{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 50,
  "global_step": 436,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022935779816513763,
      "grad_norm": 5.604829014017037,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.619293689727783,
      "logits/rejected": -2.5524563789367676,
      "logps/chosen": -265.4638977050781,
      "logps/rejected": -236.136962890625,
      "loss": 0.1942,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": -0.00018078314315062016,
      "rewards/margins": -0.00011108322360087186,
      "rewards/rejected": -6.969989772187546e-05,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 5.250583616242791,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.6577353477478027,
      "logits/rejected": -2.575864315032959,
      "logps/chosen": -298.8517150878906,
      "logps/rejected": -274.293212890625,
      "loss": 0.1959,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.0006643417873419821,
      "rewards/margins": 0.0011681303149089217,
      "rewards/rejected": -0.0018324721604585648,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 4.980077918182304,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.675644636154175,
      "logits/rejected": -2.6021487712860107,
      "logps/chosen": -290.4170227050781,
      "logps/rejected": -234.3779296875,
      "loss": 0.1923,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.009157607331871986,
      "rewards/margins": 0.013118197210133076,
      "rewards/rejected": -0.003960589878261089,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 5.215480488397355,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.660599946975708,
      "logits/rejected": -2.6108055114746094,
      "logps/chosen": -281.02667236328125,
      "logps/rejected": -267.7845764160156,
      "loss": 0.1835,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.04142485558986664,
      "rewards/margins": 0.04134037345647812,
      "rewards/rejected": 8.447784784948453e-05,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 5.6967811731593905,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.619861602783203,
      "logits/rejected": -2.6124091148376465,
      "logps/chosen": -293.50762939453125,
      "logps/rejected": -303.9714660644531,
      "loss": 0.1732,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": 0.027435380965471268,
      "rewards/margins": 0.0753081664443016,
      "rewards/rejected": -0.04787278175354004,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.5723471641540527,
      "eval_logits/rejected": -2.49267315864563,
      "eval_logps/chosen": -283.6944580078125,
      "eval_logps/rejected": -257.6913146972656,
      "eval_loss": 0.1639617681503296,
      "eval_rewards/accuracies": 0.7068965435028076,
      "eval_rewards/chosen": 0.013956375420093536,
      "eval_rewards/margins": 0.12564486265182495,
      "eval_rewards/rejected": -0.11168847978115082,
      "eval_runtime": 95.9923,
      "eval_samples_per_second": 18.939,
      "eval_steps_per_second": 0.302,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 7.248589740263705,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.5621228218078613,
      "logits/rejected": -2.503438949584961,
      "logps/chosen": -293.131103515625,
      "logps/rejected": -272.94793701171875,
      "loss": 0.1592,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.03646404668688774,
      "rewards/margins": 0.11744984239339828,
      "rewards/rejected": -0.15391388535499573,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 7.653772817525038,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.4572176933288574,
      "logits/rejected": -2.407254219055176,
      "logps/chosen": -327.6614685058594,
      "logps/rejected": -284.7535705566406,
      "loss": 0.1557,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.0009046800551004708,
      "rewards/margins": 0.23255476355552673,
      "rewards/rejected": -0.23345942795276642,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 8.127224192262021,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.3246560096740723,
      "logits/rejected": -2.2035281658172607,
      "logps/chosen": -297.11651611328125,
      "logps/rejected": -290.16302490234375,
      "loss": 0.1475,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.11266676336526871,
      "rewards/margins": 0.2624030113220215,
      "rewards/rejected": -0.3750697076320648,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 8.777001707426516,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -1.9738223552703857,
      "logits/rejected": -1.9115241765975952,
      "logps/chosen": -299.46820068359375,
      "logps/rejected": -291.00994873046875,
      "loss": 0.1449,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.11874761432409286,
      "rewards/margins": 0.3233392536640167,
      "rewards/rejected": -0.4420868456363678,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 9.678897311244762,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -1.0515625476837158,
      "logits/rejected": -1.0063254833221436,
      "logps/chosen": -311.99945068359375,
      "logps/rejected": -342.3149719238281,
      "loss": 0.1412,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.39574748277664185,
      "rewards/margins": 0.3081180453300476,
      "rewards/rejected": -0.7038655281066895,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -1.1028852462768555,
      "eval_logits/rejected": -0.8337642550468445,
      "eval_logps/chosen": -341.0970458984375,
      "eval_logps/rejected": -340.4026794433594,
      "eval_loss": 0.13921599090099335,
      "eval_rewards/accuracies": 0.7198275923728943,
      "eval_rewards/chosen": -0.5600696802139282,
      "eval_rewards/margins": 0.3787323534488678,
      "eval_rewards/rejected": -0.9388020038604736,
      "eval_runtime": 95.7518,
      "eval_samples_per_second": 18.987,
      "eval_steps_per_second": 0.303,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 8.075119669591446,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -1.309912085533142,
      "logits/rejected": -1.0928936004638672,
      "logps/chosen": -323.1109313964844,
      "logps/rejected": -320.5851135253906,
      "loss": 0.1371,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.5792280435562134,
      "rewards/margins": 0.2892912030220032,
      "rewards/rejected": -0.868519127368927,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 11.527192184915142,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": -1.2554824352264404,
      "logits/rejected": -1.2111155986785889,
      "logps/chosen": -315.65582275390625,
      "logps/rejected": -323.9941101074219,
      "loss": 0.1347,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.39836472272872925,
      "rewards/margins": 0.3269372582435608,
      "rewards/rejected": -0.7253020405769348,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 10.253713539855786,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": -0.9448979496955872,
      "logits/rejected": -0.6800575256347656,
      "logps/chosen": -294.2066345214844,
      "logps/rejected": -319.7179870605469,
      "loss": 0.1367,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.38026565313339233,
      "rewards/margins": 0.39078813791275024,
      "rewards/rejected": -0.7710537910461426,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 11.166916308293251,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": -0.8638193011283875,
      "logits/rejected": -0.5810004472732544,
      "logps/chosen": -307.26165771484375,
      "logps/rejected": -287.42547607421875,
      "loss": 0.134,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.3107650578022003,
      "rewards/margins": 0.3205246925354004,
      "rewards/rejected": -0.6312897801399231,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 10.491465778765697,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": -0.785676121711731,
      "logits/rejected": -0.4156356453895569,
      "logps/chosen": -306.7540588378906,
      "logps/rejected": -312.4991149902344,
      "loss": 0.1327,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.40210476517677307,
      "rewards/margins": 0.39101067185401917,
      "rewards/rejected": -0.793115496635437,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": -0.934603214263916,
      "eval_logits/rejected": -0.4213140308856964,
      "eval_logps/chosen": -321.6920166015625,
      "eval_logps/rejected": -333.3775939941406,
      "eval_loss": 0.1256996989250183,
      "eval_rewards/accuracies": 0.7241379022598267,
      "eval_rewards/chosen": -0.36601918935775757,
      "eval_rewards/margins": 0.5025321841239929,
      "eval_rewards/rejected": -0.8685513734817505,
      "eval_runtime": 95.7427,
      "eval_samples_per_second": 18.988,
      "eval_steps_per_second": 0.303,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 10.606572226202415,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": -0.6937485933303833,
      "logits/rejected": -0.39156150817871094,
      "logps/chosen": -341.2958068847656,
      "logps/rejected": -376.64788818359375,
      "loss": 0.1226,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.4474472105503082,
      "rewards/margins": 0.4593452513217926,
      "rewards/rejected": -0.9067924618721008,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 7.909933445936232,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": -0.5848881602287292,
      "logits/rejected": -0.295288622379303,
      "logps/chosen": -314.4745178222656,
      "logps/rejected": -355.83209228515625,
      "loss": 0.1194,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.42873620986938477,
      "rewards/margins": 0.4288958013057709,
      "rewards/rejected": -0.857632040977478,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 9.296204512324422,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": -0.4405382573604584,
      "logits/rejected": 0.17199954390525818,
      "logps/chosen": -351.5772399902344,
      "logps/rejected": -332.4178771972656,
      "loss": 0.1256,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.48219355940818787,
      "rewards/margins": 0.4570356011390686,
      "rewards/rejected": -0.9392291903495789,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 9.788859469537575,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": -0.9616500735282898,
      "logits/rejected": -0.5780771970748901,
      "logps/chosen": -284.89788818359375,
      "logps/rejected": -339.87139892578125,
      "loss": 0.124,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.2678338885307312,
      "rewards/margins": 0.487435519695282,
      "rewards/rejected": -0.7552694082260132,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 8.31792118291438,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": -0.9191502332687378,
      "logits/rejected": -0.3312894105911255,
      "logps/chosen": -313.3896179199219,
      "logps/rejected": -324.57000732421875,
      "loss": 0.1242,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.3510800302028656,
      "rewards/margins": 0.4560603201389313,
      "rewards/rejected": -0.8071403503417969,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": -0.9084411859512329,
      "eval_logits/rejected": -0.1497526615858078,
      "eval_logps/chosen": -319.3252258300781,
      "eval_logps/rejected": -331.5806884765625,
      "eval_loss": 0.11925708502531052,
      "eval_rewards/accuracies": 0.7456896305084229,
      "eval_rewards/chosen": -0.34235113859176636,
      "eval_rewards/margins": 0.5082312226295471,
      "eval_rewards/rejected": -0.8505823612213135,
      "eval_runtime": 93.7494,
      "eval_samples_per_second": 19.392,
      "eval_steps_per_second": 0.309,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 7.584339731807936,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": -0.6695399284362793,
      "logits/rejected": -0.2294241189956665,
      "logps/chosen": -311.20343017578125,
      "logps/rejected": -338.2936096191406,
      "loss": 0.1173,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.3789902925491333,
      "rewards/margins": 0.4605233073234558,
      "rewards/rejected": -0.8395135998725891,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 10.1245037946263,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": -0.6778780221939087,
      "logits/rejected": 0.15124213695526123,
      "logps/chosen": -365.30010986328125,
      "logps/rejected": -385.3759460449219,
      "loss": 0.1146,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.42619770765304565,
      "rewards/margins": 0.5778595209121704,
      "rewards/rejected": -1.0040571689605713,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 8.611834221562093,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": -0.5495241284370422,
      "logits/rejected": -0.10778895765542984,
      "logps/chosen": -308.42864990234375,
      "logps/rejected": -350.29937744140625,
      "loss": 0.1211,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.4048517346382141,
      "rewards/margins": 0.4419426918029785,
      "rewards/rejected": -0.8467944264411926,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 8.84744756928296,
      "learning_rate": 2.5e-07,
      "logits/chosen": -0.8606957197189331,
      "logits/rejected": -0.25019243359565735,
      "logps/chosen": -323.67840576171875,
      "logps/rejected": -341.2455749511719,
      "loss": 0.1198,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.3614657521247864,
      "rewards/margins": 0.4374900758266449,
      "rewards/rejected": -0.7989557981491089,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 8.762072273295908,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": -0.8679742813110352,
      "logits/rejected": -0.18818725645542145,
      "logps/chosen": -340.2317810058594,
      "logps/rejected": -349.81414794921875,
      "loss": 0.118,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.3643796145915985,
      "rewards/margins": 0.5293234586715698,
      "rewards/rejected": -0.8937031030654907,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": -0.8147245645523071,
      "eval_logits/rejected": -0.025059914216399193,
      "eval_logps/chosen": -322.4342041015625,
      "eval_logps/rejected": -343.3682556152344,
      "eval_loss": 0.11674168705940247,
      "eval_rewards/accuracies": 0.7543103694915771,
      "eval_rewards/chosen": -0.3734411895275116,
      "eval_rewards/margins": 0.5950165390968323,
      "eval_rewards/rejected": -0.9684576988220215,
      "eval_runtime": 95.6284,
      "eval_samples_per_second": 19.011,
      "eval_steps_per_second": 0.303,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 8.055173398595795,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": -0.5792064666748047,
      "logits/rejected": -0.028709029778838158,
      "logps/chosen": -319.5377197265625,
      "logps/rejected": -383.94610595703125,
      "loss": 0.1186,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.4882320463657379,
      "rewards/margins": 0.6072918176651001,
      "rewards/rejected": -1.0955239534378052,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 9.235774055264796,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": -0.37728264927864075,
      "logits/rejected": 0.017757948487997055,
      "logps/chosen": -318.92266845703125,
      "logps/rejected": -360.00250244140625,
      "loss": 0.1169,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.5286036729812622,
      "rewards/margins": 0.5529505014419556,
      "rewards/rejected": -1.0815541744232178,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 9.286061693337006,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": -0.7753764986991882,
      "logits/rejected": -0.038079630583524704,
      "logps/chosen": -314.3120422363281,
      "logps/rejected": -365.4776611328125,
      "loss": 0.1216,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.4131316542625427,
      "rewards/margins": 0.6036869883537292,
      "rewards/rejected": -1.016818642616272,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 10.391059671359317,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": -0.6743412017822266,
      "logits/rejected": -0.05127597600221634,
      "logps/chosen": -326.2575378417969,
      "logps/rejected": -345.9309387207031,
      "loss": 0.1144,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.3852667212486267,
      "rewards/margins": 0.5348750352859497,
      "rewards/rejected": -0.9201416969299316,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 11.142166437672913,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": -0.6289922595024109,
      "logits/rejected": -0.2742362916469574,
      "logps/chosen": -350.96099853515625,
      "logps/rejected": -345.9538879394531,
      "loss": 0.1236,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.45011693239212036,
      "rewards/margins": 0.42082324624061584,
      "rewards/rejected": -0.8709400296211243,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": -0.5284780263900757,
      "eval_logits/rejected": 0.2890760600566864,
      "eval_logps/chosen": -332.8935546875,
      "eval_logps/rejected": -351.01617431640625,
      "eval_loss": 0.11446455121040344,
      "eval_rewards/accuracies": 0.7629310488700867,
      "eval_rewards/chosen": -0.4780345857143402,
      "eval_rewards/margins": 0.5669029355049133,
      "eval_rewards/rejected": -1.0449376106262207,
      "eval_runtime": 94.2008,
      "eval_samples_per_second": 19.299,
      "eval_steps_per_second": 0.308,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 9.249175988688211,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": -0.5085445642471313,
      "logits/rejected": 0.12936155498027802,
      "logps/chosen": -332.48468017578125,
      "logps/rejected": -366.32489013671875,
      "loss": 0.1229,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.5451949238777161,
      "rewards/margins": 0.4535568654537201,
      "rewards/rejected": -0.9987518191337585,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 9.319929398035963,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": -0.37469834089279175,
      "logits/rejected": 0.34990087151527405,
      "logps/chosen": -323.309326171875,
      "logps/rejected": -343.5829772949219,
      "loss": 0.1184,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.49077337980270386,
      "rewards/margins": 0.6001543402671814,
      "rewards/rejected": -1.0909278392791748,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 10.87259656283017,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": -0.20644304156303406,
      "logits/rejected": 0.2470076084136963,
      "logps/chosen": -329.26171875,
      "logps/rejected": -346.26666259765625,
      "loss": 0.1182,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.5473949909210205,
      "rewards/margins": 0.44427090883255005,
      "rewards/rejected": -0.9916658401489258,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 11.502524720737265,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": -0.511077344417572,
      "logits/rejected": -0.07181791216135025,
      "logps/chosen": -338.0530700683594,
      "logps/rejected": -354.2392578125,
      "loss": 0.1129,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.5006048679351807,
      "rewards/margins": 0.4419478476047516,
      "rewards/rejected": -0.9425527453422546,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 10.417951906388708,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": -0.4384083151817322,
      "logits/rejected": 0.22936472296714783,
      "logps/chosen": -327.1330261230469,
      "logps/rejected": -339.122314453125,
      "loss": 0.122,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.526419460773468,
      "rewards/margins": 0.48818930983543396,
      "rewards/rejected": -1.0146087408065796,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": -0.48056575655937195,
      "eval_logits/rejected": 0.33971619606018066,
      "eval_logps/chosen": -330.8570251464844,
      "eval_logps/rejected": -350.28302001953125,
      "eval_loss": 0.11267489194869995,
      "eval_rewards/accuracies": 0.767241358757019,
      "eval_rewards/chosen": -0.45766934752464294,
      "eval_rewards/margins": 0.5799364447593689,
      "eval_rewards/rejected": -1.0376057624816895,
      "eval_runtime": 94.5966,
      "eval_samples_per_second": 19.218,
      "eval_steps_per_second": 0.307,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 8.120253834632708,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": -0.412276029586792,
      "logits/rejected": 0.03293975815176964,
      "logps/chosen": -295.88397216796875,
      "logps/rejected": -349.67828369140625,
      "loss": 0.1126,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.43563956022262573,
      "rewards/margins": 0.47577181458473206,
      "rewards/rejected": -0.9114114046096802,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 8.937083414679394,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": -0.7006546258926392,
      "logits/rejected": -0.1087837815284729,
      "logps/chosen": -344.0120544433594,
      "logps/rejected": -357.1898498535156,
      "loss": 0.1124,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.4442760944366455,
      "rewards/margins": 0.5018762350082397,
      "rewards/rejected": -0.9461522102355957,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 9.250415689561605,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": -0.6538180112838745,
      "logits/rejected": 0.11514680087566376,
      "logps/chosen": -334.7962951660156,
      "logps/rejected": -353.5735778808594,
      "loss": 0.1165,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.3841826617717743,
      "rewards/margins": 0.5749909281730652,
      "rewards/rejected": -0.9591735601425171,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 9.532136069213038,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": -0.4097678065299988,
      "logits/rejected": 0.13332059979438782,
      "logps/chosen": -294.68280029296875,
      "logps/rejected": -319.02825927734375,
      "loss": 0.122,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.4272535443305969,
      "rewards/margins": 0.4699665606021881,
      "rewards/rejected": -0.8972201347351074,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 9.826819983657527,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": -0.3362836241722107,
      "logits/rejected": 0.09532846510410309,
      "logps/chosen": -262.51324462890625,
      "logps/rejected": -315.52349853515625,
      "loss": 0.12,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.37953639030456543,
      "rewards/margins": 0.48354291915893555,
      "rewards/rejected": -0.8630792498588562,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": -0.6159119606018066,
      "eval_logits/rejected": 0.19038529694080353,
      "eval_logps/chosen": -325.2745361328125,
      "eval_logps/rejected": -342.9399108886719,
      "eval_loss": 0.11256735771894455,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": -0.40184491872787476,
      "eval_rewards/margins": 0.5623298287391663,
      "eval_rewards/rejected": -0.9641746282577515,
      "eval_runtime": 93.497,
      "eval_samples_per_second": 19.444,
      "eval_steps_per_second": 0.31,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 9.708164546519836,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": -0.34120887517929077,
      "logits/rejected": 0.403463214635849,
      "logps/chosen": -314.7442321777344,
      "logps/rejected": -337.3968811035156,
      "loss": 0.1168,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.43792352080345154,
      "rewards/margins": 0.5555781126022339,
      "rewards/rejected": -0.9935015439987183,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 10.889878191306106,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": -0.31919345259666443,
      "logits/rejected": 0.3534119725227356,
      "logps/chosen": -306.3693542480469,
      "logps/rejected": -349.92462158203125,
      "loss": 0.1147,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.4693039059638977,
      "rewards/margins": 0.5711686015129089,
      "rewards/rejected": -1.040472388267517,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 9.153719920278654,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": -0.556784451007843,
      "logits/rejected": 0.2818067669868469,
      "logps/chosen": -345.05450439453125,
      "logps/rejected": -335.64605712890625,
      "loss": 0.1155,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.41879063844680786,
      "rewards/margins": 0.49730420112609863,
      "rewards/rejected": -0.9160947799682617,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.13197535372108493,
      "train_runtime": 11523.0815,
      "train_samples_per_second": 4.839,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 10,
  "max_steps": 436,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}