{
  "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": 5.597671583305064,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.6193342208862305,
      "logits/rejected": -2.5525386333465576,
      "logps/chosen": -265.3854064941406,
      "logps/rejected": -236.1589813232422,
      "loss": 0.1941,
      "rewards/accuracies": 0.41874998807907104,
      "rewards/chosen": 0.0006044252077117562,
      "rewards/margins": 0.0008946189773268998,
      "rewards/rejected": -0.000290193798718974,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 5.252146327563864,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.657933473587036,
      "logits/rejected": -2.576028347015381,
      "logps/chosen": -298.8043518066406,
      "logps/rejected": -274.30718994140625,
      "loss": 0.1958,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": -0.00019080772472079843,
      "rewards/margins": 0.0017816700274124742,
      "rewards/rejected": -0.001972477650269866,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 4.974318928321663,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6759729385375977,
      "logits/rejected": -2.6023097038269043,
      "logps/chosen": -290.3868713378906,
      "logps/rejected": -234.3571319580078,
      "loss": 0.1923,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.009458948858082294,
      "rewards/margins": 0.013211230747401714,
      "rewards/rejected": -0.0037522814236581326,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 5.209557025843711,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.659982204437256,
      "logits/rejected": -2.6102194786071777,
      "logps/chosen": -280.9975891113281,
      "logps/rejected": -267.7278747558594,
      "loss": 0.1836,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.04171619936823845,
      "rewards/margins": 0.041064582765102386,
      "rewards/rejected": 0.0006516153225675225,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 5.582751836908906,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.6198437213897705,
      "logits/rejected": -2.6123862266540527,
      "logps/chosen": -293.4254455566406,
      "logps/rejected": -303.90386962890625,
      "loss": 0.1732,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.028257910162210464,
      "rewards/margins": 0.0754549652338028,
      "rewards/rejected": -0.047197047621011734,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.5737149715423584,
      "eval_logits/rejected": -2.494300365447998,
      "eval_logps/chosen": -283.4266662597656,
      "eval_logps/rejected": -257.4446105957031,
      "eval_loss": 0.16405636072158813,
      "eval_rewards/accuracies": 0.7068965435028076,
      "eval_rewards/chosen": 0.016634328290820122,
      "eval_rewards/margins": 0.12585583329200745,
      "eval_rewards/rejected": -0.10922150313854218,
      "eval_runtime": 93.7026,
      "eval_samples_per_second": 19.402,
      "eval_steps_per_second": 0.309,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 7.241200335566244,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.562969207763672,
      "logits/rejected": -2.5053610801696777,
      "logps/chosen": -293.1821594238281,
      "logps/rejected": -272.9414367675781,
      "loss": 0.1593,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.03697499260306358,
      "rewards/margins": 0.11687393486499786,
      "rewards/rejected": -0.15384894609451294,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 7.058665086800562,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.4592204093933105,
      "logits/rejected": -2.4099252223968506,
      "logps/chosen": -327.544189453125,
      "logps/rejected": -284.48760986328125,
      "loss": 0.1564,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.00026793667348101735,
      "rewards/margins": 0.23106786608695984,
      "rewards/rejected": -0.23079994320869446,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 8.05919017424508,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.3360755443573,
      "logits/rejected": -2.2115492820739746,
      "logps/chosen": -298.69482421875,
      "logps/rejected": -292.111328125,
      "loss": 0.1478,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.128449946641922,
      "rewards/margins": 0.26610296964645386,
      "rewards/rejected": -0.39455294609069824,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 8.402102216426163,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -1.9170547723770142,
      "logits/rejected": -1.8550583124160767,
      "logps/chosen": -299.20281982421875,
      "logps/rejected": -289.49163818359375,
      "loss": 0.1441,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.1160941943526268,
      "rewards/margins": 0.3108092248439789,
      "rewards/rejected": -0.4269034266471863,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 8.625504630985954,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -0.48900121450424194,
      "logits/rejected": -0.44961875677108765,
      "logps/chosen": -311.1794738769531,
      "logps/rejected": -339.92254638671875,
      "loss": 0.1412,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.3875480592250824,
      "rewards/margins": 0.2923927903175354,
      "rewards/rejected": -0.6799408197402954,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -0.4184306561946869,
      "eval_logits/rejected": -0.1537734419107437,
      "eval_logps/chosen": -328.5106201171875,
      "eval_logps/rejected": -326.1405944824219,
      "eval_loss": 0.13468901813030243,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -0.4342052936553955,
      "eval_rewards/margins": 0.3619759678840637,
      "eval_rewards/rejected": -0.796181321144104,
      "eval_runtime": 92.0893,
      "eval_samples_per_second": 19.742,
      "eval_steps_per_second": 0.315,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 10.103043851045701,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -0.40221118927001953,
      "logits/rejected": -0.13967064023017883,
      "logps/chosen": -312.624267578125,
      "logps/rejected": -311.3497009277344,
      "loss": 0.1354,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.4743613302707672,
      "rewards/margins": 0.3018040060997009,
      "rewards/rejected": -0.776165246963501,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 10.671011000324334,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": -0.04719700291752815,
      "logits/rejected": 0.09439365565776825,
      "logps/chosen": -324.4268798828125,
      "logps/rejected": -335.16424560546875,
      "loss": 0.1353,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.4860754609107971,
      "rewards/margins": 0.3509276807308197,
      "rewards/rejected": -0.8370031118392944,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 11.348522180816168,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": -0.058751799166202545,
      "logits/rejected": 0.29970529675483704,
      "logps/chosen": -296.40399169921875,
      "logps/rejected": -316.8427429199219,
      "loss": 0.1373,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.4022388458251953,
      "rewards/margins": 0.3400627374649048,
      "rewards/rejected": -0.7423015832901001,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 25.0836527688945,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": -0.29172104597091675,
      "logits/rejected": 0.11182677745819092,
      "logps/chosen": -303.4015197753906,
      "logps/rejected": -286.45526123046875,
      "loss": 0.1334,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.2721634805202484,
      "rewards/margins": 0.34942418336868286,
      "rewards/rejected": -0.6215876340866089,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 11.640337926409853,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": 0.08257939666509628,
      "logits/rejected": 0.5254168510437012,
      "logps/chosen": -308.8121337890625,
      "logps/rejected": -314.32220458984375,
      "loss": 0.1307,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.42268529534339905,
      "rewards/margins": 0.3886614143848419,
      "rewards/rejected": -0.8113466501235962,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": 0.018126631155610085,
      "eval_logits/rejected": 0.7143937349319458,
      "eval_logps/chosen": -320.6210021972656,
      "eval_logps/rejected": -332.353271484375,
      "eval_loss": 0.12613436579704285,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -0.3553086221218109,
      "eval_rewards/margins": 0.5029994249343872,
      "eval_rewards/rejected": -0.8583080768585205,
      "eval_runtime": 91.5757,
      "eval_samples_per_second": 19.852,
      "eval_steps_per_second": 0.317,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 11.56271576551216,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": 0.02932182513177395,
      "logits/rejected": 0.4751170575618744,
      "logps/chosen": -337.37237548828125,
      "logps/rejected": -373.10382080078125,
      "loss": 0.1247,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.40821242332458496,
      "rewards/margins": 0.46313947439193726,
      "rewards/rejected": -0.8713518381118774,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 7.985791463305948,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": 0.06568197906017303,
      "logits/rejected": 0.3852883279323578,
      "logps/chosen": -311.62481689453125,
      "logps/rejected": -353.85443115234375,
      "loss": 0.1191,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.40023931860923767,
      "rewards/margins": 0.4376160204410553,
      "rewards/rejected": -0.837855339050293,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 9.33389793686436,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": 0.15779247879981995,
      "logits/rejected": 0.6506751775741577,
      "logps/chosen": -348.14398193359375,
      "logps/rejected": -325.84637451171875,
      "loss": 0.1278,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.44786080718040466,
      "rewards/margins": 0.4256533682346344,
      "rewards/rejected": -0.8735141754150391,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 10.313716932222244,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": -0.43816322088241577,
      "logits/rejected": 0.023333895951509476,
      "logps/chosen": -283.8222351074219,
      "logps/rejected": -337.9872741699219,
      "loss": 0.1236,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.2570773959159851,
      "rewards/margins": 0.47935089468955994,
      "rewards/rejected": -0.7364283800125122,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 9.248153089741166,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": -0.08292710781097412,
      "logits/rejected": 0.5214661359786987,
      "logps/chosen": -312.56500244140625,
      "logps/rejected": -325.7038269042969,
      "loss": 0.1238,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.34283357858657837,
      "rewards/margins": 0.4756447672843933,
      "rewards/rejected": -0.8184784054756165,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": 0.2969372272491455,
      "eval_logits/rejected": 1.2989132404327393,
      "eval_logps/chosen": -326.1716613769531,
      "eval_logps/rejected": -341.28619384765625,
      "eval_loss": 0.11991516500711441,
      "eval_rewards/accuracies": 0.732758641242981,
      "eval_rewards/chosen": -0.41081586480140686,
      "eval_rewards/margins": 0.5368219614028931,
      "eval_rewards/rejected": -0.9476377367973328,
      "eval_runtime": 92.1824,
      "eval_samples_per_second": 19.722,
      "eval_steps_per_second": 0.315,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 8.486691996298621,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": 0.504795253276825,
      "logits/rejected": 1.0519931316375732,
      "logps/chosen": -317.31756591796875,
      "logps/rejected": -346.00689697265625,
      "loss": 0.1175,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.44013166427612305,
      "rewards/margins": 0.47651535272598267,
      "rewards/rejected": -0.9166469573974609,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 9.587760322945808,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": 0.056784339249134064,
      "logits/rejected": 1.057108998298645,
      "logps/chosen": -358.15313720703125,
      "logps/rejected": -375.32452392578125,
      "loss": 0.1134,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.3547281324863434,
      "rewards/margins": 0.548815131187439,
      "rewards/rejected": -0.9035432934761047,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 9.093354835969592,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": -0.12167753279209137,
      "logits/rejected": 0.5305444002151489,
      "logps/chosen": -298.165283203125,
      "logps/rejected": -341.3959045410156,
      "loss": 0.1206,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.3022175431251526,
      "rewards/margins": 0.4555422365665436,
      "rewards/rejected": -0.7577598094940186,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 9.439667913770359,
      "learning_rate": 2.5e-07,
      "logits/chosen": -0.21128520369529724,
      "logits/rejected": 0.5969688296318054,
      "logps/chosen": -321.5914306640625,
      "logps/rejected": -339.0221862792969,
      "loss": 0.1218,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.34059566259384155,
      "rewards/margins": 0.43612655997276306,
      "rewards/rejected": -0.7767221927642822,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 9.292342727311365,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": -0.020272482186555862,
      "logits/rejected": 0.808585524559021,
      "logps/chosen": -337.55767822265625,
      "logps/rejected": -346.179443359375,
      "loss": 0.1185,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.337638795375824,
      "rewards/margins": 0.5197167992591858,
      "rewards/rejected": -0.8573554754257202,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": -0.1745007336139679,
      "eval_logits/rejected": 0.8515735268592834,
      "eval_logps/chosen": -315.9549560546875,
      "eval_logps/rejected": -335.7632751464844,
      "eval_loss": 0.11658623069524765,
      "eval_rewards/accuracies": 0.7543103694915771,
      "eval_rewards/chosen": -0.3086487948894501,
      "eval_rewards/margins": 0.5837592482566833,
      "eval_rewards/rejected": -0.8924079537391663,
      "eval_runtime": 92.3611,
      "eval_samples_per_second": 19.684,
      "eval_steps_per_second": 0.314,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 8.186794570934321,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": -0.006458556745201349,
      "logits/rejected": 0.6227847933769226,
      "logps/chosen": -308.7506408691406,
      "logps/rejected": -372.9380798339844,
      "loss": 0.1168,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.3803611397743225,
      "rewards/margins": 0.6050828695297241,
      "rewards/rejected": -0.9854438900947571,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 10.337683797583873,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": 0.44111576676368713,
      "logits/rejected": 0.8292443156242371,
      "logps/chosen": -313.4148254394531,
      "logps/rejected": -354.33258056640625,
      "loss": 0.1168,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.47352510690689087,
      "rewards/margins": 0.5513299703598022,
      "rewards/rejected": -1.0248548984527588,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 10.116957119119249,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": 0.1775570809841156,
      "logits/rejected": 1.0465366840362549,
      "logps/chosen": -317.04180908203125,
      "logps/rejected": -370.72479248046875,
      "loss": 0.1221,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.4404297471046448,
      "rewards/margins": 0.6288600564002991,
      "rewards/rejected": -1.0692898035049438,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 9.609746165977107,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": 0.02145857736468315,
      "logits/rejected": 0.823032021522522,
      "logps/chosen": -322.24542236328125,
      "logps/rejected": -342.22503662109375,
      "loss": 0.1156,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.3451458811759949,
      "rewards/margins": 0.5379365086555481,
      "rewards/rejected": -0.883082389831543,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 9.08082569283608,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": -0.18167677521705627,
      "logits/rejected": 0.25138911604881287,
      "logps/chosen": -341.6298828125,
      "logps/rejected": -338.09869384765625,
      "loss": 0.1228,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.3568056523799896,
      "rewards/margins": 0.43558257818222046,
      "rewards/rejected": -0.7923881411552429,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": -0.13157324492931366,
      "eval_logits/rejected": 0.8573818206787109,
      "eval_logps/chosen": -322.0434265136719,
      "eval_logps/rejected": -339.1875305175781,
      "eval_loss": 0.11548212170600891,
      "eval_rewards/accuracies": 0.7456896305084229,
      "eval_rewards/chosen": -0.3695334494113922,
      "eval_rewards/margins": 0.5571174621582031,
      "eval_rewards/rejected": -0.9266510009765625,
      "eval_runtime": 91.8257,
      "eval_samples_per_second": 19.798,
      "eval_steps_per_second": 0.316,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 9.32642322141669,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": -0.07652176916599274,
      "logits/rejected": 0.6259859800338745,
      "logps/chosen": -321.20538330078125,
      "logps/rejected": -355.15618896484375,
      "loss": 0.1239,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.43240228295326233,
      "rewards/margins": 0.4546627104282379,
      "rewards/rejected": -0.887065052986145,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 9.870138396091912,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": 0.15437743067741394,
      "logits/rejected": 0.9939621686935425,
      "logps/chosen": -314.3326416015625,
      "logps/rejected": -335.50787353515625,
      "loss": 0.1193,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.40100646018981934,
      "rewards/margins": 0.6091704964637756,
      "rewards/rejected": -1.0101768970489502,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 10.470680354911353,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": 0.41432857513427734,
      "logits/rejected": 0.9864773750305176,
      "logps/chosen": -319.98114013671875,
      "logps/rejected": -336.5364074707031,
      "loss": 0.1183,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.45458975434303284,
      "rewards/margins": 0.43977364897727966,
      "rewards/rejected": -0.8943634033203125,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 10.41854471188631,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": 0.3841426968574524,
      "logits/rejected": 0.9351291656494141,
      "logps/chosen": -334.6560974121094,
      "logps/rejected": -349.6623229980469,
      "loss": 0.1152,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.4666348397731781,
      "rewards/margins": 0.4301484227180481,
      "rewards/rejected": -0.8967832326889038,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 9.996573814918259,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": 0.6401292085647583,
      "logits/rejected": 1.3996922969818115,
      "logps/chosen": -324.6096496582031,
      "logps/rejected": -337.38555908203125,
      "loss": 0.1213,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.5011857151985168,
      "rewards/margins": 0.4960555136203766,
      "rewards/rejected": -0.997241199016571,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": 0.5151604413986206,
      "eval_logits/rejected": 1.5740225315093994,
      "eval_logps/chosen": -329.048583984375,
      "eval_logps/rejected": -348.0972900390625,
      "eval_loss": 0.113578200340271,
      "eval_rewards/accuracies": 0.7629310488700867,
      "eval_rewards/chosen": -0.4395850598812103,
      "eval_rewards/margins": 0.5761635303497314,
      "eval_rewards/rejected": -1.0157485008239746,
      "eval_runtime": 92.2267,
      "eval_samples_per_second": 19.712,
      "eval_steps_per_second": 0.314,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 8.23863269387286,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": 0.49835261702537537,
      "logits/rejected": 1.0901247262954712,
      "logps/chosen": -292.41424560546875,
      "logps/rejected": -345.85540771484375,
      "loss": 0.1131,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.40094178915023804,
      "rewards/margins": 0.4722411632537842,
      "rewards/rejected": -0.8731829524040222,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 10.10305933177313,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": 0.16921785473823547,
      "logits/rejected": 0.9673768877983093,
      "logps/chosen": -340.80059814453125,
      "logps/rejected": -353.12274169921875,
      "loss": 0.1129,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.41216129064559937,
      "rewards/margins": 0.4933200776576996,
      "rewards/rejected": -0.9054813385009766,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 9.150940276647933,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": 0.3220874071121216,
      "logits/rejected": 1.2661678791046143,
      "logps/chosen": -332.2058410644531,
      "logps/rejected": -351.96856689453125,
      "loss": 0.1174,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.35827815532684326,
      "rewards/margins": 0.5848454236984253,
      "rewards/rejected": -0.9431236386299133,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 9.488885903438641,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": 0.6681944727897644,
      "logits/rejected": 1.2935806512832642,
      "logps/chosen": -292.4158630371094,
      "logps/rejected": -315.7539978027344,
      "loss": 0.123,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.40458402037620544,
      "rewards/margins": 0.4598938524723053,
      "rewards/rejected": -0.8644779324531555,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 9.000204295912624,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": 0.7832537889480591,
      "logits/rejected": 1.2279746532440186,
      "logps/chosen": -260.876708984375,
      "logps/rejected": -314.25390625,
      "loss": 0.12,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.3631708323955536,
      "rewards/margins": 0.4872121214866638,
      "rewards/rejected": -0.8503829836845398,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": 0.40672120451927185,
      "eval_logits/rejected": 1.4352502822875977,
      "eval_logps/chosen": -324.5692443847656,
      "eval_logps/rejected": -342.49981689453125,
      "eval_loss": 0.11321888864040375,
      "eval_rewards/accuracies": 0.7543103694915771,
      "eval_rewards/chosen": -0.39479103684425354,
      "eval_rewards/margins": 0.5649827122688293,
      "eval_rewards/rejected": -0.9597737789154053,
      "eval_runtime": 92.1464,
      "eval_samples_per_second": 19.729,
      "eval_steps_per_second": 0.315,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 12.797858678996972,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": 0.733371913433075,
      "logits/rejected": 1.6556295156478882,
      "logps/chosen": -313.32049560546875,
      "logps/rejected": -336.0906677246094,
      "loss": 0.1171,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.42368635535240173,
      "rewards/margins": 0.5567531585693359,
      "rewards/rejected": -0.9804395437240601,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 11.330514295442143,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": 0.8590083122253418,
      "logits/rejected": 1.5339264869689941,
      "logps/chosen": -304.95831298828125,
      "logps/rejected": -347.80169677734375,
      "loss": 0.1153,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.4551934599876404,
      "rewards/margins": 0.5640500783920288,
      "rewards/rejected": -1.019243597984314,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 9.086504274831077,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": 0.5266289710998535,
      "logits/rejected": 1.5435163974761963,
      "logps/chosen": -344.6377868652344,
      "logps/rejected": -336.5421447753906,
      "loss": 0.1158,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.41462403535842896,
      "rewards/margins": 0.5104321837425232,
      "rewards/rejected": -0.9250561594963074,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.13218348616853767,
      "train_runtime": 11343.5766,
      "train_samples_per_second": 4.915,
      "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
}