{
  "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": 54.05966214679458,
      "learning_rate": 5e-07,
      "logits/chosen": -2.740234851837158,
      "logits/rejected": -2.7408976554870605,
      "logps/chosen": -260.93475341796875,
      "logps/rejected": -211.115966796875,
      "loss": 0.6911,
      "rewards/accuracies": 0.36250001192092896,
      "rewards/chosen": 0.019142702221870422,
      "rewards/margins": 0.0086396224796772,
      "rewards/rejected": 0.010503082536160946,
      "step": 5
    },
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 48.125498996039966,
      "learning_rate": 1e-06,
      "logits/chosen": -2.709315776824951,
      "logits/rejected": -2.7182488441467285,
      "logps/chosen": -258.59228515625,
      "logps/rejected": -225.8120880126953,
      "loss": 0.6539,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": 0.472653865814209,
      "rewards/margins": 0.14731308817863464,
      "rewards/rejected": 0.32534077763557434,
      "step": 10
    },
    {
      "epoch": 0.16042780748663102,
      "grad_norm": 44.380678098165866,
      "learning_rate": 9.991477798614637e-07,
      "logits/chosen": -2.6404635906219482,
      "logits/rejected": -2.6321334838867188,
      "logps/chosen": -245.1427764892578,
      "logps/rejected": -204.20162963867188,
      "loss": 0.6283,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 1.1205874681472778,
      "rewards/margins": 0.548652172088623,
      "rewards/rejected": 0.5719352960586548,
      "step": 15
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 45.00568883373122,
      "learning_rate": 9.965940245625131e-07,
      "logits/chosen": -2.5277099609375,
      "logits/rejected": -2.532583236694336,
      "logps/chosen": -265.33990478515625,
      "logps/rejected": -236.9266815185547,
      "loss": 0.6455,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 1.2614521980285645,
      "rewards/margins": 0.9182946085929871,
      "rewards/rejected": 0.34315773844718933,
      "step": 20
    },
    {
      "epoch": 0.26737967914438504,
      "grad_norm": 49.76731551004973,
      "learning_rate": 9.923474395499264e-07,
      "logits/chosen": -2.448563814163208,
      "logits/rejected": -2.4464802742004395,
      "logps/chosen": -234.9072723388672,
      "logps/rejected": -208.8161163330078,
      "loss": 0.6152,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.3159387111663818,
      "rewards/margins": 1.1293729543685913,
      "rewards/rejected": 0.18656572699546814,
      "step": 25
    },
    {
      "epoch": 0.32085561497326204,
      "grad_norm": 37.50219389396827,
      "learning_rate": 9.86422500924775e-07,
      "logits/chosen": -2.470647096633911,
      "logits/rejected": -2.449976682662964,
      "logps/chosen": -256.57049560546875,
      "logps/rejected": -202.13229370117188,
      "loss": 0.5704,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.158319354057312,
      "rewards/margins": 0.9538521766662598,
      "rewards/rejected": 0.20446734130382538,
      "step": 30
    },
    {
      "epoch": 0.37433155080213903,
      "grad_norm": 39.982844211127215,
      "learning_rate": 9.788394060951227e-07,
      "logits/chosen": -2.488635540008545,
      "logits/rejected": -2.4961345195770264,
      "logps/chosen": -247.16757202148438,
      "logps/rejected": -212.7868194580078,
      "loss": 0.6044,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.8941676020622253,
      "rewards/margins": 0.9727466702461243,
      "rewards/rejected": -0.07857902348041534,
      "step": 35
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 43.02526224638259,
      "learning_rate": 9.696240049254742e-07,
      "logits/chosen": -2.5331573486328125,
      "logits/rejected": -2.5155434608459473,
      "logps/chosen": -243.4714813232422,
      "logps/rejected": -212.7275848388672,
      "loss": 0.567,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 0.5864531397819519,
      "rewards/margins": 1.3161426782608032,
      "rewards/rejected": -0.7296894788742065,
      "step": 40
    },
    {
      "epoch": 0.48128342245989303,
      "grad_norm": 38.62113257038537,
      "learning_rate": 9.588077116176756e-07,
      "logits/chosen": -2.5625014305114746,
      "logits/rejected": -2.5336291790008545,
      "logps/chosen": -257.4891662597656,
      "logps/rejected": -241.340576171875,
      "loss": 0.5311,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 0.603995144367218,
      "rewards/margins": 1.318690538406372,
      "rewards/rejected": -0.7146955132484436,
      "step": 45
    },
    {
      "epoch": 0.5347593582887701,
      "grad_norm": 37.205626252672076,
      "learning_rate": 9.464273976236516e-07,
      "logits/chosen": -2.555994749069214,
      "logits/rejected": -2.5348007678985596,
      "logps/chosen": -246.0524444580078,
      "logps/rejected": -224.68038940429688,
      "loss": 0.5372,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.7058802843093872,
      "rewards/margins": 1.3246526718139648,
      "rewards/rejected": -0.6187724471092224,
      "step": 50
    },
    {
      "epoch": 0.5347593582887701,
      "eval_logits/chosen": -2.565211772918701,
      "eval_logits/rejected": -2.5557875633239746,
      "eval_logps/chosen": -242.1534881591797,
      "eval_logps/rejected": -226.5532989501953,
      "eval_loss": 0.5589047074317932,
      "eval_rewards/accuracies": 0.7708333134651184,
      "eval_rewards/chosen": 0.888508141040802,
      "eval_rewards/margins": 1.6572500467300415,
      "eval_rewards/rejected": -0.76874178647995,
      "eval_runtime": 179.6866,
      "eval_samples_per_second": 14.804,
      "eval_steps_per_second": 0.234,
      "step": 50
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 41.38413864195239,
      "learning_rate": 9.325252659550308e-07,
      "logits/chosen": -2.564159393310547,
      "logits/rejected": -2.562894105911255,
      "logps/chosen": -239.49169921875,
      "logps/rejected": -221.1046905517578,
      "loss": 0.5664,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.8376470804214478,
      "rewards/margins": 1.171051263809204,
      "rewards/rejected": -0.3334041237831116,
      "step": 55
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 38.183548449406196,
      "learning_rate": 9.171487073181197e-07,
      "logits/chosen": -2.5205636024475098,
      "logits/rejected": -2.5134239196777344,
      "logps/chosen": -247.12869262695312,
      "logps/rejected": -223.0773468017578,
      "loss": 0.5445,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 0.6957036256790161,
      "rewards/margins": 1.3918036222457886,
      "rewards/rejected": -0.6960999965667725,
      "step": 60
    },
    {
      "epoch": 0.6951871657754011,
      "grad_norm": 38.105501822781726,
      "learning_rate": 9.003501385646448e-07,
      "logits/chosen": -2.5033719539642334,
      "logits/rejected": -2.473064422607422,
      "logps/chosen": -230.86160278320312,
      "logps/rejected": -228.2729949951172,
      "loss": 0.5983,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.8065773248672485,
      "rewards/margins": 1.1710764169692993,
      "rewards/rejected": -0.3644990026950836,
      "step": 65
    },
    {
      "epoch": 0.7486631016042781,
      "grad_norm": 32.7431204523646,
      "learning_rate": 8.821868240089676e-07,
      "logits/chosen": -2.4544219970703125,
      "logits/rejected": -2.4576900005340576,
      "logps/chosen": -241.304443359375,
      "logps/rejected": -234.3035125732422,
      "loss": 0.5341,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": 1.1656383275985718,
      "rewards/margins": 2.016019582748413,
      "rewards/rejected": -0.8503812551498413,
      "step": 70
    },
    {
      "epoch": 0.8021390374331551,
      "grad_norm": 35.17657050274663,
      "learning_rate": 8.62720680220876e-07,
      "logits/chosen": -2.4114229679107666,
      "logits/rejected": -2.374790906906128,
      "logps/chosen": -244.16860961914062,
      "logps/rejected": -230.75845336914062,
      "loss": 0.5453,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 0.4619104862213135,
      "rewards/margins": 1.3655352592468262,
      "rewards/rejected": -0.9036247134208679,
      "step": 75
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 41.012118211896635,
      "learning_rate": 8.420180649593929e-07,
      "logits/chosen": -2.25986647605896,
      "logits/rejected": -2.228325366973877,
      "logps/chosen": -249.28677368164062,
      "logps/rejected": -226.74453735351562,
      "loss": 0.5689,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.25673240423202515,
      "rewards/margins": 1.7667404413223267,
      "rewards/rejected": -1.5100078582763672,
      "step": 80
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 32.649741196796036,
      "learning_rate": 8.201495509671036e-07,
      "logits/chosen": -2.2017948627471924,
      "logits/rejected": -2.1529741287231445,
      "logps/chosen": -272.0859069824219,
      "logps/rejected": -241.6033172607422,
      "loss": 0.4923,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.13092926144599915,
      "rewards/margins": 1.9662458896636963,
      "rewards/rejected": -1.8353164196014404,
      "step": 85
    },
    {
      "epoch": 0.9625668449197861,
      "grad_norm": 36.664818682351935,
      "learning_rate": 7.971896853961042e-07,
      "logits/chosen": -2.264296531677246,
      "logits/rejected": -2.21388840675354,
      "logps/chosen": -250.2943572998047,
      "logps/rejected": -235.5312042236328,
      "loss": 0.4863,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.07481655478477478,
      "rewards/margins": 1.6215366125106812,
      "rewards/rejected": -1.546720027923584,
      "step": 90
    },
    {
      "epoch": 1.0160427807486632,
      "grad_norm": 17.846722548016714,
      "learning_rate": 7.732167356856654e-07,
      "logits/chosen": -2.2805099487304688,
      "logits/rejected": -2.2655348777770996,
      "logps/chosen": -253.1170196533203,
      "logps/rejected": -233.9530487060547,
      "loss": 0.4497,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.8429386019706726,
      "rewards/margins": 2.112081289291382,
      "rewards/rejected": -1.269142746925354,
      "step": 95
    },
    {
      "epoch": 1.0695187165775402,
      "grad_norm": 23.335980228800327,
      "learning_rate": 7.48312422757881e-07,
      "logits/chosen": -2.319995164871216,
      "logits/rejected": -2.2926743030548096,
      "logps/chosen": -255.7274627685547,
      "logps/rejected": -237.66000366210938,
      "loss": 0.2535,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.8762633204460144,
      "rewards/margins": 2.882870674133301,
      "rewards/rejected": -2.0066072940826416,
      "step": 100
    },
    {
      "epoch": 1.0695187165775402,
      "eval_logits/chosen": -2.3488855361938477,
      "eval_logits/rejected": -2.315967559814453,
      "eval_logps/chosen": -244.4204864501953,
      "eval_logps/rejected": -233.458984375,
      "eval_loss": 0.5071699619293213,
      "eval_rewards/accuracies": 0.8154761791229248,
      "eval_rewards/chosen": 0.6618079543113708,
      "eval_rewards/margins": 2.1211202144622803,
      "eval_rewards/rejected": -1.459312081336975,
      "eval_runtime": 179.4652,
      "eval_samples_per_second": 14.822,
      "eval_steps_per_second": 0.234,
      "step": 100
    },
    {
      "epoch": 1.1229946524064172,
      "grad_norm": 18.045597792369282,
      "learning_rate": 7.225616424408044e-07,
      "logits/chosen": -2.3656322956085205,
      "logits/rejected": -2.3468470573425293,
      "logps/chosen": -224.35452270507812,
      "logps/rejected": -254.05020141601562,
      "loss": 0.248,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.2632067203521729,
      "rewards/margins": 3.3555703163146973,
      "rewards/rejected": -2.0923633575439453,
      "step": 105
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 19.830510405928298,
      "learning_rate": 6.96052176068713e-07,
      "logits/chosen": -2.3559365272521973,
      "logits/rejected": -2.32434344291687,
      "logps/chosen": -221.80929565429688,
      "logps/rejected": -223.14456176757812,
      "loss": 0.2495,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.21290123462677,
      "rewards/margins": 3.029930591583252,
      "rewards/rejected": -1.817029356956482,
      "step": 110
    },
    {
      "epoch": 1.2299465240641712,
      "grad_norm": 20.921727727462144,
      "learning_rate": 6.688743912460229e-07,
      "logits/chosen": -2.3118255138397217,
      "logits/rejected": -2.272325038909912,
      "logps/chosen": -224.7656707763672,
      "logps/rejected": -231.08969116210938,
      "loss": 0.2426,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 1.3237473964691162,
      "rewards/margins": 3.1390082836151123,
      "rewards/rejected": -1.8152612447738647,
      "step": 115
    },
    {
      "epoch": 1.2834224598930482,
      "grad_norm": 18.131576812677366,
      "learning_rate": 6.411209337949213e-07,
      "logits/chosen": -2.2410776615142822,
      "logits/rejected": -2.2443623542785645,
      "logps/chosen": -234.05697631835938,
      "logps/rejected": -252.6307373046875,
      "loss": 0.245,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 1.2462797164916992,
      "rewards/margins": 3.2672533988952637,
      "rewards/rejected": -2.0209736824035645,
      "step": 120
    },
    {
      "epoch": 1.3368983957219251,
      "grad_norm": 20.669432062488653,
      "learning_rate": 6.128864119368233e-07,
      "logits/chosen": -2.222790002822876,
      "logits/rejected": -2.178349256515503,
      "logps/chosen": -243.15139770507812,
      "logps/rejected": -244.16476440429688,
      "loss": 0.2539,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.1168737411499023,
      "rewards/margins": 3.560255765914917,
      "rewards/rejected": -2.4433822631835938,
      "step": 125
    },
    {
      "epoch": 1.3903743315508021,
      "grad_norm": 22.81645285626285,
      "learning_rate": 5.842670737842467e-07,
      "logits/chosen": -2.2315919399261475,
      "logits/rejected": -2.2085089683532715,
      "logps/chosen": -281.7901306152344,
      "logps/rejected": -241.4110107421875,
      "loss": 0.2643,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.744272232055664,
      "rewards/margins": 3.986635684967041,
      "rewards/rejected": -2.242363929748535,
      "step": 130
    },
    {
      "epoch": 1.4438502673796791,
      "grad_norm": 20.139205741708395,
      "learning_rate": 5.553604792424922e-07,
      "logits/chosen": -2.235245943069458,
      "logits/rejected": -2.2212934494018555,
      "logps/chosen": -251.8795928955078,
      "logps/rejected": -226.3660125732422,
      "loss": 0.2567,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.2548778057098389,
      "rewards/margins": 3.875605821609497,
      "rewards/rejected": -2.620728015899658,
      "step": 135
    },
    {
      "epoch": 1.4973262032085561,
      "grad_norm": 21.60802685620548,
      "learning_rate": 5.262651674395798e-07,
      "logits/chosen": -2.259312391281128,
      "logits/rejected": -2.211486339569092,
      "logps/chosen": -244.6748504638672,
      "logps/rejected": -232.0905303955078,
      "loss": 0.2779,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.3069220781326294,
      "rewards/margins": 3.8860652446746826,
      "rewards/rejected": -2.5791430473327637,
      "step": 140
    },
    {
      "epoch": 1.5508021390374331,
      "grad_norm": 19.201634463890933,
      "learning_rate": 4.970803208181314e-07,
      "logits/chosen": -2.251796245574951,
      "logits/rejected": -2.234971761703491,
      "logps/chosen": -253.1035919189453,
      "logps/rejected": -227.96426391601562,
      "loss": 0.2623,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.1842231750488281,
      "rewards/margins": 3.6010851860046387,
      "rewards/rejected": -2.4168620109558105,
      "step": 145
    },
    {
      "epoch": 1.6042780748663101,
      "grad_norm": 23.363468017275885,
      "learning_rate": 4.679054270342702e-07,
      "logits/chosen": -2.2931339740753174,
      "logits/rejected": -2.239758253097534,
      "logps/chosen": -252.3613739013672,
      "logps/rejected": -249.5667724609375,
      "loss": 0.2927,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 0.7659964561462402,
      "rewards/margins": 3.969538927078247,
      "rewards/rejected": -3.2035422325134277,
      "step": 150
    },
    {
      "epoch": 1.6042780748663101,
      "eval_logits/chosen": -2.2895591259002686,
      "eval_logits/rejected": -2.2493326663970947,
      "eval_logps/chosen": -248.8394012451172,
      "eval_logps/rejected": -243.3928985595703,
      "eval_loss": 0.5304741263389587,
      "eval_rewards/accuracies": 0.8065476417541504,
      "eval_rewards/chosen": 0.21991625428199768,
      "eval_rewards/margins": 2.6726186275482178,
      "eval_rewards/rejected": -2.452702760696411,
      "eval_runtime": 180.4978,
      "eval_samples_per_second": 14.737,
      "eval_steps_per_second": 0.233,
      "step": 150
    },
    {
      "epoch": 1.6577540106951871,
      "grad_norm": 20.076408881524735,
      "learning_rate": 4.3883993981608567e-07,
      "logits/chosen": -2.278191566467285,
      "logits/rejected": -2.2216737270355225,
      "logps/chosen": -275.7608642578125,
      "logps/rejected": -237.4488067626953,
      "loss": 0.2842,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 0.8101941347122192,
      "rewards/margins": 3.2896721363067627,
      "rewards/rejected": -2.479477882385254,
      "step": 155
    },
    {
      "epoch": 1.7112299465240641,
      "grad_norm": 19.861126459874942,
      "learning_rate": 4.0998293993775234e-07,
      "logits/chosen": -2.2196192741394043,
      "logits/rejected": -2.1777396202087402,
      "logps/chosen": -252.45205688476562,
      "logps/rejected": -251.8419189453125,
      "loss": 0.2776,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.1267125606536865,
      "rewards/margins": 3.9922633171081543,
      "rewards/rejected": -2.865550994873047,
      "step": 160
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 18.719503087791026,
      "learning_rate": 3.814327974650066e-07,
      "logits/chosen": -2.08862566947937,
      "logits/rejected": -2.0716958045959473,
      "logps/chosen": -239.0010528564453,
      "logps/rejected": -232.9193572998047,
      "loss": 0.3089,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.8289812207221985,
      "rewards/margins": 3.4957098960876465,
      "rewards/rejected": -2.666728973388672,
      "step": 165
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 22.16135538684767,
      "learning_rate": 3.532868364233416e-07,
      "logits/chosen": -2.05873966217041,
      "logits/rejected": -2.0186731815338135,
      "logps/chosen": -236.9549560546875,
      "logps/rejected": -251.7572784423828,
      "loss": 0.2894,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.7584196925163269,
      "rewards/margins": 3.4695358276367188,
      "rewards/rejected": -2.711116075515747,
      "step": 170
    },
    {
      "epoch": 1.8716577540106951,
      "grad_norm": 24.703187440369405,
      "learning_rate": 3.256410030320304e-07,
      "logits/chosen": -2.0796573162078857,
      "logits/rejected": -1.9813162088394165,
      "logps/chosen": -253.44644165039062,
      "logps/rejected": -235.35342407226562,
      "loss": 0.2781,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.8340685963630676,
      "rewards/margins": 3.6132419109344482,
      "rewards/rejected": -2.7791733741760254,
      "step": 175
    },
    {
      "epoch": 1.9251336898395723,
      "grad_norm": 22.202361530980614,
      "learning_rate": 2.985895386349233e-07,
      "logits/chosen": -2.0965940952301025,
      "logits/rejected": -2.0198874473571777,
      "logps/chosen": -247.35818481445312,
      "logps/rejected": -219.046142578125,
      "loss": 0.317,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.740056037902832,
      "rewards/margins": 3.212428569793701,
      "rewards/rejected": -2.472372531890869,
      "step": 180
    },
    {
      "epoch": 1.9786096256684491,
      "grad_norm": 24.566891431592218,
      "learning_rate": 2.7222465844296514e-07,
      "logits/chosen": -2.090613842010498,
      "logits/rejected": -2.0187251567840576,
      "logps/chosen": -235.1741485595703,
      "logps/rejected": -242.4431915283203,
      "loss": 0.2846,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.8609472513198853,
      "rewards/margins": 3.340895891189575,
      "rewards/rejected": -2.4799487590789795,
      "step": 185
    },
    {
      "epoch": 2.0320855614973263,
      "grad_norm": 17.297791640690683,
      "learning_rate": 2.466362371835544e-07,
      "logits/chosen": -2.10907244682312,
      "logits/rejected": -2.0432229042053223,
      "logps/chosen": -234.08016967773438,
      "logps/rejected": -258.77655029296875,
      "loss": 0.2268,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.8330599665641785,
      "rewards/margins": 3.731884479522705,
      "rewards/rejected": -2.898824453353882,
      "step": 190
    },
    {
      "epoch": 2.085561497326203,
      "grad_norm": 16.320947260850904,
      "learning_rate": 2.2191150272833386e-07,
      "logits/chosen": -2.153738260269165,
      "logits/rejected": -2.116497755050659,
      "logps/chosen": -256.75152587890625,
      "logps/rejected": -259.03619384765625,
      "loss": 0.1797,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.3713855743408203,
      "rewards/margins": 4.130622386932373,
      "rewards/rejected": -2.7592365741729736,
      "step": 195
    },
    {
      "epoch": 2.1390374331550803,
      "grad_norm": 20.361872847150345,
      "learning_rate": 1.9813473874379395e-07,
      "logits/chosen": -2.188068389892578,
      "logits/rejected": -2.12589430809021,
      "logps/chosen": -259.9382629394531,
      "logps/rejected": -239.9225616455078,
      "loss": 0.1722,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.5985621213912964,
      "rewards/margins": 4.089066982269287,
      "rewards/rejected": -2.4905049800872803,
      "step": 200
    },
    {
      "epoch": 2.1390374331550803,
      "eval_logits/chosen": -2.1841375827789307,
      "eval_logits/rejected": -2.126652479171753,
      "eval_logps/chosen": -245.60311889648438,
      "eval_logps/rejected": -240.4331817626953,
      "eval_loss": 0.49720147252082825,
      "eval_rewards/accuracies": 0.8303571343421936,
      "eval_rewards/chosen": 0.5435419678688049,
      "eval_rewards/margins": 2.700273036956787,
      "eval_rewards/rejected": -2.1567306518554688,
      "eval_runtime": 179.6054,
      "eval_samples_per_second": 14.81,
      "eval_steps_per_second": 0.234,
      "step": 200
    },
    {
      "epoch": 2.192513368983957,
      "grad_norm": 20.023776997346594,
      "learning_rate": 1.7538699737832237e-07,
      "logits/chosen": -2.167818546295166,
      "logits/rejected": -2.1456615924835205,
      "logps/chosen": -244.92453002929688,
      "logps/rejected": -239.2166290283203,
      "loss": 0.1663,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.2993017435073853,
      "rewards/margins": 4.142195224761963,
      "rewards/rejected": -2.842893362045288,
      "step": 205
    },
    {
      "epoch": 2.2459893048128343,
      "grad_norm": 10.710923875253577,
      "learning_rate": 1.5374582296511053e-07,
      "logits/chosen": -2.172811269760132,
      "logits/rejected": -2.1587228775024414,
      "logps/chosen": -229.21676635742188,
      "logps/rejected": -241.53805541992188,
      "loss": 0.1463,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.2647745609283447,
      "rewards/margins": 3.8649630546569824,
      "rewards/rejected": -2.6001884937286377,
      "step": 210
    },
    {
      "epoch": 2.299465240641711,
      "grad_norm": 14.963800725254186,
      "learning_rate": 1.3328498768278418e-07,
      "logits/chosen": -2.1864559650421143,
      "logits/rejected": -2.1280677318573,
      "logps/chosen": -242.97134399414062,
      "logps/rejected": -235.1353759765625,
      "loss": 0.1548,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.3184950351715088,
      "rewards/margins": 4.027216911315918,
      "rewards/rejected": -2.7087221145629883,
      "step": 215
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 16.654181879826538,
      "learning_rate": 1.1407424007485928e-07,
      "logits/chosen": -2.198726177215576,
      "logits/rejected": -2.162065029144287,
      "logps/chosen": -264.01043701171875,
      "logps/rejected": -252.46572875976562,
      "loss": 0.1595,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.1831305027008057,
      "rewards/margins": 4.2487406730651855,
      "rewards/rejected": -3.065610408782959,
      "step": 220
    },
    {
      "epoch": 2.406417112299465,
      "grad_norm": 15.572352212494664,
      "learning_rate": 9.617906728528679e-08,
      "logits/chosen": -2.221874713897705,
      "logits/rejected": -2.161702871322632,
      "logps/chosen": -241.98342895507812,
      "logps/rejected": -266.36346435546875,
      "loss": 0.1549,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.7255210876464844,
      "rewards/margins": 4.4200758934021,
      "rewards/rejected": -2.6945548057556152,
      "step": 225
    },
    {
      "epoch": 2.4598930481283423,
      "grad_norm": 15.999724580618913,
      "learning_rate": 7.966047182060226e-08,
      "logits/chosen": -2.2201499938964844,
      "logits/rejected": -2.157503843307495,
      "logps/chosen": -270.21563720703125,
      "logps/rejected": -256.5613098144531,
      "loss": 0.1484,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 0.8697912096977234,
      "rewards/margins": 3.904383897781372,
      "rewards/rejected": -3.034592866897583,
      "step": 230
    },
    {
      "epoch": 2.5133689839572195,
      "grad_norm": 16.887962505823275,
      "learning_rate": 6.457476359966684e-08,
      "logits/chosen": -2.214259147644043,
      "logits/rejected": -2.1480872631073,
      "logps/chosen": -240.1714324951172,
      "logps/rejected": -239.39773559570312,
      "loss": 0.1549,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.0319417715072632,
      "rewards/margins": 3.902411937713623,
      "rewards/rejected": -2.8704705238342285,
      "step": 235
    },
    {
      "epoch": 2.5668449197860963,
      "grad_norm": 23.342009758498257,
      "learning_rate": 5.097336799988067e-08,
      "logits/chosen": -2.2076547145843506,
      "logits/rejected": -2.154020309448242,
      "logps/chosen": -265.31988525390625,
      "logps/rejected": -231.6819305419922,
      "loss": 0.1448,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.2512500286102295,
      "rewards/margins": 4.000115394592285,
      "rewards/rejected": -2.7488656044006348,
      "step": 240
    },
    {
      "epoch": 2.620320855614973,
      "grad_norm": 19.548359496095816,
      "learning_rate": 3.8902650554212826e-08,
      "logits/chosen": -2.1908040046691895,
      "logits/rejected": -2.155940532684326,
      "logps/chosen": -244.70938110351562,
      "logps/rejected": -229.6440887451172,
      "loss": 0.1683,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.8019378781318665,
      "rewards/margins": 3.827380418777466,
      "rewards/rejected": -3.025442600250244,
      "step": 245
    },
    {
      "epoch": 2.6737967914438503,
      "grad_norm": 19.711354702595525,
      "learning_rate": 2.8403758896638707e-08,
      "logits/chosen": -2.18560791015625,
      "logits/rejected": -2.1317005157470703,
      "logps/chosen": -240.10330200195312,
      "logps/rejected": -229.8534698486328,
      "loss": 0.1412,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 0.7915533185005188,
      "rewards/margins": 4.206446647644043,
      "rewards/rejected": -3.414893388748169,
      "step": 250
    },
    {
      "epoch": 2.6737967914438503,
      "eval_logits/chosen": -2.2052550315856934,
      "eval_logits/rejected": -2.1488451957702637,
      "eval_logps/chosen": -248.0778045654297,
      "eval_logps/rejected": -244.66810607910156,
      "eval_loss": 0.5014069676399231,
      "eval_rewards/accuracies": 0.8214285969734192,
      "eval_rewards/chosen": 0.2960735559463501,
      "eval_rewards/margins": 2.876296281814575,
      "eval_rewards/rejected": -2.5802230834960938,
      "eval_runtime": 180.0422,
      "eval_samples_per_second": 14.774,
      "eval_steps_per_second": 0.233,
      "step": 250
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 11.553793242094493,
      "learning_rate": 1.951248249476961e-08,
      "logits/chosen": -2.1934921741485596,
      "logits/rejected": -2.1499221324920654,
      "logps/chosen": -241.71261596679688,
      "logps/rejected": -252.1171875,
      "loss": 0.1516,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.3082354068756104,
      "rewards/margins": 4.448694705963135,
      "rewards/rejected": -3.1404590606689453,
      "step": 255
    },
    {
      "epoch": 2.7807486631016043,
      "grad_norm": 11.957978779660227,
      "learning_rate": 1.2259130647833626e-08,
      "logits/chosen": -2.1855366230010986,
      "logits/rejected": -2.1587939262390137,
      "logps/chosen": -223.84475708007812,
      "logps/rejected": -237.4518585205078,
      "loss": 0.1438,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.0836931467056274,
      "rewards/margins": 4.059792518615723,
      "rewards/rejected": -2.9760990142822266,
      "step": 260
    },
    {
      "epoch": 2.834224598930481,
      "grad_norm": 18.617117203727886,
      "learning_rate": 6.668429165893996e-09,
      "logits/chosen": -2.1889359951019287,
      "logits/rejected": -2.160160541534424,
      "logps/chosen": -240.87222290039062,
      "logps/rejected": -270.87615966796875,
      "loss": 0.1605,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.1605539321899414,
      "rewards/margins": 4.490550994873047,
      "rewards/rejected": -3.3299968242645264,
      "step": 265
    },
    {
      "epoch": 2.8877005347593583,
      "grad_norm": 16.02689723420078,
      "learning_rate": 2.759436082516664e-09,
      "logits/chosen": -2.205754041671753,
      "logits/rejected": -2.1528849601745605,
      "logps/chosen": -226.815185546875,
      "logps/rejected": -249.47933959960938,
      "loss": 0.1511,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 0.9036574363708496,
      "rewards/margins": 4.245236873626709,
      "rewards/rejected": -3.3415794372558594,
      "step": 270
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 15.563348909167164,
      "learning_rate": 5.454766882097006e-10,
      "logits/chosen": -2.2089016437530518,
      "logits/rejected": -2.166327714920044,
      "logps/chosen": -245.7669677734375,
      "logps/rejected": -246.909912109375,
      "loss": 0.1517,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 1.596333622932434,
      "rewards/margins": 4.828945159912109,
      "rewards/rejected": -3.2326114177703857,
      "step": 275
    },
    {
      "epoch": 2.983957219251337,
      "step": 279,
      "total_flos": 3289753017384960.0,
      "train_loss": 0.3353286673945765,
      "train_runtime": 10412.0978,
      "train_samples_per_second": 6.896,
      "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
}