{
  "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": 8.180421521208332,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.6708250045776367,
      "logits/rejected": -2.6100573539733887,
      "logps/chosen": -280.91131591796875,
      "logps/rejected": -254.9091033935547,
      "loss": 0.6931,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 2.8257753001525998e-05,
      "rewards/margins": -1.8637976609170437e-05,
      "rewards/rejected": 4.689567140303552e-05,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 10.108880291822361,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.6290431022644043,
      "logits/rejected": -2.5482747554779053,
      "logps/chosen": -295.01123046875,
      "logps/rejected": -235.47012329101562,
      "loss": 0.6922,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": 0.0019573778845369816,
      "rewards/margins": 0.002028578193858266,
      "rewards/rejected": -7.120029476936907e-05,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 7.7116147316900845,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6313440799713135,
      "logits/rejected": -2.608391046524048,
      "logps/chosen": -299.28961181640625,
      "logps/rejected": -286.8223571777344,
      "loss": 0.6878,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": 0.011437652632594109,
      "rewards/margins": 0.009487219154834747,
      "rewards/rejected": 0.0019504327792674303,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 8.453965686447793,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.6830360889434814,
      "logits/rejected": -2.6185193061828613,
      "logps/chosen": -314.0823669433594,
      "logps/rejected": -274.99749755859375,
      "loss": 0.6763,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": 0.04395774006843567,
      "rewards/margins": 0.034831658005714417,
      "rewards/rejected": 0.009126082062721252,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 8.520035411779036,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.5832035541534424,
      "logits/rejected": -2.560774564743042,
      "logps/chosen": -294.65460205078125,
      "logps/rejected": -279.75335693359375,
      "loss": 0.6573,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.010480286553502083,
      "rewards/margins": 0.08574546873569489,
      "rewards/rejected": -0.07526517659425735,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.516770124435425,
      "eval_logits/rejected": -2.4140138626098633,
      "eval_logps/chosen": -289.93359375,
      "eval_logps/rejected": -261.017333984375,
      "eval_loss": 0.6482123136520386,
      "eval_rewards/accuracies": 0.642241358757019,
      "eval_rewards/chosen": -0.059276144951581955,
      "eval_rewards/margins": 0.09442207962274551,
      "eval_rewards/rejected": -0.15369825065135956,
      "eval_runtime": 91.2745,
      "eval_samples_per_second": 19.918,
      "eval_steps_per_second": 0.318,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 15.433607591426341,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.520766496658325,
      "logits/rejected": -2.4426732063293457,
      "logps/chosen": -295.76214599609375,
      "logps/rejected": -292.59765625,
      "loss": 0.6376,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.0482044480741024,
      "rewards/margins": 0.11990388482809067,
      "rewards/rejected": -0.16810834407806396,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 14.264817414504142,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.4215266704559326,
      "logits/rejected": -2.30562162399292,
      "logps/chosen": -306.8285217285156,
      "logps/rejected": -274.642333984375,
      "loss": 0.6185,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.15291962027549744,
      "rewards/margins": 0.16322971880435944,
      "rewards/rejected": -0.3161493241786957,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 13.495382013613375,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.31144380569458,
      "logits/rejected": -2.2500224113464355,
      "logps/chosen": -320.6394348144531,
      "logps/rejected": -311.4682922363281,
      "loss": 0.6172,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": -0.4840007424354553,
      "rewards/margins": 0.20117254555225372,
      "rewards/rejected": -0.6851732134819031,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 15.127866867375296,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -1.7405483722686768,
      "logits/rejected": -1.5851049423217773,
      "logps/chosen": -324.17535400390625,
      "logps/rejected": -355.6795959472656,
      "loss": 0.6095,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -0.6825101971626282,
      "rewards/margins": 0.29254621267318726,
      "rewards/rejected": -0.9750563502311707,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 18.548127952359756,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -1.1513614654541016,
      "logits/rejected": -0.8654650449752808,
      "logps/chosen": -322.5659484863281,
      "logps/rejected": -328.91851806640625,
      "loss": 0.5517,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.3547930121421814,
      "rewards/margins": 0.43715816736221313,
      "rewards/rejected": -0.7919511795043945,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -0.635834276676178,
      "eval_logits/rejected": -0.16341222822666168,
      "eval_logps/chosen": -332.67840576171875,
      "eval_logps/rejected": -346.1421813964844,
      "eval_loss": 0.5831290483474731,
      "eval_rewards/accuracies": 0.693965494632721,
      "eval_rewards/chosen": -0.48672425746917725,
      "eval_rewards/margins": 0.5182227492332458,
      "eval_rewards/rejected": -1.0049471855163574,
      "eval_runtime": 92.7607,
      "eval_samples_per_second": 19.599,
      "eval_steps_per_second": 0.313,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 21.86823905005592,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -0.5007633566856384,
      "logits/rejected": -0.005487987305969,
      "logps/chosen": -308.6052551269531,
      "logps/rejected": -346.922607421875,
      "loss": 0.582,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.43811899423599243,
      "rewards/margins": 0.6061130166053772,
      "rewards/rejected": -1.0442321300506592,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 25.537041781100793,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": -0.5123111605644226,
      "logits/rejected": 0.09549371898174286,
      "logps/chosen": -376.27618408203125,
      "logps/rejected": -340.7986145019531,
      "loss": 0.5541,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.5627577304840088,
      "rewards/margins": 0.48137393593788147,
      "rewards/rejected": -1.0441316366195679,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 23.843161239454492,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": -0.09251334518194199,
      "logits/rejected": 0.23536929488182068,
      "logps/chosen": -317.888427734375,
      "logps/rejected": -373.44952392578125,
      "loss": 0.5659,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.7276519536972046,
      "rewards/margins": 0.4391087591648102,
      "rewards/rejected": -1.1667606830596924,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 31.20904640703908,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": -0.3974105715751648,
      "logits/rejected": 0.24647757411003113,
      "logps/chosen": -351.69537353515625,
      "logps/rejected": -372.2206115722656,
      "loss": 0.5453,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.6338341236114502,
      "rewards/margins": 0.5683634877204895,
      "rewards/rejected": -1.202197551727295,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 24.781311573851625,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": -0.8739379048347473,
      "logits/rejected": -0.45592212677001953,
      "logps/chosen": -330.25274658203125,
      "logps/rejected": -367.8900146484375,
      "loss": 0.5596,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.5034725069999695,
      "rewards/margins": 0.5031090974807739,
      "rewards/rejected": -1.0065815448760986,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": -1.00446355342865,
      "eval_logits/rejected": -0.24214474856853485,
      "eval_logps/chosen": -326.3860778808594,
      "eval_logps/rejected": -345.77154541015625,
      "eval_loss": 0.5497105717658997,
      "eval_rewards/accuracies": 0.7241379022598267,
      "eval_rewards/chosen": -0.42380115389823914,
      "eval_rewards/margins": 0.577439546585083,
      "eval_rewards/rejected": -1.001240611076355,
      "eval_runtime": 92.0532,
      "eval_samples_per_second": 19.749,
      "eval_steps_per_second": 0.315,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 22.924194135406648,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": -0.7640997767448425,
      "logits/rejected": 0.009471100755035877,
      "logps/chosen": -348.05657958984375,
      "logps/rejected": -402.06524658203125,
      "loss": 0.5524,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.5906395316123962,
      "rewards/margins": 0.6577303409576416,
      "rewards/rejected": -1.248369812965393,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 20.23639815537178,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": -0.6523529291152954,
      "logits/rejected": -0.03978896886110306,
      "logps/chosen": -377.78094482421875,
      "logps/rejected": -413.15313720703125,
      "loss": 0.5447,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.7973796129226685,
      "rewards/margins": 0.6492636799812317,
      "rewards/rejected": -1.446643352508545,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 18.90429435262536,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": -0.7708563208580017,
      "logits/rejected": -0.08572294563055038,
      "logps/chosen": -347.65289306640625,
      "logps/rejected": -362.48541259765625,
      "loss": 0.5317,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.7364877462387085,
      "rewards/margins": 0.6042462587356567,
      "rewards/rejected": -1.3407340049743652,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 31.220611180756855,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": -0.38854315876960754,
      "logits/rejected": 0.21274462342262268,
      "logps/chosen": -350.7613525390625,
      "logps/rejected": -399.95086669921875,
      "loss": 0.5437,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.8297553062438965,
      "rewards/margins": 0.6977420449256897,
      "rewards/rejected": -1.5274971723556519,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 30.74591682430673,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": 0.04708825796842575,
      "logits/rejected": 0.7175595760345459,
      "logps/chosen": -389.9985656738281,
      "logps/rejected": -427.6483459472656,
      "loss": 0.557,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.0787255764007568,
      "rewards/margins": 0.6673825979232788,
      "rewards/rejected": -1.7461086511611938,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": -0.24332188069820404,
      "eval_logits/rejected": 0.630587100982666,
      "eval_logps/chosen": -360.6939392089844,
      "eval_logps/rejected": -391.9894714355469,
      "eval_loss": 0.5398454666137695,
      "eval_rewards/accuracies": 0.732758641242981,
      "eval_rewards/chosen": -0.7668798565864563,
      "eval_rewards/margins": 0.6965396404266357,
      "eval_rewards/rejected": -1.4634195566177368,
      "eval_runtime": 92.2093,
      "eval_samples_per_second": 19.716,
      "eval_steps_per_second": 0.315,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 20.053636726221647,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": -0.5000969171524048,
      "logits/rejected": 0.287570059299469,
      "logps/chosen": -364.197021484375,
      "logps/rejected": -397.1741943359375,
      "loss": 0.521,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.5910893082618713,
      "rewards/margins": 0.7975226044654846,
      "rewards/rejected": -1.388611912727356,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 29.09938855135104,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": -0.06796761602163315,
      "logits/rejected": 0.6887415647506714,
      "logps/chosen": -362.93719482421875,
      "logps/rejected": -396.042724609375,
      "loss": 0.5366,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.6787573099136353,
      "rewards/margins": 0.7150012850761414,
      "rewards/rejected": -1.3937586545944214,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 30.18521722411796,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": 0.06652301549911499,
      "logits/rejected": 0.9086186289787292,
      "logps/chosen": -363.90216064453125,
      "logps/rejected": -403.9359436035156,
      "loss": 0.5387,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.8593271970748901,
      "rewards/margins": 0.7014445066452026,
      "rewards/rejected": -1.5607718229293823,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 27.269767101203968,
      "learning_rate": 2.5e-07,
      "logits/chosen": -0.38139885663986206,
      "logits/rejected": 0.7918380498886108,
      "logps/chosen": -391.44293212890625,
      "logps/rejected": -436.5362243652344,
      "loss": 0.5051,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.770719051361084,
      "rewards/margins": 0.832466721534729,
      "rewards/rejected": -1.6031854152679443,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 25.963088610849898,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": 0.0945362076163292,
      "logits/rejected": 1.0674773454666138,
      "logps/chosen": -381.2936706542969,
      "logps/rejected": -445.334716796875,
      "loss": 0.5483,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.8827805519104004,
      "rewards/margins": 0.6815775036811829,
      "rewards/rejected": -1.564357876777649,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": 0.15348348021507263,
      "eval_logits/rejected": 1.1693741083145142,
      "eval_logps/chosen": -374.9230651855469,
      "eval_logps/rejected": -410.466064453125,
      "eval_loss": 0.5334360599517822,
      "eval_rewards/accuracies": 0.7370689511299133,
      "eval_rewards/chosen": -0.9091711640357971,
      "eval_rewards/margins": 0.7390145659446716,
      "eval_rewards/rejected": -1.6481858491897583,
      "eval_runtime": 91.8172,
      "eval_samples_per_second": 19.8,
      "eval_steps_per_second": 0.316,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 24.61101929030597,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": 0.17619961500167847,
      "logits/rejected": 0.9847076535224915,
      "logps/chosen": -403.1936340332031,
      "logps/rejected": -437.13897705078125,
      "loss": 0.5437,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.0121877193450928,
      "rewards/margins": 0.678584098815918,
      "rewards/rejected": -1.6907718181610107,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 20.367684952689917,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": -0.2150786817073822,
      "logits/rejected": 0.4295934736728668,
      "logps/chosen": -364.59625244140625,
      "logps/rejected": -382.64544677734375,
      "loss": 0.5429,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.7735622525215149,
      "rewards/margins": 0.5857953429222107,
      "rewards/rejected": -1.3593575954437256,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 22.118134031317915,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": 0.00539786834269762,
      "logits/rejected": 1.0036160945892334,
      "logps/chosen": -326.5462341308594,
      "logps/rejected": -401.6683654785156,
      "loss": 0.5232,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.7456392645835876,
      "rewards/margins": 0.8400524854660034,
      "rewards/rejected": -1.5856916904449463,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 25.56546771816306,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": 0.18758408725261688,
      "logits/rejected": 1.4861918687820435,
      "logps/chosen": -357.71759033203125,
      "logps/rejected": -386.7574768066406,
      "loss": 0.5156,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.7674424648284912,
      "rewards/margins": 0.8558618426322937,
      "rewards/rejected": -1.6233041286468506,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 24.341607485547637,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": 0.2733023762702942,
      "logits/rejected": 1.3938772678375244,
      "logps/chosen": -372.9563903808594,
      "logps/rejected": -409.0280456542969,
      "loss": 0.5338,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.7974545955657959,
      "rewards/margins": 0.7862997055053711,
      "rewards/rejected": -1.583754301071167,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": 0.05302007496356964,
      "eval_logits/rejected": 1.1201387643814087,
      "eval_logps/chosen": -354.7213439941406,
      "eval_logps/rejected": -390.7056579589844,
      "eval_loss": 0.5226743817329407,
      "eval_rewards/accuracies": 0.7241379022598267,
      "eval_rewards/chosen": -0.7071539163589478,
      "eval_rewards/margins": 0.7434275150299072,
      "eval_rewards/rejected": -1.450581431388855,
      "eval_runtime": 91.9962,
      "eval_samples_per_second": 19.762,
      "eval_steps_per_second": 0.315,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 23.93721697368582,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": 0.01756982132792473,
      "logits/rejected": 0.9820684194564819,
      "logps/chosen": -369.3329772949219,
      "logps/rejected": -423.09552001953125,
      "loss": 0.5194,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.7456832528114319,
      "rewards/margins": 0.7584327459335327,
      "rewards/rejected": -1.5041159391403198,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 27.593197014560943,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": 0.2417779266834259,
      "logits/rejected": 1.0749890804290771,
      "logps/chosen": -362.8450927734375,
      "logps/rejected": -409.4270324707031,
      "loss": 0.5134,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.771125316619873,
      "rewards/margins": 0.7913394570350647,
      "rewards/rejected": -1.5624648332595825,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 24.639500410291397,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": 0.31850242614746094,
      "logits/rejected": 1.264216423034668,
      "logps/chosen": -382.9512634277344,
      "logps/rejected": -421.80419921875,
      "loss": 0.5279,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.9504399299621582,
      "rewards/margins": 0.6338992118835449,
      "rewards/rejected": -1.5843393802642822,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 27.22903391760232,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": 0.4879975914955139,
      "logits/rejected": 1.7275480031967163,
      "logps/chosen": -377.9282531738281,
      "logps/rejected": -405.58038330078125,
      "loss": 0.531,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.9674369692802429,
      "rewards/margins": 0.7435644865036011,
      "rewards/rejected": -1.7110016345977783,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 23.43707551698937,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": 0.2636318504810333,
      "logits/rejected": 1.2225837707519531,
      "logps/chosen": -377.68994140625,
      "logps/rejected": -407.3072814941406,
      "loss": 0.5111,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.6589912176132202,
      "rewards/margins": 0.8298226594924927,
      "rewards/rejected": -1.4888139963150024,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": 0.2474285364151001,
      "eval_logits/rejected": 1.3205074071884155,
      "eval_logps/chosen": -361.77734375,
      "eval_logps/rejected": -398.4795837402344,
      "eval_loss": 0.5173320770263672,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -0.777714192867279,
      "eval_rewards/margins": 0.7506070137023926,
      "eval_rewards/rejected": -1.528321385383606,
      "eval_runtime": 91.2793,
      "eval_samples_per_second": 19.917,
      "eval_steps_per_second": 0.318,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 24.131364526405257,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": 0.3964577317237854,
      "logits/rejected": 1.2822027206420898,
      "logps/chosen": -376.3506164550781,
      "logps/rejected": -434.9208068847656,
      "loss": 0.5363,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.8864477276802063,
      "rewards/margins": 0.8174076080322266,
      "rewards/rejected": -1.703855276107788,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 24.753082391706794,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": 0.3074328601360321,
      "logits/rejected": 1.4143335819244385,
      "logps/chosen": -383.6210632324219,
      "logps/rejected": -398.04010009765625,
      "loss": 0.5163,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.8147560954093933,
      "rewards/margins": 0.836268424987793,
      "rewards/rejected": -1.651024580001831,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 24.185031971952714,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": 0.6229906678199768,
      "logits/rejected": 1.3948328495025635,
      "logps/chosen": -357.06268310546875,
      "logps/rejected": -401.7029724121094,
      "loss": 0.5288,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.9318356513977051,
      "rewards/margins": 0.685466468334198,
      "rewards/rejected": -1.6173019409179688,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 25.966162156707668,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": 0.46141189336776733,
      "logits/rejected": 1.5275745391845703,
      "logps/chosen": -373.74420166015625,
      "logps/rejected": -420.73968505859375,
      "loss": 0.5163,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.8795710802078247,
      "rewards/margins": 0.8060371279716492,
      "rewards/rejected": -1.685608148574829,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 22.442914644150154,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": 0.351333886384964,
      "logits/rejected": 1.4142162799835205,
      "logps/chosen": -361.7117004394531,
      "logps/rejected": -423.2845764160156,
      "loss": 0.5185,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.775847315788269,
      "rewards/margins": 0.930949866771698,
      "rewards/rejected": -1.7067972421646118,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": 0.3004731237888336,
      "eval_logits/rejected": 1.3740564584732056,
      "eval_logps/chosen": -365.25396728515625,
      "eval_logps/rejected": -402.4430847167969,
      "eval_loss": 0.5178534984588623,
      "eval_rewards/accuracies": 0.7241379022598267,
      "eval_rewards/chosen": -0.8124799132347107,
      "eval_rewards/margins": 0.7554758191108704,
      "eval_rewards/rejected": -1.567955493927002,
      "eval_runtime": 92.4415,
      "eval_samples_per_second": 19.666,
      "eval_steps_per_second": 0.314,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 22.619715289441434,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": 0.658234715461731,
      "logits/rejected": 1.4394605159759521,
      "logps/chosen": -335.9058837890625,
      "logps/rejected": -414.93212890625,
      "loss": 0.5145,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.927649199962616,
      "rewards/margins": 0.7706761360168457,
      "rewards/rejected": -1.6983253955841064,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 25.83898128993406,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": 0.38683071732521057,
      "logits/rejected": 1.4439704418182373,
      "logps/chosen": -358.71893310546875,
      "logps/rejected": -380.32879638671875,
      "loss": 0.5317,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.8638426065444946,
      "rewards/margins": 0.7388638257980347,
      "rewards/rejected": -1.6027065515518188,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 22.9017291219974,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": 0.3319759964942932,
      "logits/rejected": 1.3225994110107422,
      "logps/chosen": -365.75030517578125,
      "logps/rejected": -427.2379455566406,
      "loss": 0.5134,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.9414383172988892,
      "rewards/margins": 0.7166720628738403,
      "rewards/rejected": -1.6581103801727295,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.5596197334998244,
      "train_runtime": 11808.0241,
      "train_samples_per_second": 4.722,
      "train_steps_per_second": 0.037
    }
  ],
  "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
}