{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 355,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 1.6023182644205256,
      "learning_rate": 1.3888888888888888e-07,
      "logits/chosen": -2.861618995666504,
      "logits/rejected": -2.8205904960632324,
      "logps/chosen": -271.06011962890625,
      "logps/rejected": -211.1704559326172,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/margins_max": 0.0,
      "rewards/margins_min": 0.0,
      "rewards/margins_std": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.801703666985174,
      "learning_rate": 1.3888888888888892e-06,
      "logits/chosen": -2.833193778991699,
      "logits/rejected": -2.7905588150024414,
      "logps/chosen": -324.9515380859375,
      "logps/rejected": -274.97601318359375,
      "loss": 0.6927,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.0009881681762635708,
      "rewards/margins": 0.001091944519430399,
      "rewards/margins_max": 0.005423115566372871,
      "rewards/margins_min": -0.0020756451413035393,
      "rewards/margins_std": 0.0033777032513171434,
      "rewards/rejected": -0.00010377627768320963,
      "step": 10
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.8295784413634215,
      "learning_rate": 2.7777777777777783e-06,
      "logits/chosen": -2.7249505519866943,
      "logits/rejected": -2.7066354751586914,
      "logps/chosen": -293.10626220703125,
      "logps/rejected": -215.7418670654297,
      "loss": 0.6894,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.007196377031505108,
      "rewards/margins": 0.0068252841010689735,
      "rewards/margins_max": 0.015210196375846863,
      "rewards/margins_min": -0.0003913335967808962,
      "rewards/margins_std": 0.006930469069629908,
      "rewards/rejected": 0.0003710930759552866,
      "step": 20
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.0991251258429813,
      "learning_rate": 4.166666666666667e-06,
      "logits/chosen": -2.816267490386963,
      "logits/rejected": -2.7468106746673584,
      "logps/chosen": -300.4118347167969,
      "logps/rejected": -232.51480102539062,
      "loss": 0.6751,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.03294036164879799,
      "rewards/margins": 0.03294597938656807,
      "rewards/margins_max": 0.06541319936513901,
      "rewards/margins_min": 0.006184346973896027,
      "rewards/margins_std": 0.02764921821653843,
      "rewards/rejected": -5.6155026868509594e-06,
      "step": 30
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.7320040744134495,
      "learning_rate": 4.998060489154965e-06,
      "logits/chosen": -2.8301820755004883,
      "logits/rejected": -2.751152753829956,
      "logps/chosen": -271.0854187011719,
      "logps/rejected": -225.5413360595703,
      "loss": 0.6584,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.06399233639240265,
      "rewards/margins": 0.0620587058365345,
      "rewards/margins_max": 0.128224715590477,
      "rewards/margins_min": 0.005723648704588413,
      "rewards/margins_std": 0.05614202097058296,
      "rewards/rejected": 0.0019336309051141143,
      "step": 40
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.9286136496428266,
      "learning_rate": 4.976275538042932e-06,
      "logits/chosen": -2.7784149646759033,
      "logits/rejected": -2.7074830532073975,
      "logps/chosen": -263.76910400390625,
      "logps/rejected": -236.1763458251953,
      "loss": 0.6279,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.1180100291967392,
      "rewards/margins": 0.14100034534931183,
      "rewards/margins_max": 0.2934209704399109,
      "rewards/margins_min": 0.029913101345300674,
      "rewards/margins_std": 0.12193763256072998,
      "rewards/rejected": -0.022990306839346886,
      "step": 50
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.19542131799087,
      "learning_rate": 4.93049306999712e-06,
      "logits/chosen": -2.6885108947753906,
      "logits/rejected": -2.653852939605713,
      "logps/chosen": -303.560791015625,
      "logps/rejected": -276.5751647949219,
      "loss": 0.5911,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.10078845918178558,
      "rewards/margins": 0.23882508277893066,
      "rewards/margins_max": 0.41835322976112366,
      "rewards/margins_min": 0.06585223972797394,
      "rewards/margins_std": 0.1579587757587433,
      "rewards/rejected": -0.1380366086959839,
      "step": 60
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.8726629315170749,
      "learning_rate": 4.861156761634014e-06,
      "logits/chosen": -2.6910977363586426,
      "logits/rejected": -2.637085437774658,
      "logps/chosen": -308.81890869140625,
      "logps/rejected": -251.6500701904297,
      "loss": 0.5567,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.14042340219020844,
      "rewards/margins": 0.3005082309246063,
      "rewards/margins_max": 0.6100845336914062,
      "rewards/margins_min": 0.049360163509845734,
      "rewards/margins_std": 0.25483688712120056,
      "rewards/rejected": -0.16008484363555908,
      "step": 70
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.2506013460624406,
      "learning_rate": 4.7689385491773934e-06,
      "logits/chosen": -2.6962907314300537,
      "logits/rejected": -2.648644208908081,
      "logps/chosen": -310.2759704589844,
      "logps/rejected": -320.2915344238281,
      "loss": 0.5126,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.1099453717470169,
      "rewards/margins": 0.4490734934806824,
      "rewards/margins_max": 0.8111162185668945,
      "rewards/margins_min": 0.09584008157253265,
      "rewards/margins_std": 0.3334144353866577,
      "rewards/rejected": -0.3391281068325043,
      "step": 80
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.7658324125158935,
      "learning_rate": 4.654732116743193e-06,
      "logits/chosen": -2.5777134895324707,
      "logits/rejected": -2.555579423904419,
      "logps/chosen": -265.8702697753906,
      "logps/rejected": -246.2513427734375,
      "loss": 0.456,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.10124516487121582,
      "rewards/margins": 0.5738865733146667,
      "rewards/margins_max": 0.9613016843795776,
      "rewards/margins_min": 0.21029099822044373,
      "rewards/margins_std": 0.3518252670764923,
      "rewards/rejected": -0.4726414680480957,
      "step": 90
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.900318438715333,
      "learning_rate": 4.5196442356717526e-06,
      "logits/chosen": -2.587555408477783,
      "logits/rejected": -2.5667903423309326,
      "logps/chosen": -279.7422790527344,
      "logps/rejected": -321.7631530761719,
      "loss": 0.4419,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.01669424958527088,
      "rewards/margins": 0.6124511957168579,
      "rewards/margins_max": 1.1129316091537476,
      "rewards/margins_min": 0.16525402665138245,
      "rewards/margins_std": 0.43622416257858276,
      "rewards/rejected": -0.5957569479942322,
      "step": 100
    },
    {
      "epoch": 0.28,
      "eval_logits/chosen": -2.5718588829040527,
      "eval_logits/rejected": -2.544903039932251,
      "eval_logps/chosen": -329.8811340332031,
      "eval_logps/rejected": -318.3518371582031,
      "eval_loss": 0.6649842262268066,
      "eval_rewards/accuracies": 0.6240000128746033,
      "eval_rewards/chosen": -0.4528772234916687,
      "eval_rewards/margins": 0.14485196769237518,
      "eval_rewards/margins_max": 0.8618865609169006,
      "eval_rewards/margins_min": -0.5417762994766235,
      "eval_rewards/margins_std": 0.46923625469207764,
      "eval_rewards/rejected": -0.5977292656898499,
      "eval_runtime": 428.7091,
      "eval_samples_per_second": 4.665,
      "eval_steps_per_second": 0.292,
      "step": 100
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.5447046302664655,
      "learning_rate": 4.364984038837727e-06,
      "logits/chosen": -2.6017634868621826,
      "logits/rejected": -2.543693780899048,
      "logps/chosen": -371.7081604003906,
      "logps/rejected": -368.0250549316406,
      "loss": 0.3877,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.028975462540984154,
      "rewards/margins": 0.844887912273407,
      "rewards/margins_max": 1.4032989740371704,
      "rewards/margins_min": 0.22432711720466614,
      "rewards/margins_std": 0.5291948914527893,
      "rewards/rejected": -0.8159124255180359,
      "step": 110
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.8541926547535676,
      "learning_rate": 4.192250333880045e-06,
      "logits/chosen": -2.566699504852295,
      "logits/rejected": -2.523707866668701,
      "logps/chosen": -360.9166564941406,
      "logps/rejected": -359.21478271484375,
      "loss": 0.393,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.12530414760112762,
      "rewards/margins": 0.8049014210700989,
      "rewards/margins_max": 1.5037436485290527,
      "rewards/margins_min": 0.22369906306266785,
      "rewards/margins_std": 0.5746256709098816,
      "rewards/rejected": -0.9302055239677429,
      "step": 120
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.676765116222706,
      "learning_rate": 4.0031170782990214e-06,
      "logits/chosen": -2.488229751586914,
      "logits/rejected": -2.4302055835723877,
      "logps/chosen": -412.65087890625,
      "logps/rejected": -429.1094665527344,
      "loss": 0.3581,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.3028665781021118,
      "rewards/margins": 0.9839191436767578,
      "rewards/margins_max": 1.651447057723999,
      "rewards/margins_min": 0.29168885946273804,
      "rewards/margins_std": 0.603247344493866,
      "rewards/rejected": -1.2867857217788696,
      "step": 130
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.271231871514864,
      "learning_rate": 3.7994171571810756e-06,
      "logits/chosen": -2.3792262077331543,
      "logits/rejected": -2.348208427429199,
      "logps/chosen": -337.531005859375,
      "logps/rejected": -398.3355407714844,
      "loss": 0.3623,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.19116966426372528,
      "rewards/margins": 1.0936906337738037,
      "rewards/margins_max": 1.9557828903198242,
      "rewards/margins_min": 0.2191341668367386,
      "rewards/margins_std": 0.7831128835678101,
      "rewards/rejected": -1.284860372543335,
      "step": 140
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.267090889931009,
      "learning_rate": 3.5831246207606597e-06,
      "logits/chosen": -2.244743585586548,
      "logits/rejected": -2.207181453704834,
      "logps/chosen": -340.91412353515625,
      "logps/rejected": -383.2621765136719,
      "loss": 0.3039,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.5222915410995483,
      "rewards/margins": 1.1795024871826172,
      "rewards/margins_max": 2.02280855178833,
      "rewards/margins_min": 0.273987352848053,
      "rewards/margins_std": 0.7898725271224976,
      "rewards/rejected": -1.701793909072876,
      "step": 150
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.800242307139694,
      "learning_rate": 3.3563355539546795e-06,
      "logits/chosen": -2.081348419189453,
      "logits/rejected": -2.017423629760742,
      "logps/chosen": -342.2528991699219,
      "logps/rejected": -432.951171875,
      "loss": 0.2389,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.41706037521362305,
      "rewards/margins": 1.5685182809829712,
      "rewards/margins_max": 2.661461114883423,
      "rewards/margins_min": 0.5422362685203552,
      "rewards/margins_std": 0.971255898475647,
      "rewards/rejected": -1.9855785369873047,
      "step": 160
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.002190159330027,
      "learning_rate": 3.121247763262235e-06,
      "logits/chosen": -2.031193971633911,
      "logits/rejected": -1.9694950580596924,
      "logps/chosen": -359.4656982421875,
      "logps/rejected": -506.41217041015625,
      "loss": 0.2274,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.2984406650066376,
      "rewards/margins": 1.8030246496200562,
      "rewards/margins_max": 2.6862552165985107,
      "rewards/margins_min": 0.6847059726715088,
      "rewards/margins_std": 0.9101365208625793,
      "rewards/rejected": -2.1014654636383057,
      "step": 170
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.495552270047072,
      "learning_rate": 2.8801394778833475e-06,
      "logits/chosen": -1.804452657699585,
      "logits/rejected": -1.6888850927352905,
      "logps/chosen": -424.28533935546875,
      "logps/rejected": -582.659423828125,
      "loss": 0.2208,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.923845112323761,
      "rewards/margins": 1.9609006643295288,
      "rewards/margins_max": 3.038311719894409,
      "rewards/margins_min": 0.9176443219184875,
      "rewards/margins_std": 0.9313791394233704,
      "rewards/rejected": -2.8847455978393555,
      "step": 180
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.26489542539375,
      "learning_rate": 2.6353472714635443e-06,
      "logits/chosen": -1.6563403606414795,
      "logits/rejected": -1.4980707168579102,
      "logps/chosen": -365.2936096191406,
      "logps/rejected": -520.5008544921875,
      "loss": 0.1887,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.4333272874355316,
      "rewards/margins": 2.3968050479888916,
      "rewards/margins_max": 3.5410873889923096,
      "rewards/margins_min": 1.2592780590057373,
      "rewards/margins_std": 1.0371885299682617,
      "rewards/rejected": -2.830132007598877,
      "step": 190
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.247380737778327,
      "learning_rate": 2.3892434184240536e-06,
      "logits/chosen": -1.6041126251220703,
      "logits/rejected": -1.4733774662017822,
      "logps/chosen": -478.11541748046875,
      "logps/rejected": -663.8895874023438,
      "loss": 0.1872,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -1.3858261108398438,
      "rewards/margins": 2.5970957279205322,
      "rewards/margins_max": 4.095944404602051,
      "rewards/margins_min": 0.5976935625076294,
      "rewards/margins_std": 1.6118961572647095,
      "rewards/rejected": -3.982921600341797,
      "step": 200
    },
    {
      "epoch": 0.56,
      "eval_logits/chosen": -1.4771034717559814,
      "eval_logits/rejected": -1.4158743619918823,
      "eval_logps/chosen": -568.5177612304688,
      "eval_logps/rejected": -612.582763671875,
      "eval_loss": 0.7859476804733276,
      "eval_rewards/accuracies": 0.6629999876022339,
      "eval_rewards/chosen": -2.839242935180664,
      "eval_rewards/margins": 0.70079505443573,
      "eval_rewards/margins_max": 3.4374256134033203,
      "eval_rewards/margins_min": -2.182706356048584,
      "eval_rewards/margins_std": 1.9158276319503784,
      "eval_rewards/rejected": -3.5400378704071045,
      "eval_runtime": 428.5233,
      "eval_samples_per_second": 4.667,
      "eval_steps_per_second": 0.292,
      "step": 200
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.190377965258524,
      "learning_rate": 2.1442129043167877e-06,
      "logits/chosen": -1.4293570518493652,
      "logits/rejected": -1.3583998680114746,
      "logps/chosen": -448.363037109375,
      "logps/rejected": -679.0516357421875,
      "loss": 0.1699,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -1.281787395477295,
      "rewards/margins": 2.944481134414673,
      "rewards/margins_max": 4.665500640869141,
      "rewards/margins_min": 0.9343290328979492,
      "rewards/margins_std": 1.711601972579956,
      "rewards/rejected": -4.226268291473389,
      "step": 210
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.464327030535654,
      "learning_rate": 1.9026303129961049e-06,
      "logits/chosen": -1.474830150604248,
      "logits/rejected": -1.3183205127716064,
      "logps/chosen": -591.7376708984375,
      "logps/rejected": -873.0149536132812,
      "loss": 0.1359,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -2.382814407348633,
      "rewards/margins": 3.604060411453247,
      "rewards/margins_max": 4.968972206115723,
      "rewards/margins_min": 1.4786490201950073,
      "rewards/margins_std": 1.6280790567398071,
      "rewards/rejected": -5.986874580383301,
      "step": 220
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.671302383042187,
      "learning_rate": 1.66683681459314e-06,
      "logits/chosen": -1.4693548679351807,
      "logits/rejected": -1.3056604862213135,
      "logps/chosen": -644.6790771484375,
      "logps/rejected": -878.6692504882812,
      "loss": 0.146,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -2.730034351348877,
      "rewards/margins": 3.4357237815856934,
      "rewards/margins_max": 5.224689483642578,
      "rewards/margins_min": 1.111033320426941,
      "rewards/margins_std": 1.8828909397125244,
      "rewards/rejected": -6.165757656097412,
      "step": 230
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.647778822191043,
      "learning_rate": 1.4391174773015836e-06,
      "logits/chosen": -1.4095017910003662,
      "logits/rejected": -1.2933498620986938,
      "logps/chosen": -636.0205688476562,
      "logps/rejected": -929.3019409179688,
      "loss": 0.1482,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -3.1122348308563232,
      "rewards/margins": 3.3490359783172607,
      "rewards/margins_max": 5.252045631408691,
      "rewards/margins_min": 1.3964684009552002,
      "rewards/margins_std": 1.7266426086425781,
      "rewards/rejected": -6.461270809173584,
      "step": 240
    },
    {
      "epoch": 0.7,
      "grad_norm": 29.012737765276803,
      "learning_rate": 1.2216791228457778e-06,
      "logits/chosen": -1.3994617462158203,
      "logits/rejected": -1.2709558010101318,
      "logps/chosen": -621.2999267578125,
      "logps/rejected": -928.4699096679688,
      "loss": 0.1235,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -3.1001884937286377,
      "rewards/margins": 3.6740684509277344,
      "rewards/margins_max": 5.534354209899902,
      "rewards/margins_min": 1.508026361465454,
      "rewards/margins_std": 1.7670818567276,
      "rewards/rejected": -6.774256706237793,
      "step": 250
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.553216346552263,
      "learning_rate": 1.0166289402331391e-06,
      "logits/chosen": -1.4789546728134155,
      "logits/rejected": -1.3316059112548828,
      "logps/chosen": -634.4136962890625,
      "logps/rejected": -945.2862548828125,
      "loss": 0.1438,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -3.4232017993927,
      "rewards/margins": 3.440350294113159,
      "rewards/margins_max": 5.2788987159729,
      "rewards/margins_min": 1.4436607360839844,
      "rewards/margins_std": 1.7658751010894775,
      "rewards/rejected": -6.863552093505859,
      "step": 260
    },
    {
      "epoch": 0.76,
      "grad_norm": 6.13292881557691,
      "learning_rate": 8.259540650444736e-07,
      "logits/chosen": -1.3842976093292236,
      "logits/rejected": -1.286069631576538,
      "logps/chosen": -662.1298217773438,
      "logps/rejected": -982.2066650390625,
      "loss": 0.1621,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -3.530142307281494,
      "rewards/margins": 3.6669273376464844,
      "rewards/margins_max": 5.465144157409668,
      "rewards/margins_min": 1.4210524559020996,
      "rewards/margins_std": 1.8017559051513672,
      "rewards/rejected": -7.197070121765137,
      "step": 270
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.782327714810695,
      "learning_rate": 6.515023221586722e-07,
      "logits/chosen": -1.3697433471679688,
      "logits/rejected": -1.28987455368042,
      "logps/chosen": -655.1025390625,
      "logps/rejected": -1005.9417724609375,
      "loss": 0.1048,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -3.516455888748169,
      "rewards/margins": 3.8169894218444824,
      "rewards/margins_max": 5.726956367492676,
      "rewards/margins_min": 1.5342118740081787,
      "rewards/margins_std": 1.8832015991210938,
      "rewards/rejected": -7.3334455490112305,
      "step": 280
    },
    {
      "epoch": 0.82,
      "grad_norm": 16.94834231932571,
      "learning_rate": 4.949643185335288e-07,
      "logits/chosen": -1.4026738405227661,
      "logits/rejected": -1.3095262050628662,
      "logps/chosen": -593.7080078125,
      "logps/rejected": -906.31298828125,
      "loss": 0.1582,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -3.067917823791504,
      "rewards/margins": 3.355029344558716,
      "rewards/margins_max": 5.543763637542725,
      "rewards/margins_min": 1.00192391872406,
      "rewards/margins_std": 2.0951552391052246,
      "rewards/rejected": -6.422947883605957,
      "step": 290
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.830855192004018,
      "learning_rate": 3.578570595810274e-07,
      "logits/chosen": -1.5070269107818604,
      "logits/rejected": -1.3751003742218018,
      "logps/chosen": -646.901123046875,
      "logps/rejected": -984.0985107421875,
      "loss": 0.1102,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -3.0038809776306152,
      "rewards/margins": 3.9427695274353027,
      "rewards/margins_max": 5.859181880950928,
      "rewards/margins_min": 1.6614525318145752,
      "rewards/margins_std": 1.9274225234985352,
      "rewards/rejected": -6.946650505065918,
      "step": 300
    },
    {
      "epoch": 0.85,
      "eval_logits/chosen": -1.4022948741912842,
      "eval_logits/rejected": -1.3416800498962402,
      "eval_logps/chosen": -773.4503173828125,
      "eval_logps/rejected": -840.8230590820312,
      "eval_loss": 0.8934583067893982,
      "eval_rewards/accuracies": 0.6470000147819519,
      "eval_rewards/chosen": -4.888567924499512,
      "eval_rewards/margins": 0.9338728189468384,
      "eval_rewards/margins_max": 4.508862495422363,
      "eval_rewards/margins_min": -2.803253412246704,
      "eval_rewards/margins_std": 2.4985644817352295,
      "eval_rewards/rejected": -5.822441577911377,
      "eval_runtime": 428.3228,
      "eval_samples_per_second": 4.669,
      "eval_steps_per_second": 0.292,
      "step": 300
    },
    {
      "epoch": 0.87,
      "grad_norm": 25.66825260695654,
      "learning_rate": 2.4150924791035037e-07,
      "logits/chosen": -1.4382137060165405,
      "logits/rejected": -1.2731282711029053,
      "logps/chosen": -688.03857421875,
      "logps/rejected": -948.1687622070312,
      "loss": 0.1452,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -3.9022278785705566,
      "rewards/margins": 3.437220335006714,
      "rewards/margins_max": 5.5707316398620605,
      "rewards/margins_min": 1.006291389465332,
      "rewards/margins_std": 2.0285696983337402,
      "rewards/rejected": -7.339447975158691,
      "step": 310
    },
    {
      "epoch": 0.9,
      "grad_norm": 24.605253871204138,
      "learning_rate": 1.4704840690808658e-07,
      "logits/chosen": -1.4029080867767334,
      "logits/rejected": -1.3075344562530518,
      "logps/chosen": -682.196044921875,
      "logps/rejected": -989.9052734375,
      "loss": 0.1673,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -3.739229202270508,
      "rewards/margins": 3.553121566772461,
      "rewards/margins_max": 5.583710670471191,
      "rewards/margins_min": 0.8181284070014954,
      "rewards/margins_std": 2.200822114944458,
      "rewards/rejected": -7.292350769042969,
      "step": 320
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.663524467768484,
      "learning_rate": 7.538995394063996e-08,
      "logits/chosen": -1.5248098373413086,
      "logits/rejected": -1.3775807619094849,
      "logps/chosen": -683.1748046875,
      "logps/rejected": -983.3424682617188,
      "loss": 0.1405,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -3.2885494232177734,
      "rewards/margins": 3.875479221343994,
      "rewards/margins_max": 5.90716552734375,
      "rewards/margins_min": 1.6211687326431274,
      "rewards/margins_std": 1.9896876811981201,
      "rewards/rejected": -7.164028167724609,
      "step": 330
    },
    {
      "epoch": 0.96,
      "grad_norm": 3.7848904322458856,
      "learning_rate": 2.722832907015971e-08,
      "logits/chosen": -1.389829397201538,
      "logits/rejected": -1.2816789150238037,
      "logps/chosen": -664.4754638671875,
      "logps/rejected": -1003.7318115234375,
      "loss": 0.115,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -3.6462998390197754,
      "rewards/margins": 3.879322052001953,
      "rewards/margins_max": 5.659389019012451,
      "rewards/margins_min": 1.880358338356018,
      "rewards/margins_std": 1.7462043762207031,
      "rewards/rejected": -7.5256218910217285,
      "step": 340
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.242067065610791,
      "learning_rate": 3.030265255329623e-09,
      "logits/chosen": -1.377044439315796,
      "logits/rejected": -1.2859034538269043,
      "logps/chosen": -652.96337890625,
      "logps/rejected": -1036.343505859375,
      "loss": 0.1061,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -3.380042552947998,
      "rewards/margins": 4.118873119354248,
      "rewards/margins_max": 5.629302501678467,
      "rewards/margins_min": 2.122800827026367,
      "rewards/margins_std": 1.6110029220581055,
      "rewards/rejected": -7.4989166259765625,
      "step": 350
    },
    {
      "epoch": 1.0,
      "step": 355,
      "total_flos": 0.0,
      "train_loss": 0.3073710792501208,
      "train_runtime": 4217.5244,
      "train_samples_per_second": 1.346,
      "train_steps_per_second": 0.084
    }
  ],
  "logging_steps": 10,
  "max_steps": 355,
  "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
}