{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.983957219251337,
  "eval_steps": 50,
  "global_step": 279,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.053475935828877004,
      "grad_norm": 65.08164204257967,
      "learning_rate": 5e-07,
      "logits/chosen": -2.733124017715454,
      "logits/rejected": -2.7089576721191406,
      "logps/chosen": -271.7913818359375,
      "logps/rejected": -220.1679229736328,
      "loss": 0.6907,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": 0.01848032884299755,
      "rewards/margins": 0.006348843220621347,
      "rewards/rejected": 0.01213148795068264,
      "step": 5
    },
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 50.59833986698045,
      "learning_rate": 1e-06,
      "logits/chosen": -2.66473388671875,
      "logits/rejected": -2.6718990802764893,
      "logps/chosen": -258.00506591796875,
      "logps/rejected": -206.7321319580078,
      "loss": 0.6381,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": 0.5401543378829956,
      "rewards/margins": 0.19901855289936066,
      "rewards/rejected": 0.3411358594894409,
      "step": 10
    },
    {
      "epoch": 0.16042780748663102,
      "grad_norm": 50.14290028396245,
      "learning_rate": 9.991477798614637e-07,
      "logits/chosen": -2.508185863494873,
      "logits/rejected": -2.524341106414795,
      "logps/chosen": -271.5559997558594,
      "logps/rejected": -203.901123046875,
      "loss": 0.6476,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 1.8893938064575195,
      "rewards/margins": 0.9772478342056274,
      "rewards/rejected": 0.9121459126472473,
      "step": 15
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 42.09940353495551,
      "learning_rate": 9.965940245625131e-07,
      "logits/chosen": -2.3450875282287598,
      "logits/rejected": -2.3332550525665283,
      "logps/chosen": -231.1672821044922,
      "logps/rejected": -189.68919372558594,
      "loss": 0.6411,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 1.9618078470230103,
      "rewards/margins": 1.288054347038269,
      "rewards/rejected": 0.6737534403800964,
      "step": 20
    },
    {
      "epoch": 0.26737967914438504,
      "grad_norm": 50.008448633201716,
      "learning_rate": 9.923474395499264e-07,
      "logits/chosen": -2.2368509769439697,
      "logits/rejected": -2.2110252380371094,
      "logps/chosen": -238.6009063720703,
      "logps/rejected": -201.56163024902344,
      "loss": 0.6356,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.4835152626037598,
      "rewards/margins": 0.7462359666824341,
      "rewards/rejected": 0.7372792959213257,
      "step": 25
    },
    {
      "epoch": 0.32085561497326204,
      "grad_norm": 37.46999153089867,
      "learning_rate": 9.86422500924775e-07,
      "logits/chosen": -2.279388904571533,
      "logits/rejected": -2.2622551918029785,
      "logps/chosen": -210.3643798828125,
      "logps/rejected": -200.84750366210938,
      "loss": 0.5918,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.0033924579620361,
      "rewards/margins": 0.8053790330886841,
      "rewards/rejected": 0.1980135291814804,
      "step": 30
    },
    {
      "epoch": 0.37433155080213903,
      "grad_norm": 40.36951895662085,
      "learning_rate": 9.788394060951227e-07,
      "logits/chosen": -2.380889654159546,
      "logits/rejected": -2.3877792358398438,
      "logps/chosen": -242.85470581054688,
      "logps/rejected": -207.0241241455078,
      "loss": 0.6045,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.8806816935539246,
      "rewards/margins": 0.9049984812736511,
      "rewards/rejected": -0.024316836148500443,
      "step": 35
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 49.54969179550948,
      "learning_rate": 9.696240049254742e-07,
      "logits/chosen": -2.361086368560791,
      "logits/rejected": -2.358868360519409,
      "logps/chosen": -250.7805633544922,
      "logps/rejected": -223.0339813232422,
      "loss": 0.6251,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.8420997858047485,
      "rewards/margins": 1.0070801973342896,
      "rewards/rejected": -0.16498039662837982,
      "step": 40
    },
    {
      "epoch": 0.48128342245989303,
      "grad_norm": 38.326754460806335,
      "learning_rate": 9.588077116176756e-07,
      "logits/chosen": -2.3445451259613037,
      "logits/rejected": -2.329608917236328,
      "logps/chosen": -244.2638702392578,
      "logps/rejected": -213.9591522216797,
      "loss": 0.6052,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.9911876916885376,
      "rewards/margins": 0.9155967831611633,
      "rewards/rejected": 0.07559084892272949,
      "step": 45
    },
    {
      "epoch": 0.5347593582887701,
      "grad_norm": 41.70114459904792,
      "learning_rate": 9.464273976236516e-07,
      "logits/chosen": -2.3615097999572754,
      "logits/rejected": -2.375725269317627,
      "logps/chosen": -235.63998413085938,
      "logps/rejected": -205.462158203125,
      "loss": 0.5778,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.0482008457183838,
      "rewards/margins": 1.3194622993469238,
      "rewards/rejected": -0.27126139402389526,
      "step": 50
    },
    {
      "epoch": 0.5347593582887701,
      "eval_logits/chosen": -2.382444381713867,
      "eval_logits/rejected": -2.353994369506836,
      "eval_logps/chosen": -246.02862548828125,
      "eval_logps/rejected": -197.20895385742188,
      "eval_loss": 0.5711008310317993,
      "eval_rewards/accuracies": 0.7559523582458496,
      "eval_rewards/chosen": 1.21534264087677,
      "eval_rewards/margins": 1.2842590808868408,
      "eval_rewards/rejected": -0.0689164251089096,
      "eval_runtime": 178.2179,
      "eval_samples_per_second": 14.926,
      "eval_steps_per_second": 0.236,
      "step": 50
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 37.61770640976083,
      "learning_rate": 9.325252659550308e-07,
      "logits/chosen": -2.4103198051452637,
      "logits/rejected": -2.375884532928467,
      "logps/chosen": -232.7388458251953,
      "logps/rejected": -197.4444122314453,
      "loss": 0.5385,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.110247015953064,
      "rewards/margins": 1.345210075378418,
      "rewards/rejected": -0.23496298491954803,
      "step": 55
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 36.284977037429975,
      "learning_rate": 9.171487073181197e-07,
      "logits/chosen": -2.45546817779541,
      "logits/rejected": -2.4268698692321777,
      "logps/chosen": -264.7644958496094,
      "logps/rejected": -237.2240753173828,
      "loss": 0.5661,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 1.084460973739624,
      "rewards/margins": 1.5218535661697388,
      "rewards/rejected": -0.4373924136161804,
      "step": 60
    },
    {
      "epoch": 0.6951871657754011,
      "grad_norm": 37.36342990536676,
      "learning_rate": 9.003501385646448e-07,
      "logits/chosen": -2.4519729614257812,
      "logits/rejected": -2.4639739990234375,
      "logps/chosen": -251.09231567382812,
      "logps/rejected": -201.4525909423828,
      "loss": 0.5596,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 1.1350815296173096,
      "rewards/margins": 1.5527232885360718,
      "rewards/rejected": -0.4176417887210846,
      "step": 65
    },
    {
      "epoch": 0.7486631016042781,
      "grad_norm": 33.50003576842058,
      "learning_rate": 8.821868240089676e-07,
      "logits/chosen": -2.400097608566284,
      "logits/rejected": -2.377209186553955,
      "logps/chosen": -265.53155517578125,
      "logps/rejected": -226.5992431640625,
      "loss": 0.5263,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.7660782337188721,
      "rewards/margins": 1.5463287830352783,
      "rewards/rejected": -0.7802505493164062,
      "step": 70
    },
    {
      "epoch": 0.8021390374331551,
      "grad_norm": 35.415664255337724,
      "learning_rate": 8.62720680220876e-07,
      "logits/chosen": -2.389862060546875,
      "logits/rejected": -2.379544258117676,
      "logps/chosen": -250.21115112304688,
      "logps/rejected": -234.2509765625,
      "loss": 0.5073,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.7700754404067993,
      "rewards/margins": 1.3676471710205078,
      "rewards/rejected": -0.597571849822998,
      "step": 75
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 38.29403188930404,
      "learning_rate": 8.420180649593929e-07,
      "logits/chosen": -2.42349910736084,
      "logits/rejected": -2.4062836170196533,
      "logps/chosen": -261.1897277832031,
      "logps/rejected": -225.38619995117188,
      "loss": 0.5523,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.5706126689910889,
      "rewards/margins": 1.5985792875289917,
      "rewards/rejected": -1.0279667377471924,
      "step": 80
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 34.7463437300499,
      "learning_rate": 8.201495509671036e-07,
      "logits/chosen": -2.4246888160705566,
      "logits/rejected": -2.438143253326416,
      "logps/chosen": -259.6698303222656,
      "logps/rejected": -213.34518432617188,
      "loss": 0.5362,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 1.5000348091125488,
      "rewards/margins": 1.8813749551773071,
      "rewards/rejected": -0.3813403248786926,
      "step": 85
    },
    {
      "epoch": 0.9625668449197861,
      "grad_norm": 34.01709977955445,
      "learning_rate": 7.971896853961042e-07,
      "logits/chosen": -2.4786272048950195,
      "logits/rejected": -2.4621052742004395,
      "logps/chosen": -252.6436004638672,
      "logps/rejected": -250.2777557373047,
      "loss": 0.5064,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.4901584386825562,
      "rewards/margins": 1.7402337789535522,
      "rewards/rejected": -0.25007539987564087,
      "step": 90
    },
    {
      "epoch": 1.0160427807486632,
      "grad_norm": 24.058087864441212,
      "learning_rate": 7.732167356856654e-07,
      "logits/chosen": -2.5284178256988525,
      "logits/rejected": -2.500856876373291,
      "logps/chosen": -245.199462890625,
      "logps/rejected": -247.8142852783203,
      "loss": 0.4555,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 1.14225172996521,
      "rewards/margins": 2.6005594730377197,
      "rewards/rejected": -1.4583076238632202,
      "step": 95
    },
    {
      "epoch": 1.0695187165775402,
      "grad_norm": 34.758664424975294,
      "learning_rate": 7.48312422757881e-07,
      "logits/chosen": -2.516972064971924,
      "logits/rejected": -2.514902114868164,
      "logps/chosen": -236.08767700195312,
      "logps/rejected": -226.1017608642578,
      "loss": 0.232,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.0098340511322021,
      "rewards/margins": 3.1043362617492676,
      "rewards/rejected": -2.0945024490356445,
      "step": 100
    },
    {
      "epoch": 1.0695187165775402,
      "eval_logits/chosen": -2.5262153148651123,
      "eval_logits/rejected": -2.4977195262908936,
      "eval_logps/chosen": -248.79345703125,
      "eval_logps/rejected": -207.30593872070312,
      "eval_loss": 0.49347466230392456,
      "eval_rewards/accuracies": 0.8035714030265808,
      "eval_rewards/chosen": 0.938858687877655,
      "eval_rewards/margins": 2.0174732208251953,
      "eval_rewards/rejected": -1.078614354133606,
      "eval_runtime": 177.1838,
      "eval_samples_per_second": 15.013,
      "eval_steps_per_second": 0.237,
      "step": 100
    },
    {
      "epoch": 1.1229946524064172,
      "grad_norm": 23.33568729706663,
      "learning_rate": 7.225616424408044e-07,
      "logits/chosen": -2.514841318130493,
      "logits/rejected": -2.4978649616241455,
      "logps/chosen": -235.51412963867188,
      "logps/rejected": -235.5688018798828,
      "loss": 0.2233,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.734924554824829,
      "rewards/margins": 3.03052020072937,
      "rewards/rejected": -1.2955958843231201,
      "step": 105
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 34.28901770913126,
      "learning_rate": 6.96052176068713e-07,
      "logits/chosen": -2.4421210289001465,
      "logits/rejected": -2.439652919769287,
      "logps/chosen": -253.3994140625,
      "logps/rejected": -220.8712615966797,
      "loss": 0.2622,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.7797740697860718,
      "rewards/margins": 3.0600085258483887,
      "rewards/rejected": -1.2802345752716064,
      "step": 110
    },
    {
      "epoch": 1.2299465240641712,
      "grad_norm": 25.642631790616498,
      "learning_rate": 6.688743912460229e-07,
      "logits/chosen": -2.429490566253662,
      "logits/rejected": -2.395030975341797,
      "logps/chosen": -220.0726318359375,
      "logps/rejected": -223.9545135498047,
      "loss": 0.2755,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.3883707523345947,
      "rewards/margins": 3.119737386703491,
      "rewards/rejected": -1.731366515159607,
      "step": 115
    },
    {
      "epoch": 1.2834224598930482,
      "grad_norm": 20.428351141406825,
      "learning_rate": 6.411209337949213e-07,
      "logits/chosen": -2.4585251808166504,
      "logits/rejected": -2.4200589656829834,
      "logps/chosen": -248.60250854492188,
      "logps/rejected": -234.059814453125,
      "loss": 0.2087,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.7997968196868896,
      "rewards/margins": 3.555243730545044,
      "rewards/rejected": -1.755446195602417,
      "step": 120
    },
    {
      "epoch": 1.3368983957219251,
      "grad_norm": 23.894924535095097,
      "learning_rate": 6.128864119368233e-07,
      "logits/chosen": -2.4363160133361816,
      "logits/rejected": -2.409313201904297,
      "logps/chosen": -238.1552734375,
      "logps/rejected": -241.06640625,
      "loss": 0.311,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.5000994205474854,
      "rewards/margins": 3.579270839691162,
      "rewards/rejected": -2.0791714191436768,
      "step": 125
    },
    {
      "epoch": 1.3903743315508021,
      "grad_norm": 25.85137765086269,
      "learning_rate": 5.842670737842467e-07,
      "logits/chosen": -2.4068820476531982,
      "logits/rejected": -2.3998050689697266,
      "logps/chosen": -243.46554565429688,
      "logps/rejected": -254.32815551757812,
      "loss": 0.2738,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.553800106048584,
      "rewards/margins": 3.620004653930664,
      "rewards/rejected": -2.06620454788208,
      "step": 130
    },
    {
      "epoch": 1.4438502673796791,
      "grad_norm": 23.3391055271991,
      "learning_rate": 5.553604792424922e-07,
      "logits/chosen": -2.402261734008789,
      "logits/rejected": -2.388066530227661,
      "logps/chosen": -257.2569580078125,
      "logps/rejected": -232.8003387451172,
      "loss": 0.2369,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.5672048330307007,
      "rewards/margins": 3.8800740242004395,
      "rewards/rejected": -2.312869071960449,
      "step": 135
    },
    {
      "epoch": 1.4973262032085561,
      "grad_norm": 22.523493114917184,
      "learning_rate": 5.262651674395798e-07,
      "logits/chosen": -2.3823370933532715,
      "logits/rejected": -2.363351583480835,
      "logps/chosen": -251.1205291748047,
      "logps/rejected": -235.83895874023438,
      "loss": 0.2554,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.209763526916504,
      "rewards/margins": 3.3722991943359375,
      "rewards/rejected": -2.1625359058380127,
      "step": 140
    },
    {
      "epoch": 1.5508021390374331,
      "grad_norm": 22.629390451583,
      "learning_rate": 4.970803208181314e-07,
      "logits/chosen": -2.3576979637145996,
      "logits/rejected": -2.341073513031006,
      "logps/chosen": -247.33438110351562,
      "logps/rejected": -270.8658142089844,
      "loss": 0.2808,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 1.388020634651184,
      "rewards/margins": 3.6264255046844482,
      "rewards/rejected": -2.2384045124053955,
      "step": 145
    },
    {
      "epoch": 1.6042780748663101,
      "grad_norm": 17.721318468581117,
      "learning_rate": 4.679054270342702e-07,
      "logits/chosen": -2.3148550987243652,
      "logits/rejected": -2.2610244750976562,
      "logps/chosen": -278.3253479003906,
      "logps/rejected": -230.291259765625,
      "loss": 0.2427,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.1150439977645874,
      "rewards/margins": 3.5642380714416504,
      "rewards/rejected": -2.4491934776306152,
      "step": 150
    },
    {
      "epoch": 1.6042780748663101,
      "eval_logits/chosen": -2.2463791370391846,
      "eval_logits/rejected": -2.1962924003601074,
      "eval_logps/chosen": -252.03550720214844,
      "eval_logps/rejected": -213.7987823486328,
      "eval_loss": 0.4994485378265381,
      "eval_rewards/accuracies": 0.8184523582458496,
      "eval_rewards/chosen": 0.6146563291549683,
      "eval_rewards/margins": 2.342554807662964,
      "eval_rewards/rejected": -1.727898359298706,
      "eval_runtime": 177.1502,
      "eval_samples_per_second": 15.016,
      "eval_steps_per_second": 0.237,
      "step": 150
    },
    {
      "epoch": 1.6577540106951871,
      "grad_norm": 21.833026684822734,
      "learning_rate": 4.3883993981608567e-07,
      "logits/chosen": -2.187971591949463,
      "logits/rejected": -2.1873555183410645,
      "logps/chosen": -255.7908935546875,
      "logps/rejected": -228.4761962890625,
      "loss": 0.2553,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.4208053350448608,
      "rewards/margins": 3.8419246673583984,
      "rewards/rejected": -2.421119213104248,
      "step": 155
    },
    {
      "epoch": 1.7112299465240641,
      "grad_norm": 21.527834234868564,
      "learning_rate": 4.0998293993775234e-07,
      "logits/chosen": -2.1791014671325684,
      "logits/rejected": -2.1374058723449707,
      "logps/chosen": -249.51611328125,
      "logps/rejected": -237.10806274414062,
      "loss": 0.2627,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.1588197946548462,
      "rewards/margins": 3.3558552265167236,
      "rewards/rejected": -2.197035789489746,
      "step": 160
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 22.141924882028945,
      "learning_rate": 3.814327974650066e-07,
      "logits/chosen": -2.1892764568328857,
      "logits/rejected": -2.154066562652588,
      "logps/chosen": -245.1927032470703,
      "logps/rejected": -229.3457794189453,
      "loss": 0.2499,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.5559265613555908,
      "rewards/margins": 3.307781219482422,
      "rewards/rejected": -1.751854658126831,
      "step": 165
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 18.747114781201187,
      "learning_rate": 3.532868364233416e-07,
      "logits/chosen": -2.2867817878723145,
      "logits/rejected": -2.241206645965576,
      "logps/chosen": -235.2709503173828,
      "logps/rejected": -254.90792846679688,
      "loss": 0.2275,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.4621918201446533,
      "rewards/margins": 3.464813232421875,
      "rewards/rejected": -2.0026211738586426,
      "step": 170
    },
    {
      "epoch": 1.8716577540106951,
      "grad_norm": 22.34846640164264,
      "learning_rate": 3.256410030320304e-07,
      "logits/chosen": -2.348212480545044,
      "logits/rejected": -2.3325066566467285,
      "logps/chosen": -267.20892333984375,
      "logps/rejected": -237.50350952148438,
      "loss": 0.2749,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.7395166158676147,
      "rewards/margins": 4.169873237609863,
      "rewards/rejected": -2.430356502532959,
      "step": 175
    },
    {
      "epoch": 1.9251336898395723,
      "grad_norm": 22.707246637809515,
      "learning_rate": 2.985895386349233e-07,
      "logits/chosen": -2.4144673347473145,
      "logits/rejected": -2.401940107345581,
      "logps/chosen": -228.4314422607422,
      "logps/rejected": -238.22073364257812,
      "loss": 0.2368,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.1973397731781006,
      "rewards/margins": 3.5035414695739746,
      "rewards/rejected": -2.3062021732330322,
      "step": 180
    },
    {
      "epoch": 1.9786096256684491,
      "grad_norm": 27.972224583259134,
      "learning_rate": 2.7222465844296514e-07,
      "logits/chosen": -2.4469401836395264,
      "logits/rejected": -2.4393134117126465,
      "logps/chosen": -246.10189819335938,
      "logps/rejected": -265.03594970703125,
      "loss": 0.2766,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.647485375404358,
      "rewards/margins": 3.863076686859131,
      "rewards/rejected": -2.2155914306640625,
      "step": 185
    },
    {
      "epoch": 2.0320855614973263,
      "grad_norm": 16.97425283327867,
      "learning_rate": 2.466362371835544e-07,
      "logits/chosen": -2.438560962677002,
      "logits/rejected": -2.436560869216919,
      "logps/chosen": -251.0770263671875,
      "logps/rejected": -256.6495666503906,
      "loss": 0.1829,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 1.3526352643966675,
      "rewards/margins": 4.015156269073486,
      "rewards/rejected": -2.6625208854675293,
      "step": 190
    },
    {
      "epoch": 2.085561497326203,
      "grad_norm": 17.33378544754746,
      "learning_rate": 2.2191150272833386e-07,
      "logits/chosen": -2.4327006340026855,
      "logits/rejected": -2.426313877105713,
      "logps/chosen": -239.974609375,
      "logps/rejected": -263.90679931640625,
      "loss": 0.1495,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.4122151136398315,
      "rewards/margins": 4.446558475494385,
      "rewards/rejected": -3.0343427658081055,
      "step": 195
    },
    {
      "epoch": 2.1390374331550803,
      "grad_norm": 13.9334572780982,
      "learning_rate": 1.9813473874379395e-07,
      "logits/chosen": -2.4262537956237793,
      "logits/rejected": -2.377641201019287,
      "logps/chosen": -258.02496337890625,
      "logps/rejected": -253.2255859375,
      "loss": 0.1239,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.7413403987884521,
      "rewards/margins": 4.427127838134766,
      "rewards/rejected": -2.6857874393463135,
      "step": 200
    },
    {
      "epoch": 2.1390374331550803,
      "eval_logits/chosen": -2.3958089351654053,
      "eval_logits/rejected": -2.360522508621216,
      "eval_logps/chosen": -251.51199340820312,
      "eval_logps/rejected": -215.51217651367188,
      "eval_loss": 0.4940880835056305,
      "eval_rewards/accuracies": 0.8065476417541504,
      "eval_rewards/chosen": 0.6670060753822327,
      "eval_rewards/margins": 2.5662453174591064,
      "eval_rewards/rejected": -1.8992393016815186,
      "eval_runtime": 177.0357,
      "eval_samples_per_second": 15.025,
      "eval_steps_per_second": 0.237,
      "step": 200
    },
    {
      "epoch": 2.192513368983957,
      "grad_norm": 11.582273345860106,
      "learning_rate": 1.7538699737832237e-07,
      "logits/chosen": -2.3853468894958496,
      "logits/rejected": -2.3745884895324707,
      "logps/chosen": -230.9375457763672,
      "logps/rejected": -217.5093994140625,
      "loss": 0.1469,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 1.4965482950210571,
      "rewards/margins": 4.085664749145508,
      "rewards/rejected": -2.5891165733337402,
      "step": 205
    },
    {
      "epoch": 2.2459893048128343,
      "grad_norm": 19.256743465727453,
      "learning_rate": 1.5374582296511053e-07,
      "logits/chosen": -2.3758203983306885,
      "logits/rejected": -2.350308895111084,
      "logps/chosen": -257.522705078125,
      "logps/rejected": -234.826171875,
      "loss": 0.1481,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.6718244552612305,
      "rewards/margins": 4.5953168869018555,
      "rewards/rejected": -2.923491954803467,
      "step": 210
    },
    {
      "epoch": 2.299465240641711,
      "grad_norm": 13.139408171077235,
      "learning_rate": 1.3328498768278418e-07,
      "logits/chosen": -2.3815619945526123,
      "logits/rejected": -2.3640360832214355,
      "logps/chosen": -238.7193603515625,
      "logps/rejected": -232.0906219482422,
      "loss": 0.1182,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 1.7218825817108154,
      "rewards/margins": 4.676151275634766,
      "rewards/rejected": -2.9542689323425293,
      "step": 215
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 13.083761092577205,
      "learning_rate": 1.1407424007485928e-07,
      "logits/chosen": -2.351168394088745,
      "logits/rejected": -2.362783908843994,
      "logps/chosen": -261.3135986328125,
      "logps/rejected": -223.7679443359375,
      "loss": 0.1353,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 1.8544862270355225,
      "rewards/margins": 4.578373908996582,
      "rewards/rejected": -2.7238874435424805,
      "step": 220
    },
    {
      "epoch": 2.406417112299465,
      "grad_norm": 22.686899523674633,
      "learning_rate": 9.617906728528679e-08,
      "logits/chosen": -2.3818743228912354,
      "logits/rejected": -2.3381073474884033,
      "logps/chosen": -249.3686065673828,
      "logps/rejected": -266.00616455078125,
      "loss": 0.1406,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 1.7818124294281006,
      "rewards/margins": 4.478957176208496,
      "rewards/rejected": -2.697144031524658,
      "step": 225
    },
    {
      "epoch": 2.4598930481283423,
      "grad_norm": 17.488012587001123,
      "learning_rate": 7.966047182060226e-08,
      "logits/chosen": -2.363152027130127,
      "logits/rejected": -2.341928482055664,
      "logps/chosen": -241.06747436523438,
      "logps/rejected": -220.3070831298828,
      "loss": 0.137,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.5707038640975952,
      "rewards/margins": 4.198428153991699,
      "rewards/rejected": -2.6277241706848145,
      "step": 230
    },
    {
      "epoch": 2.5133689839572195,
      "grad_norm": 16.158136019422187,
      "learning_rate": 6.457476359966684e-08,
      "logits/chosen": -2.3634676933288574,
      "logits/rejected": -2.3343586921691895,
      "logps/chosen": -230.6551055908203,
      "logps/rejected": -232.58920288085938,
      "loss": 0.1324,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 1.5435535907745361,
      "rewards/margins": 4.363201141357422,
      "rewards/rejected": -2.819648027420044,
      "step": 235
    },
    {
      "epoch": 2.5668449197860963,
      "grad_norm": 19.585075813526288,
      "learning_rate": 5.097336799988067e-08,
      "logits/chosen": -2.3822684288024902,
      "logits/rejected": -2.33119797706604,
      "logps/chosen": -243.6968994140625,
      "logps/rejected": -252.494873046875,
      "loss": 0.1575,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.535332441329956,
      "rewards/margins": 4.624737739562988,
      "rewards/rejected": -3.089405059814453,
      "step": 240
    },
    {
      "epoch": 2.620320855614973,
      "grad_norm": 16.504173280628983,
      "learning_rate": 3.8902650554212826e-08,
      "logits/chosen": -2.377624034881592,
      "logits/rejected": -2.357320785522461,
      "logps/chosen": -234.1246337890625,
      "logps/rejected": -233.6632537841797,
      "loss": 0.1404,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.701682686805725,
      "rewards/margins": 4.359586238861084,
      "rewards/rejected": -2.6579031944274902,
      "step": 245
    },
    {
      "epoch": 2.6737967914438503,
      "grad_norm": 19.105883784745725,
      "learning_rate": 2.8403758896638707e-08,
      "logits/chosen": -2.3754684925079346,
      "logits/rejected": -2.335705280303955,
      "logps/chosen": -261.24554443359375,
      "logps/rejected": -248.94570922851562,
      "loss": 0.1607,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 1.8046457767486572,
      "rewards/margins": 4.712311744689941,
      "rewards/rejected": -2.907665729522705,
      "step": 250
    },
    {
      "epoch": 2.6737967914438503,
      "eval_logits/chosen": -2.373189926147461,
      "eval_logits/rejected": -2.334873914718628,
      "eval_logps/chosen": -251.20956420898438,
      "eval_logps/rejected": -216.7368927001953,
      "eval_loss": 0.5006157755851746,
      "eval_rewards/accuracies": 0.800595223903656,
      "eval_rewards/chosen": 0.6972481608390808,
      "eval_rewards/margins": 2.718956708908081,
      "eval_rewards/rejected": -2.0217082500457764,
      "eval_runtime": 177.6741,
      "eval_samples_per_second": 14.971,
      "eval_steps_per_second": 0.236,
      "step": 250
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 16.859647997107373,
      "learning_rate": 1.951248249476961e-08,
      "logits/chosen": -2.390385150909424,
      "logits/rejected": -2.339979648590088,
      "logps/chosen": -232.9261932373047,
      "logps/rejected": -252.1141357421875,
      "loss": 0.1515,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.974938154220581,
      "rewards/margins": 4.697868824005127,
      "rewards/rejected": -2.722930431365967,
      "step": 255
    },
    {
      "epoch": 2.7807486631016043,
      "grad_norm": 18.408639136313546,
      "learning_rate": 1.2259130647833626e-08,
      "logits/chosen": -2.384657621383667,
      "logits/rejected": -2.3508825302124023,
      "logps/chosen": -231.360595703125,
      "logps/rejected": -238.8389892578125,
      "loss": 0.1517,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.6810638904571533,
      "rewards/margins": 4.430610179901123,
      "rewards/rejected": -2.7495458126068115,
      "step": 260
    },
    {
      "epoch": 2.834224598930481,
      "grad_norm": 19.216489097516313,
      "learning_rate": 6.668429165893996e-09,
      "logits/chosen": -2.379669427871704,
      "logits/rejected": -2.361882448196411,
      "logps/chosen": -261.16961669921875,
      "logps/rejected": -269.8806457519531,
      "loss": 0.1479,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.4186723232269287,
      "rewards/margins": 4.286049842834473,
      "rewards/rejected": -2.867377758026123,
      "step": 265
    },
    {
      "epoch": 2.8877005347593583,
      "grad_norm": 18.845102458023714,
      "learning_rate": 2.759436082516664e-09,
      "logits/chosen": -2.372851848602295,
      "logits/rejected": -2.356961488723755,
      "logps/chosen": -247.1389617919922,
      "logps/rejected": -257.267822265625,
      "loss": 0.1573,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.4503118991851807,
      "rewards/margins": 4.499074935913086,
      "rewards/rejected": -3.0487632751464844,
      "step": 270
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 16.5452733632814,
      "learning_rate": 5.454766882097006e-10,
      "logits/chosen": -2.3800816535949707,
      "logits/rejected": -2.3431155681610107,
      "logps/chosen": -242.17153930664062,
      "logps/rejected": -259.4877014160156,
      "loss": 0.1708,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.4261102676391602,
      "rewards/margins": 4.574359893798828,
      "rewards/rejected": -3.148249864578247,
      "step": 275
    },
    {
      "epoch": 2.983957219251337,
      "step": 279,
      "total_flos": 3289753017384960.0,
      "train_loss": 0.3301437693684759,
      "train_runtime": 10405.088,
      "train_samples_per_second": 6.9,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 5,
  "max_steps": 279,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3289753017384960.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}