{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.5792,
  "eval_steps": 62,
  "global_step": 806,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 76.02843905366466,
      "learning_rate": 5.3191489361702125e-09,
      "logits/generated": -1.098149061203003,
      "logits/real": -0.27942562103271484,
      "logps/generated": -261.58599853515625,
      "logps/real": -202.4606475830078,
      "loss": 0.9167,
      "rewards/accuracies": 0.0,
      "rewards/generated": 0.0,
      "rewards/margins": 0.0,
      "rewards/real": 0.0,
      "step": 1
    },
    {
      "epoch": 0.032,
      "grad_norm": 83.5080388615708,
      "learning_rate": 5.3191489361702123e-08,
      "logits/generated": -0.964782178401947,
      "logits/real": -0.43690696358680725,
      "logps/generated": -277.5035400390625,
      "logps/real": -213.53489685058594,
      "loss": 0.8941,
      "rewards/accuracies": 0.3888888955116272,
      "rewards/generated": 0.007520654238760471,
      "rewards/margins": -0.009737681597471237,
      "rewards/real": -0.002217027824372053,
      "step": 10
    },
    {
      "epoch": 0.064,
      "grad_norm": 66.11667224751555,
      "learning_rate": 1.0638297872340425e-07,
      "logits/generated": -1.0118372440338135,
      "logits/real": -0.4319210946559906,
      "logps/generated": -281.8486633300781,
      "logps/real": -229.50259399414062,
      "loss": 0.8482,
      "rewards/accuracies": 0.887499988079071,
      "rewards/generated": -0.08688100427389145,
      "rewards/margins": 0.1324445903301239,
      "rewards/real": 0.04556358605623245,
      "step": 20
    },
    {
      "epoch": 0.096,
      "grad_norm": 83.1043523213915,
      "learning_rate": 1.5957446808510638e-07,
      "logits/generated": -0.911736011505127,
      "logits/real": -0.6389068365097046,
      "logps/generated": -287.8795471191406,
      "logps/real": -189.71298217773438,
      "loss": 0.6691,
      "rewards/accuracies": 1.0,
      "rewards/generated": -0.3395097553730011,
      "rewards/margins": 0.5189359188079834,
      "rewards/real": 0.1794261634349823,
      "step": 30
    },
    {
      "epoch": 0.128,
      "grad_norm": 17.588838075583457,
      "learning_rate": 2.127659574468085e-07,
      "logits/generated": -0.9573461413383484,
      "logits/real": -0.4479493498802185,
      "logps/generated": -294.65081787109375,
      "logps/real": -203.56137084960938,
      "loss": 0.4483,
      "rewards/accuracies": 1.0,
      "rewards/generated": -0.8644129633903503,
      "rewards/margins": 1.436708688735962,
      "rewards/real": 0.5722957849502563,
      "step": 40
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.9686616342963887,
      "learning_rate": 2.659574468085106e-07,
      "logits/generated": -0.8112080693244934,
      "logits/real": -0.5195661187171936,
      "logps/generated": -303.88958740234375,
      "logps/real": -186.9166259765625,
      "loss": 0.2406,
      "rewards/accuracies": 1.0,
      "rewards/generated": -1.8978115320205688,
      "rewards/margins": 3.3566880226135254,
      "rewards/real": 1.458876371383667,
      "step": 50
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.873608373543979,
      "learning_rate": 3.1914893617021275e-07,
      "logits/generated": -0.7203534841537476,
      "logits/real": -0.339523047208786,
      "logps/generated": -325.28619384765625,
      "logps/real": -161.00442504882812,
      "loss": 0.1687,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.045280933380127,
      "rewards/margins": 7.439988613128662,
      "rewards/real": 3.3947081565856934,
      "step": 60
    },
    {
      "epoch": 0.1984,
      "eval_logits/generated": -0.7218369841575623,
      "eval_logits/real": -0.40187305212020874,
      "eval_logps/generated": -335.216796875,
      "eval_logps/real": -154.80484008789062,
      "eval_loss": 0.155403733253479,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -5.2547712326049805,
      "eval_rewards/margins": 10.460118293762207,
      "eval_rewards/real": 5.205347537994385,
      "eval_runtime": 27.3659,
      "eval_samples_per_second": 7.308,
      "eval_steps_per_second": 0.475,
      "step": 62
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.9090466290588308,
      "learning_rate": 3.7234042553191484e-07,
      "logits/generated": -0.7592553496360779,
      "logits/real": -0.5271034240722656,
      "logps/generated": -341.49017333984375,
      "logps/real": -148.29876708984375,
      "loss": 0.1549,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.969511985778809,
      "rewards/margins": 10.437345504760742,
      "rewards/real": 5.467833042144775,
      "step": 70
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.5548494311175409,
      "learning_rate": 4.25531914893617e-07,
      "logits/generated": -0.7383196949958801,
      "logits/real": -0.8009077906608582,
      "logps/generated": -321.6851806640625,
      "logps/real": -123.8510971069336,
      "loss": 0.1358,
      "rewards/accuracies": 1.0,
      "rewards/generated": -3.824869155883789,
      "rewards/margins": 10.858232498168945,
      "rewards/real": 7.033363342285156,
      "step": 80
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.42292249756861683,
      "learning_rate": 4.787234042553192e-07,
      "logits/generated": -0.8187915682792664,
      "logits/real": -1.0330232381820679,
      "logps/generated": -318.3337707519531,
      "logps/real": -120.26998138427734,
      "loss": 0.1213,
      "rewards/accuracies": 1.0,
      "rewards/generated": -3.1653218269348145,
      "rewards/margins": 11.506210327148438,
      "rewards/real": 8.340888977050781,
      "step": 90
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.41767196123848027,
      "learning_rate": 4.96437054631829e-07,
      "logits/generated": -0.8708025217056274,
      "logits/real": -1.1798157691955566,
      "logps/generated": -309.2374572753906,
      "logps/real": -115.8353042602539,
      "loss": 0.1134,
      "rewards/accuracies": 1.0,
      "rewards/generated": -3.7763671875,
      "rewards/margins": 12.563192367553711,
      "rewards/real": 8.786824226379395,
      "step": 100
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.4453179950421757,
      "learning_rate": 4.904988123515439e-07,
      "logits/generated": -0.8516510725021362,
      "logits/real": -1.1489499807357788,
      "logps/generated": -326.2418518066406,
      "logps/real": -125.87767028808594,
      "loss": 0.1137,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.001976013183594,
      "rewards/margins": 13.618423461914062,
      "rewards/real": 9.616447448730469,
      "step": 110
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.39875861828610537,
      "learning_rate": 4.845605700712589e-07,
      "logits/generated": -0.7948419451713562,
      "logits/real": -1.1261526346206665,
      "logps/generated": -331.29888916015625,
      "logps/real": -118.65771484375,
      "loss": 0.1204,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.266651153564453,
      "rewards/margins": 13.141134262084961,
      "rewards/real": 8.874483108520508,
      "step": 120
    },
    {
      "epoch": 0.3968,
      "eval_logits/generated": -0.8262335658073425,
      "eval_logits/real": -1.1627280712127686,
      "eval_logps/generated": -327.9040832519531,
      "eval_logps/real": -113.16130065917969,
      "eval_loss": 0.11531589180231094,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -4.523497581481934,
      "eval_rewards/margins": 13.89319896697998,
      "eval_rewards/real": 9.369701385498047,
      "eval_runtime": 31.0275,
      "eval_samples_per_second": 6.446,
      "eval_steps_per_second": 0.419,
      "step": 124
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.35711714781872617,
      "learning_rate": 4.786223277909738e-07,
      "logits/generated": -0.8075119853019714,
      "logits/real": -1.1974109411239624,
      "logps/generated": -326.8999938964844,
      "logps/real": -121.57466125488281,
      "loss": 0.1108,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.337772369384766,
      "rewards/margins": 13.636154174804688,
      "rewards/real": 9.298380851745605,
      "step": 130
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.3901220032340114,
      "learning_rate": 4.7268408551068883e-07,
      "logits/generated": -0.8377895355224609,
      "logits/real": -1.1887013912200928,
      "logps/generated": -325.7310791015625,
      "logps/real": -123.4073715209961,
      "loss": 0.1129,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.29754114151001,
      "rewards/margins": 13.601455688476562,
      "rewards/real": 9.303915023803711,
      "step": 140
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3874502118142099,
      "learning_rate": 4.667458432304038e-07,
      "logits/generated": -0.8494235277175903,
      "logits/real": -1.280548334121704,
      "logps/generated": -323.9864807128906,
      "logps/real": -115.0036849975586,
      "loss": 0.1119,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.178435325622559,
      "rewards/margins": 13.569964408874512,
      "rewards/real": 9.391529083251953,
      "step": 150
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.408082769634744,
      "learning_rate": 4.6080760095011875e-07,
      "logits/generated": -0.8601406812667847,
      "logits/real": -1.2258448600769043,
      "logps/generated": -328.79156494140625,
      "logps/real": -124.53865814208984,
      "loss": 0.117,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.461178779602051,
      "rewards/margins": 14.086248397827148,
      "rewards/real": 9.625069618225098,
      "step": 160
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.36331526860516933,
      "learning_rate": 4.548693586698337e-07,
      "logits/generated": -0.8272736668586731,
      "logits/real": -1.278106451034546,
      "logps/generated": -331.06683349609375,
      "logps/real": -119.9251480102539,
      "loss": 0.1135,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.8740129470825195,
      "rewards/margins": 14.372535705566406,
      "rewards/real": 9.498522758483887,
      "step": 170
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.445865795650114,
      "learning_rate": 4.4893111638954866e-07,
      "logits/generated": -0.8073589205741882,
      "logits/real": -1.2638658285140991,
      "logps/generated": -342.365966796875,
      "logps/real": -106.32196044921875,
      "loss": 0.1114,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.440592288970947,
      "rewards/margins": 14.317230224609375,
      "rewards/real": 8.876636505126953,
      "step": 180
    },
    {
      "epoch": 0.5952,
      "eval_logits/generated": -0.8445993065834045,
      "eval_logits/real": -1.2393124103546143,
      "eval_logps/generated": -335.83538818359375,
      "eval_logps/real": -110.11853790283203,
      "eval_loss": 0.11254960298538208,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -5.31662654876709,
      "eval_rewards/margins": 14.990604400634766,
      "eval_rewards/real": 9.673977851867676,
      "eval_runtime": 25.7798,
      "eval_samples_per_second": 7.758,
      "eval_steps_per_second": 0.504,
      "step": 186
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.39320789077558344,
      "learning_rate": 4.429928741092636e-07,
      "logits/generated": -0.8406023979187012,
      "logits/real": -1.2843071222305298,
      "logps/generated": -316.73291015625,
      "logps/real": -109.3148193359375,
      "loss": 0.1103,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.513542175292969,
      "rewards/margins": 14.111068725585938,
      "rewards/real": 9.597526550292969,
      "step": 190
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.34126559985788285,
      "learning_rate": 4.3705463182897863e-07,
      "logits/generated": -0.8079656362533569,
      "logits/real": -1.2299845218658447,
      "logps/generated": -333.931396484375,
      "logps/real": -113.5406723022461,
      "loss": 0.1079,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.271535873413086,
      "rewards/margins": 14.859156608581543,
      "rewards/real": 9.587618827819824,
      "step": 200
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.37211551392957165,
      "learning_rate": 4.311163895486936e-07,
      "logits/generated": -0.8258484601974487,
      "logits/real": -1.2792431116104126,
      "logps/generated": -326.34539794921875,
      "logps/real": -110.4327621459961,
      "loss": 0.1062,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.544249534606934,
      "rewards/margins": 13.822877883911133,
      "rewards/real": 9.278627395629883,
      "step": 210
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.40528624225351695,
      "learning_rate": 4.251781472684085e-07,
      "logits/generated": -0.9005807638168335,
      "logits/real": -1.367760419845581,
      "logps/generated": -334.90667724609375,
      "logps/real": -108.89170837402344,
      "loss": 0.1082,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.304292678833008,
      "rewards/margins": 14.677830696105957,
      "rewards/real": 9.37353801727295,
      "step": 220
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.4299824458037775,
      "learning_rate": 4.192399049881235e-07,
      "logits/generated": -0.871414065361023,
      "logits/real": -1.2680821418762207,
      "logps/generated": -327.0166015625,
      "logps/real": -125.11517333984375,
      "loss": 0.1201,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.102071285247803,
      "rewards/margins": 15.333366394042969,
      "rewards/real": 10.231294631958008,
      "step": 230
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.3200345894830024,
      "learning_rate": 4.1330166270783846e-07,
      "logits/generated": -0.8396976590156555,
      "logits/real": -1.2651922702789307,
      "logps/generated": -349.18804931640625,
      "logps/real": -117.8864517211914,
      "loss": 0.1094,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.743192672729492,
      "rewards/margins": 15.722867965698242,
      "rewards/real": 9.979674339294434,
      "step": 240
    },
    {
      "epoch": 0.7936,
      "eval_logits/generated": -0.8538265228271484,
      "eval_logits/real": -1.2560151815414429,
      "eval_logps/generated": -337.5219421386719,
      "eval_logps/real": -108.52310180664062,
      "eval_loss": 0.11097312718629837,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -5.485285758972168,
      "eval_rewards/margins": 15.318806648254395,
      "eval_rewards/real": 9.83352279663086,
      "eval_runtime": 32.0233,
      "eval_samples_per_second": 6.245,
      "eval_steps_per_second": 0.406,
      "step": 248
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.37657892572843127,
      "learning_rate": 4.0736342042755347e-07,
      "logits/generated": -0.8381448984146118,
      "logits/real": -1.293889045715332,
      "logps/generated": -344.9915466308594,
      "logps/real": -111.15998840332031,
      "loss": 0.1113,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.526784896850586,
      "rewards/margins": 15.003395080566406,
      "rewards/real": 9.476609230041504,
      "step": 250
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.3794924953311087,
      "learning_rate": 4.0142517814726837e-07,
      "logits/generated": -0.8616489171981812,
      "logits/real": -1.342829704284668,
      "logps/generated": -333.8689270019531,
      "logps/real": -110.18806457519531,
      "loss": 0.1068,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.346303939819336,
      "rewards/margins": 15.174580574035645,
      "rewards/real": 9.828276634216309,
      "step": 260
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.38432288454636715,
      "learning_rate": 3.9548693586698333e-07,
      "logits/generated": -0.8341599702835083,
      "logits/real": -1.2853684425354004,
      "logps/generated": -340.5439147949219,
      "logps/real": -111.25767517089844,
      "loss": 0.1119,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.1574320793151855,
      "rewards/margins": 14.693034172058105,
      "rewards/real": 9.535603523254395,
      "step": 270
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.4988860977100256,
      "learning_rate": 3.8954869358669834e-07,
      "logits/generated": -0.842201828956604,
      "logits/real": -1.3558833599090576,
      "logps/generated": -338.8299560546875,
      "logps/real": -110.4022216796875,
      "loss": 0.1111,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.072219371795654,
      "rewards/margins": 14.635177612304688,
      "rewards/real": 9.562958717346191,
      "step": 280
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.38841887432580663,
      "learning_rate": 3.836104513064133e-07,
      "logits/generated": -0.8865805864334106,
      "logits/real": -1.2532027959823608,
      "logps/generated": -338.3769226074219,
      "logps/real": -120.43833923339844,
      "loss": 0.1084,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.086993217468262,
      "rewards/margins": 15.214749336242676,
      "rewards/real": 10.127756118774414,
      "step": 290
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3616732188669415,
      "learning_rate": 3.7767220902612825e-07,
      "logits/generated": -0.8655015826225281,
      "logits/real": -1.3518953323364258,
      "logps/generated": -325.9730529785156,
      "logps/real": -95.09385681152344,
      "loss": 0.1028,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.991705417633057,
      "rewards/margins": 14.74760627746582,
      "rewards/real": 8.755899429321289,
      "step": 300
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.39359159204366234,
      "learning_rate": 3.717339667458432e-07,
      "logits/generated": -0.8543545603752136,
      "logits/real": -1.3398587703704834,
      "logps/generated": -338.96466064453125,
      "logps/real": -116.5450668334961,
      "loss": 0.1115,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.719244480133057,
      "rewards/margins": 15.74243450164795,
      "rewards/real": 10.02319049835205,
      "step": 310
    },
    {
      "epoch": 0.992,
      "eval_logits/generated": -0.865822434425354,
      "eval_logits/real": -1.33043372631073,
      "eval_logps/generated": -347.4965515136719,
      "eval_logps/real": -107.73165130615234,
      "eval_loss": 0.11000669747591019,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -6.482741832733154,
      "eval_rewards/margins": 16.39541244506836,
      "eval_rewards/real": 9.912667274475098,
      "eval_runtime": 24.7895,
      "eval_samples_per_second": 8.068,
      "eval_steps_per_second": 0.524,
      "step": 310
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.39684443120667556,
      "learning_rate": 3.6579572446555817e-07,
      "logits/generated": -0.9057809710502625,
      "logits/real": -1.3521413803100586,
      "logps/generated": -349.14471435546875,
      "logps/real": -111.87225341796875,
      "loss": 0.1073,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.6079206466674805,
      "rewards/margins": 15.645317077636719,
      "rewards/real": 10.037394523620605,
      "step": 320
    },
    {
      "epoch": 1.056,
      "grad_norm": 0.39872204323404287,
      "learning_rate": 3.598574821852731e-07,
      "logits/generated": -0.8963452577590942,
      "logits/real": -1.4075753688812256,
      "logps/generated": -350.3981018066406,
      "logps/real": -111.39154052734375,
      "loss": 0.1099,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.171029090881348,
      "rewards/margins": 16.38304328918457,
      "rewards/real": 10.212014198303223,
      "step": 330
    },
    {
      "epoch": 1.088,
      "grad_norm": 0.3296521457207489,
      "learning_rate": 3.5391923990498813e-07,
      "logits/generated": -0.882519543170929,
      "logits/real": -1.3683688640594482,
      "logps/generated": -344.4193115234375,
      "logps/real": -114.99576568603516,
      "loss": 0.1121,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.464304447174072,
      "rewards/margins": 16.844728469848633,
      "rewards/real": 10.380424499511719,
      "step": 340
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.40794606558455315,
      "learning_rate": 3.479809976247031e-07,
      "logits/generated": -0.8548433184623718,
      "logits/real": -1.33034086227417,
      "logps/generated": -345.41497802734375,
      "logps/real": -108.29139709472656,
      "loss": 0.1084,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.70036506652832,
      "rewards/margins": 15.80578899383545,
      "rewards/real": 10.105424880981445,
      "step": 350
    },
    {
      "epoch": 1.152,
      "grad_norm": 0.34179390042713315,
      "learning_rate": 3.42042755344418e-07,
      "logits/generated": -0.8854677081108093,
      "logits/real": -1.3697372674942017,
      "logps/generated": -349.7736511230469,
      "logps/real": -117.81876373291016,
      "loss": 0.1098,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.514307975769043,
      "rewards/margins": 16.870037078857422,
      "rewards/real": 10.355731010437012,
      "step": 360
    },
    {
      "epoch": 1.184,
      "grad_norm": 0.38985325345599603,
      "learning_rate": 3.36104513064133e-07,
      "logits/generated": -0.8000022172927856,
      "logits/real": -1.3874753713607788,
      "logps/generated": -347.55010986328125,
      "logps/real": -113.49405670166016,
      "loss": 0.1046,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.464725494384766,
      "rewards/margins": 16.653573989868164,
      "rewards/real": 10.188849449157715,
      "step": 370
    },
    {
      "epoch": 1.1904,
      "eval_logits/generated": -0.8655997514724731,
      "eval_logits/real": -1.3632643222808838,
      "eval_logps/generated": -349.3764953613281,
      "eval_logps/real": -107.03945922851562,
      "eval_loss": 0.10925532132387161,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -6.670737266540527,
      "eval_rewards/margins": 16.652626037597656,
      "eval_rewards/real": 9.981886863708496,
      "eval_runtime": 35.169,
      "eval_samples_per_second": 5.687,
      "eval_steps_per_second": 0.37,
      "step": 372
    },
    {
      "epoch": 1.216,
      "grad_norm": 0.3722532991523644,
      "learning_rate": 3.3016627078384796e-07,
      "logits/generated": -0.864538311958313,
      "logits/real": -1.3105075359344482,
      "logps/generated": -335.27362060546875,
      "logps/real": -106.48799133300781,
      "loss": 0.1035,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.397665977478027,
      "rewards/margins": 16.154638290405273,
      "rewards/real": 9.756973266601562,
      "step": 380
    },
    {
      "epoch": 1.248,
      "grad_norm": 0.4038074067883792,
      "learning_rate": 3.2422802850356297e-07,
      "logits/generated": -0.8852449655532837,
      "logits/real": -1.4460315704345703,
      "logps/generated": -344.63311767578125,
      "logps/real": -106.79902648925781,
      "loss": 0.1027,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.748571395874023,
      "rewards/margins": 15.521743774414062,
      "rewards/real": 9.773170471191406,
      "step": 390
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.28676972353775737,
      "learning_rate": 3.182897862232779e-07,
      "logits/generated": -0.8121256828308105,
      "logits/real": -1.354479432106018,
      "logps/generated": -345.53326416015625,
      "logps/real": -107.84037017822266,
      "loss": 0.1003,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.011273384094238,
      "rewards/margins": 15.705144882202148,
      "rewards/real": 9.693869590759277,
      "step": 400
    },
    {
      "epoch": 1.312,
      "grad_norm": 0.3326501083643602,
      "learning_rate": 3.1235154394299283e-07,
      "logits/generated": -0.8699803352355957,
      "logits/real": -1.3926770687103271,
      "logps/generated": -339.2923583984375,
      "logps/real": -113.30909729003906,
      "loss": 0.1069,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.364248275756836,
      "rewards/margins": 15.17718505859375,
      "rewards/real": 9.81293773651123,
      "step": 410
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 0.41187476133988626,
      "learning_rate": 3.0641330166270784e-07,
      "logits/generated": -0.8451802134513855,
      "logits/real": -1.3541207313537598,
      "logps/generated": -344.1119384765625,
      "logps/real": -99.13960266113281,
      "loss": 0.103,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.0690155029296875,
      "rewards/margins": 15.549135208129883,
      "rewards/real": 9.480117797851562,
      "step": 420
    },
    {
      "epoch": 1.376,
      "grad_norm": 0.37308896269833713,
      "learning_rate": 3.004750593824228e-07,
      "logits/generated": -0.8354321718215942,
      "logits/real": -1.397528052330017,
      "logps/generated": -354.12841796875,
      "logps/real": -108.86185455322266,
      "loss": 0.1067,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.957143306732178,
      "rewards/margins": 16.80294418334961,
      "rewards/real": 9.845802307128906,
      "step": 430
    },
    {
      "epoch": 1.3888,
      "eval_logits/generated": -0.881392776966095,
      "eval_logits/real": -1.3897513151168823,
      "eval_logps/generated": -358.4093933105469,
      "eval_logps/real": -106.73082733154297,
      "eval_loss": 0.10892020910978317,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -7.574033260345459,
      "eval_rewards/margins": 17.586782455444336,
      "eval_rewards/real": 10.012747764587402,
      "eval_runtime": 25.6933,
      "eval_samples_per_second": 7.784,
      "eval_steps_per_second": 0.506,
      "step": 434
    },
    {
      "epoch": 1.408,
      "grad_norm": 0.35037689567096764,
      "learning_rate": 2.9453681710213776e-07,
      "logits/generated": -0.8212960362434387,
      "logits/real": -1.3137534856796265,
      "logps/generated": -355.6631774902344,
      "logps/real": -114.98287200927734,
      "loss": 0.1095,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.6356201171875,
      "rewards/margins": 16.64114761352539,
      "rewards/real": 10.00552749633789,
      "step": 440
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.36349550750442733,
      "learning_rate": 2.885985748218527e-07,
      "logits/generated": -0.8298400044441223,
      "logits/real": -1.2980700731277466,
      "logps/generated": -342.68341064453125,
      "logps/real": -89.83566284179688,
      "loss": 0.1046,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.842597961425781,
      "rewards/margins": 15.478503227233887,
      "rewards/real": 8.635905265808105,
      "step": 450
    },
    {
      "epoch": 1.472,
      "grad_norm": 0.37277388095763225,
      "learning_rate": 2.8266033254156767e-07,
      "logits/generated": -0.891290009021759,
      "logits/real": -1.394723653793335,
      "logps/generated": -341.93218994140625,
      "logps/real": -103.62208557128906,
      "loss": 0.1079,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.340664863586426,
      "rewards/margins": 15.834001541137695,
      "rewards/real": 9.493334770202637,
      "step": 460
    },
    {
      "epoch": 1.504,
      "grad_norm": 0.3303199018618516,
      "learning_rate": 2.7672209026128263e-07,
      "logits/generated": -0.8582679629325867,
      "logits/real": -1.4615408182144165,
      "logps/generated": -349.00909423828125,
      "logps/real": -104.46484375,
      "loss": 0.0966,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.171267986297607,
      "rewards/margins": 16.01020622253418,
      "rewards/real": 9.838939666748047,
      "step": 470
    },
    {
      "epoch": 1.536,
      "grad_norm": 0.3813272818464241,
      "learning_rate": 2.7078384798099764e-07,
      "logits/generated": -0.8334212303161621,
      "logits/real": -1.4117028713226318,
      "logps/generated": -349.4712829589844,
      "logps/real": -107.6310806274414,
      "loss": 0.1069,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.388515472412109,
      "rewards/margins": 16.422515869140625,
      "rewards/real": 10.034002304077148,
      "step": 480
    },
    {
      "epoch": 1.568,
      "grad_norm": 0.3427218965024406,
      "learning_rate": 2.648456057007126e-07,
      "logits/generated": -0.8402020335197449,
      "logits/real": -1.3591063022613525,
      "logps/generated": -349.86529541015625,
      "logps/real": -102.61138916015625,
      "loss": 0.1038,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.4038848876953125,
      "rewards/margins": 16.043880462646484,
      "rewards/real": 9.639993667602539,
      "step": 490
    },
    {
      "epoch": 1.5872000000000002,
      "eval_logits/generated": -0.8754745125770569,
      "eval_logits/real": -1.3614826202392578,
      "eval_logps/generated": -352.7068786621094,
      "eval_logps/real": -106.12812042236328,
      "eval_loss": 0.10832550376653671,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -7.003779888153076,
      "eval_rewards/margins": 17.076799392700195,
      "eval_rewards/real": 10.07302188873291,
      "eval_runtime": 33.3124,
      "eval_samples_per_second": 6.004,
      "eval_steps_per_second": 0.39,
      "step": 496
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.38914741214890197,
      "learning_rate": 2.589073634204275e-07,
      "logits/generated": -0.8448807597160339,
      "logits/real": -1.3226372003555298,
      "logps/generated": -351.1488037109375,
      "logps/real": -107.77713775634766,
      "loss": 0.1085,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.437747955322266,
      "rewards/margins": 16.31161117553711,
      "rewards/real": 9.87386417388916,
      "step": 500
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 0.3937890384115947,
      "learning_rate": 2.529691211401425e-07,
      "logits/generated": -0.8879863619804382,
      "logits/real": -1.3836486339569092,
      "logps/generated": -340.003662109375,
      "logps/real": -106.28749084472656,
      "loss": 0.1051,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.393700122833252,
      "rewards/margins": 16.35400390625,
      "rewards/real": 9.960305213928223,
      "step": 510
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 0.38541180062987296,
      "learning_rate": 2.4703087885985747e-07,
      "logits/generated": -0.8842099905014038,
      "logits/real": -1.3454530239105225,
      "logps/generated": -357.91693115234375,
      "logps/real": -115.9624252319336,
      "loss": 0.1108,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.042295932769775,
      "rewards/margins": 17.503582000732422,
      "rewards/real": 10.461287498474121,
      "step": 520
    },
    {
      "epoch": 1.696,
      "grad_norm": 0.3409231189845364,
      "learning_rate": 2.410926365795724e-07,
      "logits/generated": -0.8585386276245117,
      "logits/real": -1.3780685663223267,
      "logps/generated": -352.55780029296875,
      "logps/real": -104.75419616699219,
      "loss": 0.1091,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.910445213317871,
      "rewards/margins": 16.95829963684082,
      "rewards/real": 10.047855377197266,
      "step": 530
    },
    {
      "epoch": 1.728,
      "grad_norm": 0.36170229656289943,
      "learning_rate": 2.351543942992874e-07,
      "logits/generated": -0.8263759613037109,
      "logits/real": -1.477992057800293,
      "logps/generated": -342.4795227050781,
      "logps/real": -107.40743255615234,
      "loss": 0.1066,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.372612953186035,
      "rewards/margins": 16.573341369628906,
      "rewards/real": 10.200728416442871,
      "step": 540
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.3413340610553401,
      "learning_rate": 2.2921615201900234e-07,
      "logits/generated": -0.8356834650039673,
      "logits/real": -1.391550898551941,
      "logps/generated": -350.9759521484375,
      "logps/real": -100.52735137939453,
      "loss": 0.0996,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.971991539001465,
      "rewards/margins": 16.87733268737793,
      "rewards/real": 9.905343055725098,
      "step": 550
    },
    {
      "epoch": 1.7856,
      "eval_logits/generated": -0.8467180728912354,
      "eval_logits/real": -1.3430825471878052,
      "eval_logps/generated": -352.8455810546875,
      "eval_logps/real": -105.63909149169922,
      "eval_loss": 0.10793981701135635,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -7.0176496505737305,
      "eval_rewards/margins": 17.139570236206055,
      "eval_rewards/real": 10.121922492980957,
      "eval_runtime": 26.3578,
      "eval_samples_per_second": 7.588,
      "eval_steps_per_second": 0.493,
      "step": 558
    },
    {
      "epoch": 1.792,
      "grad_norm": 0.3508645377160027,
      "learning_rate": 2.2327790973871732e-07,
      "logits/generated": -0.8774440884590149,
      "logits/real": -1.3573105335235596,
      "logps/generated": -351.45697021484375,
      "logps/real": -119.85599517822266,
      "loss": 0.1099,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.625948905944824,
      "rewards/margins": 17.333721160888672,
      "rewards/real": 10.707772254943848,
      "step": 560
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 0.3654629141180253,
      "learning_rate": 2.173396674584323e-07,
      "logits/generated": -0.8835099339485168,
      "logits/real": -1.4046037197113037,
      "logps/generated": -346.0320739746094,
      "logps/real": -121.26094818115234,
      "loss": 0.1059,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.1836042404174805,
      "rewards/margins": 16.947559356689453,
      "rewards/real": 10.763954162597656,
      "step": 570
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 0.3990780890154353,
      "learning_rate": 2.1140142517814726e-07,
      "logits/generated": -0.8519056439399719,
      "logits/real": -1.4273611307144165,
      "logps/generated": -358.67572021484375,
      "logps/real": -100.94701385498047,
      "loss": 0.102,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.960443019866943,
      "rewards/margins": 16.782718658447266,
      "rewards/real": 9.822273254394531,
      "step": 580
    },
    {
      "epoch": 1.888,
      "grad_norm": 0.38514946196243705,
      "learning_rate": 2.0546318289786222e-07,
      "logits/generated": -0.8117479085922241,
      "logits/real": -1.3695944547653198,
      "logps/generated": -354.63836669921875,
      "logps/real": -92.74488830566406,
      "loss": 0.0995,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.354495048522949,
      "rewards/margins": 16.78274154663086,
      "rewards/real": 9.428245544433594,
      "step": 590
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.3924023599162714,
      "learning_rate": 1.9952494061757718e-07,
      "logits/generated": -0.8672367930412292,
      "logits/real": -1.42955482006073,
      "logps/generated": -358.61041259765625,
      "logps/real": -108.92237854003906,
      "loss": 0.1051,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.023502349853516,
      "rewards/margins": 17.467809677124023,
      "rewards/real": 10.444307327270508,
      "step": 600
    },
    {
      "epoch": 1.952,
      "grad_norm": 0.4306414835630299,
      "learning_rate": 1.9358669833729216e-07,
      "logits/generated": -0.8639044761657715,
      "logits/real": -1.4375250339508057,
      "logps/generated": -356.2347412109375,
      "logps/real": -110.20645904541016,
      "loss": 0.1031,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.728466987609863,
      "rewards/margins": 17.04363441467285,
      "rewards/real": 10.315167427062988,
      "step": 610
    },
    {
      "epoch": 1.984,
      "grad_norm": 0.3444950108876968,
      "learning_rate": 1.876484560570071e-07,
      "logits/generated": -0.9002432823181152,
      "logits/real": -1.4994289875030518,
      "logps/generated": -355.31854248046875,
      "logps/real": -111.2657699584961,
      "loss": 0.1058,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.687992095947266,
      "rewards/margins": 16.953479766845703,
      "rewards/real": 10.265488624572754,
      "step": 620
    },
    {
      "epoch": 1.984,
      "eval_logits/generated": -0.8820578455924988,
      "eval_logits/real": -1.4055497646331787,
      "eval_logps/generated": -357.4770202636719,
      "eval_logps/real": -105.37969207763672,
      "eval_loss": 0.10766133666038513,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -7.480791091918945,
      "eval_rewards/margins": 17.628650665283203,
      "eval_rewards/real": 10.14786148071289,
      "eval_runtime": 37.0785,
      "eval_samples_per_second": 5.394,
      "eval_steps_per_second": 0.351,
      "step": 620
    },
    {
      "epoch": 2.016,
      "grad_norm": 0.3780935929863943,
      "learning_rate": 1.8171021377672207e-07,
      "logits/generated": -0.8811729550361633,
      "logits/real": -1.3763405084609985,
      "logps/generated": -336.79595947265625,
      "logps/real": -113.61727142333984,
      "loss": 0.1078,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.133264064788818,
      "rewards/margins": 16.935718536376953,
      "rewards/real": 10.802453994750977,
      "step": 630
    },
    {
      "epoch": 2.048,
      "grad_norm": 0.3519668839255736,
      "learning_rate": 1.7577197149643706e-07,
      "logits/generated": -0.843653678894043,
      "logits/real": -1.4190059900283813,
      "logps/generated": -348.73944091796875,
      "logps/real": -98.12186431884766,
      "loss": 0.1021,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.104917049407959,
      "rewards/margins": 16.62804412841797,
      "rewards/real": 9.523126602172852,
      "step": 640
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.3518523871601281,
      "learning_rate": 1.6983372921615202e-07,
      "logits/generated": -0.8700889348983765,
      "logits/real": -1.4218956232070923,
      "logps/generated": -359.59912109375,
      "logps/real": -104.86639404296875,
      "loss": 0.0964,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.900285243988037,
      "rewards/margins": 17.489273071289062,
      "rewards/real": 10.588987350463867,
      "step": 650
    },
    {
      "epoch": 2.112,
      "grad_norm": 0.33640275650323853,
      "learning_rate": 1.6389548693586697e-07,
      "logits/generated": -0.9481474161148071,
      "logits/real": -1.4541842937469482,
      "logps/generated": -341.5894470214844,
      "logps/real": -103.59893798828125,
      "loss": 0.1018,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.416829586029053,
      "rewards/margins": 16.72159194946289,
      "rewards/real": 10.30476188659668,
      "step": 660
    },
    {
      "epoch": 2.144,
      "grad_norm": 0.3422223879953069,
      "learning_rate": 1.5795724465558193e-07,
      "logits/generated": -0.9395056962966919,
      "logits/real": -1.5020724534988403,
      "logps/generated": -336.66485595703125,
      "logps/real": -115.76351165771484,
      "loss": 0.0999,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.646719455718994,
      "rewards/margins": 17.59234619140625,
      "rewards/real": 10.945624351501465,
      "step": 670
    },
    {
      "epoch": 2.176,
      "grad_norm": 0.4031943667703452,
      "learning_rate": 1.520190023752969e-07,
      "logits/generated": -0.8811619877815247,
      "logits/real": -1.4061198234558105,
      "logps/generated": -358.12457275390625,
      "logps/real": -102.46736907958984,
      "loss": 0.0995,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.244171142578125,
      "rewards/margins": 17.505325317382812,
      "rewards/real": 10.261154174804688,
      "step": 680
    },
    {
      "epoch": 2.1824,
      "eval_logits/generated": -0.8780685067176819,
      "eval_logits/real": -1.4101550579071045,
      "eval_logps/generated": -354.6165771484375,
      "eval_logps/real": -105.18895721435547,
      "eval_loss": 0.10742899030447006,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -7.194748878479004,
      "eval_rewards/margins": 17.36168670654297,
      "eval_rewards/real": 10.166935920715332,
      "eval_runtime": 25.4941,
      "eval_samples_per_second": 7.845,
      "eval_steps_per_second": 0.51,
      "step": 682
    },
    {
      "epoch": 2.208,
      "grad_norm": 0.3658391972106121,
      "learning_rate": 1.4608076009501184e-07,
      "logits/generated": -0.7777714133262634,
      "logits/real": -1.4327569007873535,
      "logps/generated": -349.45123291015625,
      "logps/real": -91.0394287109375,
      "loss": 0.1006,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.516034126281738,
      "rewards/margins": 16.061874389648438,
      "rewards/real": 9.5458402633667,
      "step": 690
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.37751724197745956,
      "learning_rate": 1.4014251781472683e-07,
      "logits/generated": -0.8562400937080383,
      "logits/real": -1.4605350494384766,
      "logps/generated": -349.2660827636719,
      "logps/real": -110.1448745727539,
      "loss": 0.0985,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.469089508056641,
      "rewards/margins": 17.073518753051758,
      "rewards/real": 10.604430198669434,
      "step": 700
    },
    {
      "epoch": 2.2720000000000002,
      "grad_norm": 0.3099039440839268,
      "learning_rate": 1.342042755344418e-07,
      "logits/generated": -0.9007622003555298,
      "logits/real": -1.437908411026001,
      "logps/generated": -345.6163330078125,
      "logps/real": -93.06938171386719,
      "loss": 0.0978,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.0835700035095215,
      "rewards/margins": 15.84107780456543,
      "rewards/real": 9.757506370544434,
      "step": 710
    },
    {
      "epoch": 2.304,
      "grad_norm": 0.3864668793151262,
      "learning_rate": 1.2826603325415677e-07,
      "logits/generated": -0.8885849714279175,
      "logits/real": -1.5355165004730225,
      "logps/generated": -342.10443115234375,
      "logps/real": -104.58480072021484,
      "loss": 0.1019,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.677704811096191,
      "rewards/margins": 16.409557342529297,
      "rewards/real": 9.731854438781738,
      "step": 720
    },
    {
      "epoch": 2.336,
      "grad_norm": 0.3445053668290169,
      "learning_rate": 1.2232779097387173e-07,
      "logits/generated": -0.8385915756225586,
      "logits/real": -1.397671103477478,
      "logps/generated": -354.0185241699219,
      "logps/real": -112.39442443847656,
      "loss": 0.1017,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.097531795501709,
      "rewards/margins": 18.07855987548828,
      "rewards/real": 10.981030464172363,
      "step": 730
    },
    {
      "epoch": 2.368,
      "grad_norm": 0.3341439577424557,
      "learning_rate": 1.163895486935867e-07,
      "logits/generated": -0.8578903079032898,
      "logits/real": -1.383490800857544,
      "logps/generated": -347.2725524902344,
      "logps/real": -103.35150146484375,
      "loss": 0.1017,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.981736183166504,
      "rewards/margins": 17.021286010742188,
      "rewards/real": 10.039549827575684,
      "step": 740
    },
    {
      "epoch": 2.3808,
      "eval_logits/generated": -0.880645751953125,
      "eval_logits/real": -1.4227505922317505,
      "eval_logps/generated": -358.9116516113281,
      "eval_logps/real": -105.00930786132812,
      "eval_loss": 0.10732075572013855,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -7.624257564544678,
      "eval_rewards/margins": 17.80915641784668,
      "eval_rewards/real": 10.184901237487793,
      "eval_runtime": 36.3751,
      "eval_samples_per_second": 5.498,
      "eval_steps_per_second": 0.357,
      "step": 744
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.37872196207505543,
      "learning_rate": 1.1045130641330165e-07,
      "logits/generated": -0.9046470522880554,
      "logits/real": -1.3609508275985718,
      "logps/generated": -347.01611328125,
      "logps/real": -113.12504577636719,
      "loss": 0.1048,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.732972621917725,
      "rewards/margins": 17.482839584350586,
      "rewards/real": 10.74986457824707,
      "step": 750
    },
    {
      "epoch": 2.432,
      "grad_norm": 0.3456217830065947,
      "learning_rate": 1.0451306413301662e-07,
      "logits/generated": -0.9419130086898804,
      "logits/real": -1.448075294494629,
      "logps/generated": -350.6387023925781,
      "logps/real": -104.91294860839844,
      "loss": 0.1024,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.793516635894775,
      "rewards/margins": 17.103900909423828,
      "rewards/real": 10.310384750366211,
      "step": 760
    },
    {
      "epoch": 2.464,
      "grad_norm": 0.3510621706063929,
      "learning_rate": 9.857482185273158e-08,
      "logits/generated": -0.8172124028205872,
      "logits/real": -1.428694486618042,
      "logps/generated": -363.5953674316406,
      "logps/real": -88.23005676269531,
      "loss": 0.1012,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.76650857925415,
      "rewards/margins": 17.137920379638672,
      "rewards/real": 9.371413230895996,
      "step": 770
    },
    {
      "epoch": 2.496,
      "grad_norm": 0.29361560856209884,
      "learning_rate": 9.263657957244655e-08,
      "logits/generated": -0.8288412094116211,
      "logits/real": -1.4823163747787476,
      "logps/generated": -359.7073974609375,
      "logps/real": -97.23246002197266,
      "loss": 0.1021,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.1570844650268555,
      "rewards/margins": 17.029443740844727,
      "rewards/real": 9.872357368469238,
      "step": 780
    },
    {
      "epoch": 2.528,
      "grad_norm": 0.3698992846301525,
      "learning_rate": 8.669833729216151e-08,
      "logits/generated": -0.834811806678772,
      "logits/real": -1.4748550653457642,
      "logps/generated": -355.08685302734375,
      "logps/real": -91.9556884765625,
      "loss": 0.1018,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.218266487121582,
      "rewards/margins": 17.2064208984375,
      "rewards/real": 9.988153457641602,
      "step": 790
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.3614102743036081,
      "learning_rate": 8.076009501187649e-08,
      "logits/generated": -0.8415037989616394,
      "logits/real": -1.4521253108978271,
      "logps/generated": -342.2822265625,
      "logps/real": -102.14026641845703,
      "loss": 0.1031,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.008787631988525,
      "rewards/margins": 16.703975677490234,
      "rewards/real": 9.695188522338867,
      "step": 800
    },
    {
      "epoch": 2.5792,
      "eval_logits/generated": -0.878677248954773,
      "eval_logits/real": -1.4391472339630127,
      "eval_logps/generated": -359.2500305175781,
      "eval_logps/real": -104.75186920166016,
      "eval_loss": 0.1071779727935791,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -7.65809440612793,
      "eval_rewards/margins": 17.868741989135742,
      "eval_rewards/real": 10.210646629333496,
      "eval_runtime": 29.5912,
      "eval_samples_per_second": 6.759,
      "eval_steps_per_second": 0.439,
      "step": 806
    }
  ],
  "logging_steps": 10,
  "max_steps": 936,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 62,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}