{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9985693848354793,
  "eval_steps": 100,
  "global_step": 349,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 1483.079147156977,
      "learning_rate": 5.7142857142857136e-09,
      "logits/chosen": -4.490396976470947,
      "logits/rejected": -4.787891387939453,
      "logps/chosen": -300.56573486328125,
      "logps/rejected": -263.39849853515625,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "grad_norm": 1678.9009350663757,
      "learning_rate": 5.714285714285714e-08,
      "logits/chosen": -4.292892932891846,
      "logits/rejected": -4.4704790115356445,
      "logps/chosen": -285.7367858886719,
      "logps/rejected": -241.40879821777344,
      "loss": 0.7232,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 0.03021715022623539,
      "rewards/margins": 0.018014244735240936,
      "rewards/rejected": 0.012202905490994453,
      "step": 10
    },
    {
      "epoch": 0.06,
      "grad_norm": 1307.5912774162387,
      "learning_rate": 1.1428571428571427e-07,
      "logits/chosen": -4.212913990020752,
      "logits/rejected": -4.504936695098877,
      "logps/chosen": -312.25714111328125,
      "logps/rejected": -247.5605926513672,
      "loss": 0.6348,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": 0.25936827063560486,
      "rewards/margins": 0.15551437437534332,
      "rewards/rejected": 0.10385391861200333,
      "step": 20
    },
    {
      "epoch": 0.09,
      "grad_norm": 877.5830541670848,
      "learning_rate": 1.714285714285714e-07,
      "logits/chosen": -4.158999919891357,
      "logits/rejected": -4.378296852111816,
      "logps/chosen": -320.1895446777344,
      "logps/rejected": -273.6168518066406,
      "loss": 0.4394,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": 1.6762861013412476,
      "rewards/margins": 1.1370527744293213,
      "rewards/rejected": 0.5392333269119263,
      "step": 30
    },
    {
      "epoch": 0.11,
      "grad_norm": 1094.7268353859254,
      "learning_rate": 1.9987489925657699e-07,
      "logits/chosen": -4.327412128448486,
      "logits/rejected": -4.590438365936279,
      "logps/chosen": -272.52813720703125,
      "logps/rejected": -226.4479522705078,
      "loss": 0.3519,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 3.0370898246765137,
      "rewards/margins": 2.1227898597717285,
      "rewards/rejected": 0.9143003225326538,
      "step": 40
    },
    {
      "epoch": 0.14,
      "grad_norm": 1142.545986046605,
      "learning_rate": 1.98875970549573e-07,
      "logits/chosen": -4.374251365661621,
      "logits/rejected": -4.627996921539307,
      "logps/chosen": -264.49615478515625,
      "logps/rejected": -226.27197265625,
      "loss": 0.3689,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 4.1793413162231445,
      "rewards/margins": 2.736725091934204,
      "rewards/rejected": 1.4426158666610718,
      "step": 50
    },
    {
      "epoch": 0.17,
      "grad_norm": 635.1242404757879,
      "learning_rate": 1.968881042201029e-07,
      "logits/chosen": -4.35978889465332,
      "logits/rejected": -4.594507694244385,
      "logps/chosen": -296.86907958984375,
      "logps/rejected": -251.1016845703125,
      "loss": 0.3246,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": 4.9698944091796875,
      "rewards/margins": 3.8340022563934326,
      "rewards/rejected": 1.1358922719955444,
      "step": 60
    },
    {
      "epoch": 0.2,
      "grad_norm": 871.5895225513464,
      "learning_rate": 1.9393118250841894e-07,
      "logits/chosen": -4.349625587463379,
      "logits/rejected": -4.586968421936035,
      "logps/chosen": -285.0860595703125,
      "logps/rejected": -241.4550018310547,
      "loss": 0.3259,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 4.21413516998291,
      "rewards/margins": 3.7849984169006348,
      "rewards/rejected": 0.42913690209388733,
      "step": 70
    },
    {
      "epoch": 0.23,
      "grad_norm": 869.2800598030351,
      "learning_rate": 1.900347799523094e-07,
      "logits/chosen": -4.277868747711182,
      "logits/rejected": -4.535857677459717,
      "logps/chosen": -318.2683410644531,
      "logps/rejected": -272.18206787109375,
      "loss": 0.2501,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 4.3253655433654785,
      "rewards/margins": 4.170114040374756,
      "rewards/rejected": 0.15525080263614655,
      "step": 80
    },
    {
      "epoch": 0.26,
      "grad_norm": 616.9171514391553,
      "learning_rate": 1.8523786758850436e-07,
      "logits/chosen": -4.401219844818115,
      "logits/rejected": -4.6477251052856445,
      "logps/chosen": -280.7421569824219,
      "logps/rejected": -236.2926025390625,
      "loss": 0.2859,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 3.923023223876953,
      "rewards/margins": 4.055254936218262,
      "rewards/rejected": -0.13223108649253845,
      "step": 90
    },
    {
      "epoch": 0.29,
      "grad_norm": 735.4670744748103,
      "learning_rate": 1.795884231721841e-07,
      "logits/chosen": -4.318561553955078,
      "logits/rejected": -4.602439880371094,
      "logps/chosen": -267.45794677734375,
      "logps/rejected": -232.0417938232422,
      "loss": 0.3114,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 4.1319708824157715,
      "rewards/margins": 4.702456951141357,
      "rewards/rejected": -0.5704857110977173,
      "step": 100
    },
    {
      "epoch": 0.29,
      "eval_logits/chosen": -3.226163148880005,
      "eval_logits/rejected": -3.226163148880005,
      "eval_logps/chosen": -161.294677734375,
      "eval_logps/rejected": -161.294677734375,
      "eval_loss": 0.6931471824645996,
      "eval_rewards/accuracies": 0.0,
      "eval_rewards/chosen": -2.8588294982910156,
      "eval_rewards/margins": 0.0,
      "eval_rewards/rejected": -2.8588294982910156,
      "eval_runtime": 1.5114,
      "eval_samples_per_second": 0.662,
      "eval_steps_per_second": 0.662,
      "step": 100
    },
    {
      "epoch": 0.31,
      "grad_norm": 817.8641381018862,
      "learning_rate": 1.7314295131309637e-07,
      "logits/chosen": -4.324714660644531,
      "logits/rejected": -4.614729404449463,
      "logps/chosen": -293.42364501953125,
      "logps/rejected": -245.01620483398438,
      "loss": 0.3079,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 4.0577192306518555,
      "rewards/margins": 4.518521308898926,
      "rewards/rejected": -0.4608024060726166,
      "step": 110
    },
    {
      "epoch": 0.34,
      "grad_norm": 926.642960527248,
      "learning_rate": 1.6596591832778466e-07,
      "logits/chosen": -4.334780216217041,
      "logits/rejected": -4.567566394805908,
      "logps/chosen": -298.56976318359375,
      "logps/rejected": -266.8869323730469,
      "loss": 0.2848,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 3.83539080619812,
      "rewards/margins": 4.040234088897705,
      "rewards/rejected": -0.20484237372875214,
      "step": 120
    },
    {
      "epoch": 0.37,
      "grad_norm": 675.9811999128359,
      "learning_rate": 1.5812910746042256e-07,
      "logits/chosen": -4.3064866065979,
      "logits/rejected": -4.553826332092285,
      "logps/chosen": -286.5328063964844,
      "logps/rejected": -242.8630828857422,
      "loss": 0.2782,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 3.7822985649108887,
      "rewards/margins": 4.371214389801025,
      "rewards/rejected": -0.5889158248901367,
      "step": 130
    },
    {
      "epoch": 0.4,
      "grad_norm": 1048.4484509272795,
      "learning_rate": 1.4971090092120542e-07,
      "logits/chosen": -4.304565906524658,
      "logits/rejected": -4.545400142669678,
      "logps/chosen": -294.17095947265625,
      "logps/rejected": -243.83935546875,
      "loss": 0.269,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 3.8339390754699707,
      "rewards/margins": 4.558647155761719,
      "rewards/rejected": -0.7247086763381958,
      "step": 140
    },
    {
      "epoch": 0.43,
      "grad_norm": 844.0558457976633,
      "learning_rate": 1.4079549592320782e-07,
      "logits/chosen": -4.313709735870361,
      "logits/rejected": -4.621119499206543,
      "logps/chosen": -292.96661376953125,
      "logps/rejected": -248.0418701171875,
      "loss": 0.2727,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 3.8533847332000732,
      "rewards/margins": 4.262558937072754,
      "rewards/rejected": -0.4091736674308777,
      "step": 150
    },
    {
      "epoch": 0.46,
      "grad_norm": 684.2161612155076,
      "learning_rate": 1.3147206255874882e-07,
      "logits/chosen": -4.248518943786621,
      "logits/rejected": -4.44361686706543,
      "logps/chosen": -309.7388610839844,
      "logps/rejected": -272.90252685546875,
      "loss": 0.2961,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 4.352166652679443,
      "rewards/margins": 4.611546516418457,
      "rewards/rejected": -0.2593800127506256,
      "step": 160
    },
    {
      "epoch": 0.49,
      "grad_norm": 795.319436926025,
      "learning_rate": 1.2183385193801653e-07,
      "logits/chosen": -4.289888381958008,
      "logits/rejected": -4.489100933074951,
      "logps/chosen": -272.3153991699219,
      "logps/rejected": -237.64114379882812,
      "loss": 0.2547,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 4.135977745056152,
      "rewards/margins": 4.544943809509277,
      "rewards/rejected": -0.40896692872047424,
      "step": 170
    },
    {
      "epoch": 0.52,
      "grad_norm": 917.2913745554154,
      "learning_rate": 1.1197726351017051e-07,
      "logits/chosen": -4.37355375289917,
      "logits/rejected": -4.5346221923828125,
      "logps/chosen": -261.6694641113281,
      "logps/rejected": -230.89599609375,
      "loss": 0.3104,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 3.4763107299804688,
      "rewards/margins": 3.877528429031372,
      "rewards/rejected": -0.4012181758880615,
      "step": 180
    },
    {
      "epoch": 0.54,
      "grad_norm": 880.6071936953809,
      "learning_rate": 1.0200088089538943e-07,
      "logits/chosen": -4.33815860748291,
      "logits/rejected": -4.5543646812438965,
      "logps/chosen": -288.26873779296875,
      "logps/rejected": -253.7147216796875,
      "loss": 0.3129,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 4.4660868644714355,
      "rewards/margins": 4.566544055938721,
      "rewards/rejected": -0.10045762360095978,
      "step": 190
    },
    {
      "epoch": 0.57,
      "grad_norm": 781.3066688232569,
      "learning_rate": 9.20044858712785e-08,
      "logits/chosen": -4.292983055114746,
      "logits/rejected": -4.566291332244873,
      "logps/chosen": -315.58251953125,
      "logps/rejected": -270.0286560058594,
      "loss": 0.2741,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 4.502341270446777,
      "rewards/margins": 4.4218339920043945,
      "rewards/rejected": 0.08050797879695892,
      "step": 200
    },
    {
      "epoch": 0.57,
      "eval_logits/chosen": -3.2033309936523438,
      "eval_logits/rejected": -3.2033309936523438,
      "eval_logps/chosen": -160.92892456054688,
      "eval_logps/rejected": -160.92892456054688,
      "eval_loss": 0.6931471824645996,
      "eval_rewards/accuracies": 0.0,
      "eval_rewards/chosen": -2.675952911376953,
      "eval_rewards/margins": 0.0,
      "eval_rewards/rejected": -2.675952911376953,
      "eval_runtime": 1.48,
      "eval_samples_per_second": 0.676,
      "eval_steps_per_second": 0.676,
      "step": 200
    },
    {
      "epoch": 0.6,
      "grad_norm": 1021.5696401893487,
      "learning_rate": 8.208806037554645e-08,
      "logits/chosen": -4.327147483825684,
      "logits/rejected": -4.6312479972839355,
      "logps/chosen": -292.85919189453125,
      "logps/rejected": -236.7397918701172,
      "loss": 0.2345,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 4.086556434631348,
      "rewards/margins": 4.589372634887695,
      "rewards/rejected": -0.5028160810470581,
      "step": 210
    },
    {
      "epoch": 0.63,
      "grad_norm": 720.6830161967047,
      "learning_rate": 7.23507865067214e-08,
      "logits/chosen": -4.357525825500488,
      "logits/rejected": -4.7082133293151855,
      "logps/chosen": -298.2110900878906,
      "logps/rejected": -237.28903198242188,
      "loss": 0.241,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 3.84515118598938,
      "rewards/margins": 4.655531406402588,
      "rewards/rejected": -0.8103805780410767,
      "step": 220
    },
    {
      "epoch": 0.66,
      "grad_norm": 784.9312094778309,
      "learning_rate": 6.289005452469778e-08,
      "logits/chosen": -4.388636112213135,
      "logits/rejected": -4.748753070831299,
      "logps/chosen": -288.0061340332031,
      "logps/rejected": -236.6077880859375,
      "loss": 0.268,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 3.5908846855163574,
      "rewards/margins": 4.1128339767456055,
      "rewards/rejected": -0.5219489932060242,
      "step": 230
    },
    {
      "epoch": 0.69,
      "grad_norm": 765.61870702253,
      "learning_rate": 5.38004887728938e-08,
      "logits/chosen": -4.234927177429199,
      "logits/rejected": -4.550938606262207,
      "logps/chosen": -283.8558654785156,
      "logps/rejected": -242.86538696289062,
      "loss": 0.2908,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 3.562377452850342,
      "rewards/margins": 4.074954032897949,
      "rewards/rejected": -0.5125768780708313,
      "step": 240
    },
    {
      "epoch": 0.72,
      "grad_norm": 787.5249084258048,
      "learning_rate": 4.517300126455066e-08,
      "logits/chosen": -4.290322303771973,
      "logits/rejected": -4.522026538848877,
      "logps/chosen": -292.47442626953125,
      "logps/rejected": -257.0675354003906,
      "loss": 0.2635,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 3.95862078666687,
      "rewards/margins": 4.515480995178223,
      "rewards/rejected": -0.5568601489067078,
      "step": 250
    },
    {
      "epoch": 0.74,
      "grad_norm": 894.7487355536125,
      "learning_rate": 3.70938823990135e-08,
      "logits/chosen": -4.367636680603027,
      "logits/rejected": -4.524864196777344,
      "logps/chosen": -293.8681640625,
      "logps/rejected": -266.19622802734375,
      "loss": 0.2999,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 3.311814069747925,
      "rewards/margins": 3.883411407470703,
      "rewards/rejected": -0.5715969800949097,
      "step": 260
    },
    {
      "epoch": 0.77,
      "grad_norm": 582.5138549963486,
      "learning_rate": 2.9643937902467277e-08,
      "logits/chosen": -4.23429012298584,
      "logits/rejected": -4.521827697753906,
      "logps/chosen": -280.5379333496094,
      "logps/rejected": -241.7247314453125,
      "loss": 0.2388,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 3.911522626876831,
      "rewards/margins": 4.728894233703613,
      "rewards/rejected": -0.8173705339431763,
      "step": 270
    },
    {
      "epoch": 0.8,
      "grad_norm": 810.3954826895692,
      "learning_rate": 2.289768062527362e-08,
      "logits/chosen": -4.273584842681885,
      "logits/rejected": -4.6050519943237305,
      "logps/chosen": -283.619384765625,
      "logps/rejected": -233.9824981689453,
      "loss": 0.2895,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 3.576643466949463,
      "rewards/margins": 3.991179943084717,
      "rewards/rejected": -0.41453617811203003,
      "step": 280
    },
    {
      "epoch": 0.83,
      "grad_norm": 730.6997584190797,
      "learning_rate": 1.6922585279389035e-08,
      "logits/chosen": -4.406073093414307,
      "logits/rejected": -4.6901140213012695,
      "logps/chosen": -294.35699462890625,
      "logps/rejected": -233.74594116210938,
      "loss": 0.2669,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 3.295900344848633,
      "rewards/margins": 3.856853485107422,
      "rewards/rejected": -0.5609533786773682,
      "step": 290
    },
    {
      "epoch": 0.86,
      "grad_norm": 1254.3257141110294,
      "learning_rate": 1.1778413569831725e-08,
      "logits/chosen": -4.323153495788574,
      "logits/rejected": -4.481599807739258,
      "logps/chosen": -291.29742431640625,
      "logps/rejected": -254.1814422607422,
      "loss": 0.2695,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 3.8485710620880127,
      "rewards/margins": 3.8490569591522217,
      "rewards/rejected": -0.0004864335060119629,
      "step": 300
    },
    {
      "epoch": 0.86,
      "eval_logits/chosen": -3.199521064758301,
      "eval_logits/rejected": -3.199521064758301,
      "eval_logps/chosen": -159.83721923828125,
      "eval_logps/rejected": -159.83721923828125,
      "eval_loss": 0.6931471824645996,
      "eval_rewards/accuracies": 0.0,
      "eval_rewards/chosen": -2.1301021575927734,
      "eval_rewards/margins": 0.0,
      "eval_rewards/rejected": -2.1301021575927734,
      "eval_runtime": 1.4681,
      "eval_samples_per_second": 0.681,
      "eval_steps_per_second": 0.681,
      "step": 300
    },
    {
      "epoch": 0.89,
      "grad_norm": 1302.095010227008,
      "learning_rate": 7.516616470096315e-09,
      "logits/chosen": -4.3807525634765625,
      "logits/rejected": -4.600883483886719,
      "logps/chosen": -299.8600158691406,
      "logps/rejected": -248.4993438720703,
      "loss": 0.2545,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 4.215323448181152,
      "rewards/margins": 4.587640762329102,
      "rewards/rejected": -0.3723169267177582,
      "step": 310
    },
    {
      "epoch": 0.92,
      "grad_norm": 875.4939401121867,
      "learning_rate": 4.179819619838454e-09,
      "logits/chosen": -4.2180280685424805,
      "logits/rejected": -4.570239067077637,
      "logps/chosen": -286.0614013671875,
      "logps/rejected": -234.3032684326172,
      "loss": 0.269,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 3.9898929595947266,
      "rewards/margins": 4.13907527923584,
      "rewards/rejected": -0.1491830050945282,
      "step": 320
    },
    {
      "epoch": 0.94,
      "grad_norm": 1010.1848642033616,
      "learning_rate": 1.8013969917777483e-09,
      "logits/chosen": -4.380679130554199,
      "logits/rejected": -4.608093738555908,
      "logps/chosen": -278.8153381347656,
      "logps/rejected": -235.4047088623047,
      "loss": 0.247,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 3.914651870727539,
      "rewards/margins": 4.33533239364624,
      "rewards/rejected": -0.42068085074424744,
      "step": 330
    },
    {
      "epoch": 0.97,
      "grad_norm": 702.698454631269,
      "learning_rate": 4.051370919176289e-10,
      "logits/chosen": -4.314477920532227,
      "logits/rejected": -4.502932548522949,
      "logps/chosen": -282.89666748046875,
      "logps/rejected": -246.3631134033203,
      "loss": 0.243,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 3.8934428691864014,
      "rewards/margins": 4.308337688446045,
      "rewards/rejected": -0.41489481925964355,
      "step": 340
    },
    {
      "epoch": 1.0,
      "step": 349,
      "total_flos": 0.0,
      "train_loss": 0.30806411576476,
      "train_runtime": 5299.0292,
      "train_samples_per_second": 8.432,
      "train_steps_per_second": 0.066
    }
  ],
  "logging_steps": 10,
  "max_steps": 349,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}