      "logps/chosen": -33.34724426269531,
      "logps/rejected": -32.53419876098633,
      "loss": 0.5611,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.08312927186489105,
      "rewards/margins": 0.11589743942022324,
      "rewards/rejected": -0.032768163830041885,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.0579377374915805e-07,
      "logits/chosen": -2.091379404067993,
      "logits/rejected": -2.0756208896636963,
      "logps/chosen": -33.90237808227539,
      "logps/rejected": -33.127288818359375,
      "loss": 0.4859,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.16620711982250214,
      "rewards/margins": 0.17176242172718048,
      "rewards/rejected": -0.005555284209549427,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.2518018074041684e-07,
      "logits/chosen": -1.9639079570770264,
      "logits/rejected": -1.9629888534545898,
      "logps/chosen": -32.960304260253906,
      "logps/rejected": -32.5081787109375,
      "loss": 0.5048,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.1930757313966751,
      "rewards/margins": 0.21889439225196838,
      "rewards/rejected": -0.02581869065761566,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-08,
      "logits/chosen": -1.9195839166641235,
      "logits/rejected": -1.929896593093872,
      "logps/chosen": -31.997264862060547,
      "logps/rejected": -35.334999084472656,
      "loss": 0.5364,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.12445466220378876,
      "rewards/margins": 0.15219925343990326,
      "rewards/rejected": -0.027744591236114502,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050325e-08,
      "logits/chosen": -2.058486223220825,
      "logits/rejected": -2.051971912384033,
      "logps/chosen": -33.44487380981445,
      "logps/rejected": -29.261219024658203,
      "loss": 0.4692,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.12425780296325684,
      "rewards/margins": 0.1451319307088852,
      "rewards/rejected": -0.020874138921499252,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-09,
      "logits/chosen": -1.9182488918304443,
      "logits/rejected": -1.9204342365264893,
      "logps/chosen": -33.968048095703125,
      "logps/rejected": -30.890918731689453,
      "loss": 0.4679,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.16260387003421783,
      "rewards/margins": 0.17357683181762695,
      "rewards/rejected": -0.010972961783409119,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.5965218513042896,
      "train_runtime": 3252.2803,
      "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
}