{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 200,
  "global_step": 1563,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006397952655150352,
      "grad_norm": 412.2480368303884,
      "learning_rate": 3.1847133757961784e-09,
      "logits/generated": -2.53759503364563,
      "logits/real": -2.290781259536743,
      "logps/generated": -79.34207153320312,
      "logps/real": -142.87411499023438,
      "loss": 1.0838,
      "rewards/accuracies": 0.0,
      "rewards/generated": 0.0,
      "rewards/margins": 0.0,
      "rewards/real": 0.0,
      "step": 1
    },
    {
      "epoch": 0.006397952655150352,
      "grad_norm": 431.96820478477065,
      "learning_rate": 3.184713375796178e-08,
      "logits/generated": -2.5309884548187256,
      "logits/real": -2.523167371749878,
      "logps/generated": -88.78756713867188,
      "logps/real": -236.1966094970703,
      "loss": 1.1221,
      "rewards/accuracies": 0.5555555820465088,
      "rewards/generated": -0.010531743057072163,
      "rewards/margins": 0.02550431713461876,
      "rewards/real": 0.014972572214901447,
      "step": 10
    },
    {
      "epoch": 0.012795905310300703,
      "grad_norm": 299.862381030409,
      "learning_rate": 6.369426751592356e-08,
      "logits/generated": -2.578387498855591,
      "logits/real": -2.5629818439483643,
      "logps/generated": -90.43721771240234,
      "logps/real": -232.0134735107422,
      "loss": 0.9789,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/generated": -0.10627230256795883,
      "rewards/margins": 0.31521138548851013,
      "rewards/real": 0.2089390754699707,
      "step": 20
    },
    {
      "epoch": 0.019193857965451054,
      "grad_norm": 128.80078242264278,
      "learning_rate": 9.554140127388536e-08,
      "logits/generated": -2.639655590057373,
      "logits/real": -2.623234748840332,
      "logps/generated": -81.7702865600586,
      "logps/real": -192.8977813720703,
      "loss": 0.732,
      "rewards/accuracies": 1.0,
      "rewards/generated": -0.4113238453865051,
      "rewards/margins": 1.0968698263168335,
      "rewards/real": 0.6855460405349731,
      "step": 30
    },
    {
      "epoch": 0.025591810620601407,
      "grad_norm": 53.90372491881725,
      "learning_rate": 1.2738853503184713e-07,
      "logits/generated": -2.754666566848755,
      "logits/real": -2.771085262298584,
      "logps/generated": -93.06287384033203,
      "logps/real": -208.916015625,
      "loss": 0.551,
      "rewards/accuracies": 1.0,
      "rewards/generated": -0.8528655171394348,
      "rewards/margins": 2.697784423828125,
      "rewards/real": 1.844918966293335,
      "step": 40
    },
    {
      "epoch": 0.03198976327575176,
      "grad_norm": 36.32945064394617,
      "learning_rate": 1.592356687898089e-07,
      "logits/generated": -2.875619649887085,
      "logits/real": -2.8736255168914795,
      "logps/generated": -102.63410949707031,
      "logps/real": -185.98129272460938,
      "loss": 0.4387,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -1.5957601070404053,
      "rewards/margins": 4.849114894866943,
      "rewards/real": 3.253354549407959,
      "step": 50
    },
    {
      "epoch": 0.03838771593090211,
      "grad_norm": 22.26335616683524,
      "learning_rate": 1.9108280254777072e-07,
      "logits/generated": -2.9478697776794434,
      "logits/real": -3.007082223892212,
      "logps/generated": -110.52901458740234,
      "logps/real": -171.8012237548828,
      "loss": 0.3865,
      "rewards/accuracies": 1.0,
      "rewards/generated": -1.786839485168457,
      "rewards/margins": 6.540804862976074,
      "rewards/real": 4.753964900970459,
      "step": 60
    },
    {
      "epoch": 0.044785668586052464,
      "grad_norm": 18.72255550731988,
      "learning_rate": 2.2292993630573247e-07,
      "logits/generated": -2.9968416690826416,
      "logits/real": -3.001485824584961,
      "logps/generated": -107.23085021972656,
      "logps/real": -146.6729736328125,
      "loss": 0.3283,
      "rewards/accuracies": 1.0,
      "rewards/generated": -1.93621826171875,
      "rewards/margins": 8.847390174865723,
      "rewards/real": 6.911171913146973,
      "step": 70
    },
    {
      "epoch": 0.05118362124120281,
      "grad_norm": 9.385567995761765,
      "learning_rate": 2.5477707006369425e-07,
      "logits/generated": -2.985713481903076,
      "logits/real": -3.0017921924591064,
      "logps/generated": -111.2695083618164,
      "logps/real": -128.3047332763672,
      "loss": 0.2744,
      "rewards/accuracies": 1.0,
      "rewards/generated": -2.9947776794433594,
      "rewards/margins": 10.760424613952637,
      "rewards/real": 7.765646934509277,
      "step": 80
    },
    {
      "epoch": 0.05758157389635317,
      "grad_norm": 4.574434721884733,
      "learning_rate": 2.86624203821656e-07,
      "logits/generated": -2.93389630317688,
      "logits/real": -2.9760384559631348,
      "logps/generated": -124.31734466552734,
      "logps/real": -127.97233581542969,
      "loss": 0.269,
      "rewards/accuracies": 1.0,
      "rewards/generated": -3.8817825317382812,
      "rewards/margins": 12.403055191040039,
      "rewards/real": 8.521272659301758,
      "step": 90
    },
    {
      "epoch": 0.06397952655150352,
      "grad_norm": 9.260926951901052,
      "learning_rate": 3.184713375796178e-07,
      "logits/generated": -2.948780059814453,
      "logits/real": -2.924206495285034,
      "logps/generated": -117.82643127441406,
      "logps/real": -140.2669219970703,
      "loss": 0.2787,
      "rewards/accuracies": 1.0,
      "rewards/generated": -3.5896174907684326,
      "rewards/margins": 13.11004638671875,
      "rewards/real": 9.520429611206055,
      "step": 100
    },
    {
      "epoch": 0.07037747920665387,
      "grad_norm": 5.655236437132318,
      "learning_rate": 3.5031847133757957e-07,
      "logits/generated": -2.9382705688476562,
      "logits/real": -2.895289421081543,
      "logps/generated": -111.1765365600586,
      "logps/real": -122.95256042480469,
      "loss": 0.2607,
      "rewards/accuracies": 1.0,
      "rewards/generated": -3.181370258331299,
      "rewards/margins": 12.501322746276855,
      "rewards/real": 9.319952011108398,
      "step": 110
    },
    {
      "epoch": 0.07677543186180422,
      "grad_norm": 10.474178085979151,
      "learning_rate": 3.8216560509554143e-07,
      "logits/generated": -2.9170148372650146,
      "logits/real": -2.9494102001190186,
      "logps/generated": -120.6019515991211,
      "logps/real": -120.84870910644531,
      "loss": 0.2583,
      "rewards/accuracies": 1.0,
      "rewards/generated": -3.854489803314209,
      "rewards/margins": 13.146855354309082,
      "rewards/real": 9.292365074157715,
      "step": 120
    },
    {
      "epoch": 0.08317338451695458,
      "grad_norm": 3.425127673225565,
      "learning_rate": 4.140127388535032e-07,
      "logits/generated": -2.962573528289795,
      "logits/real": -2.9198384284973145,
      "logps/generated": -119.30030822753906,
      "logps/real": -114.26997375488281,
      "loss": 0.2475,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.209897041320801,
      "rewards/margins": 13.073972702026367,
      "rewards/real": 8.864076614379883,
      "step": 130
    },
    {
      "epoch": 0.08957133717210493,
      "grad_norm": 2.8766539348394065,
      "learning_rate": 4.4585987261146494e-07,
      "logits/generated": -2.902782440185547,
      "logits/real": -2.8874409198760986,
      "logps/generated": -138.11538696289062,
      "logps/real": -128.03103637695312,
      "loss": 0.2566,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.331695556640625,
      "rewards/margins": 15.117105484008789,
      "rewards/real": 9.785411834716797,
      "step": 140
    },
    {
      "epoch": 0.09596928982725528,
      "grad_norm": 2.928535512875351,
      "learning_rate": 4.777070063694267e-07,
      "logits/generated": -2.9390759468078613,
      "logits/real": -2.9156928062438965,
      "logps/generated": -128.9432373046875,
      "logps/real": -112.00594329833984,
      "loss": 0.2422,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.531386375427246,
      "rewards/margins": 13.208353042602539,
      "rewards/real": 8.676966667175293,
      "step": 150
    },
    {
      "epoch": 0.10236724248240563,
      "grad_norm": 4.646412917226807,
      "learning_rate": 4.989331436699858e-07,
      "logits/generated": -2.907299041748047,
      "logits/real": -2.8829636573791504,
      "logps/generated": -147.1040802001953,
      "logps/real": -130.88232421875,
      "loss": 0.2552,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.514917373657227,
      "rewards/margins": 16.280481338500977,
      "rewards/real": 10.76556396484375,
      "step": 160
    },
    {
      "epoch": 0.10876519513755598,
      "grad_norm": 2.67790411455109,
      "learning_rate": 4.953769559032717e-07,
      "logits/generated": -2.8610851764678955,
      "logits/real": -2.931917667388916,
      "logps/generated": -134.75003051757812,
      "logps/real": -93.88365173339844,
      "loss": 0.2215,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.963212013244629,
      "rewards/margins": 13.223382949829102,
      "rewards/real": 8.260170936584473,
      "step": 170
    },
    {
      "epoch": 0.11516314779270634,
      "grad_norm": 3.0194042444553317,
      "learning_rate": 4.918207681365576e-07,
      "logits/generated": -2.8236758708953857,
      "logits/real": -2.895191192626953,
      "logps/generated": -128.0216522216797,
      "logps/real": -105.7012939453125,
      "loss": 0.2328,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.834359645843506,
      "rewards/margins": 14.000758171081543,
      "rewards/real": 9.166399002075195,
      "step": 180
    },
    {
      "epoch": 0.12156110044785669,
      "grad_norm": 7.731581783616942,
      "learning_rate": 4.882645803698435e-07,
      "logits/generated": -2.880075693130493,
      "logits/real": -2.863041400909424,
      "logps/generated": -129.73480224609375,
      "logps/real": -121.33485412597656,
      "loss": 0.2407,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -4.882136344909668,
      "rewards/margins": 15.376531600952148,
      "rewards/real": 10.49439525604248,
      "step": 190
    },
    {
      "epoch": 0.12795905310300704,
      "grad_norm": 6.783378094196307,
      "learning_rate": 4.847083926031294e-07,
      "logits/generated": -2.8695290088653564,
      "logits/real": -2.863180637359619,
      "logps/generated": -143.60906982421875,
      "logps/real": -90.62736511230469,
      "loss": 0.2262,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.7865214347839355,
      "rewards/margins": 13.516904830932617,
      "rewards/real": 7.73038387298584,
      "step": 200
    },
    {
      "epoch": 0.12795905310300704,
      "eval_logits/generated": -2.833332061767578,
      "eval_logits/real": -2.8265442848205566,
      "eval_logps/generated": -126.3611831665039,
      "eval_logps/real": -117.04948425292969,
      "eval_loss": 0.2435438483953476,
      "eval_rewards/accuracies": 0.9921875,
      "eval_rewards/generated": -4.7644171714782715,
      "eval_rewards/margins": 14.327579498291016,
      "eval_rewards/real": 9.563161849975586,
      "eval_runtime": 57.2388,
      "eval_samples_per_second": 8.735,
      "eval_steps_per_second": 0.28,
      "step": 200
    },
    {
      "epoch": 0.1343570057581574,
      "grad_norm": 4.630294431619136,
      "learning_rate": 4.811522048364154e-07,
      "logits/generated": -2.848015546798706,
      "logits/real": -2.878896474838257,
      "logps/generated": -134.64739990234375,
      "logps/real": -130.23609924316406,
      "loss": 0.2535,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.333014965057373,
      "rewards/margins": 16.449054718017578,
      "rewards/real": 11.11603832244873,
      "step": 210
    },
    {
      "epoch": 0.14075495841330773,
      "grad_norm": 3.37570780842826,
      "learning_rate": 4.775960170697012e-07,
      "logits/generated": -2.9035279750823975,
      "logits/real": -2.8899435997009277,
      "logps/generated": -159.3094024658203,
      "logps/real": -124.04632568359375,
      "loss": 0.243,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.095156669616699,
      "rewards/margins": 16.67739486694336,
      "rewards/real": 10.58223819732666,
      "step": 220
    },
    {
      "epoch": 0.1471529110684581,
      "grad_norm": 2.451067761798054,
      "learning_rate": 4.7403982930298717e-07,
      "logits/generated": -2.87707257270813,
      "logits/real": -2.901421546936035,
      "logps/generated": -120.0963134765625,
      "logps/real": -123.16734313964844,
      "loss": 0.2359,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.565489768981934,
      "rewards/margins": 15.016607284545898,
      "rewards/real": 10.451117515563965,
      "step": 230
    },
    {
      "epoch": 0.15355086372360843,
      "grad_norm": 2.3458265316121145,
      "learning_rate": 4.7048364153627306e-07,
      "logits/generated": -2.845377206802368,
      "logits/real": -2.9469656944274902,
      "logps/generated": -130.70401000976562,
      "logps/real": -102.62498474121094,
      "loss": 0.2292,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.917192459106445,
      "rewards/margins": 13.963025093078613,
      "rewards/real": 9.045831680297852,
      "step": 240
    },
    {
      "epoch": 0.1599488163787588,
      "grad_norm": 2.2252235869873425,
      "learning_rate": 4.66927453769559e-07,
      "logits/generated": -2.814479351043701,
      "logits/real": -2.79661226272583,
      "logps/generated": -129.10992431640625,
      "logps/real": -110.62039947509766,
      "loss": 0.2418,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.840092182159424,
      "rewards/margins": 14.007853507995605,
      "rewards/real": 9.16776180267334,
      "step": 250
    },
    {
      "epoch": 0.16634676903390916,
      "grad_norm": 3.2120162920530775,
      "learning_rate": 4.633712660028449e-07,
      "logits/generated": -2.7691586017608643,
      "logits/real": -2.803241729736328,
      "logps/generated": -130.2401885986328,
      "logps/real": -103.78974914550781,
      "loss": 0.2207,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.814304828643799,
      "rewards/margins": 14.570037841796875,
      "rewards/real": 9.755732536315918,
      "step": 260
    },
    {
      "epoch": 0.1727447216890595,
      "grad_norm": 2.265151763557376,
      "learning_rate": 4.5981507823613085e-07,
      "logits/generated": -2.7566680908203125,
      "logits/real": -2.747866153717041,
      "logps/generated": -152.24049377441406,
      "logps/real": -107.55720520019531,
      "loss": 0.2249,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -6.19102144241333,
      "rewards/margins": 15.924577713012695,
      "rewards/real": 9.733556747436523,
      "step": 270
    },
    {
      "epoch": 0.17914267434420986,
      "grad_norm": 3.5602666832695844,
      "learning_rate": 4.562588904694168e-07,
      "logits/generated": -2.7723679542541504,
      "logits/real": -2.785216808319092,
      "logps/generated": -128.1282958984375,
      "logps/real": -111.1799545288086,
      "loss": 0.2202,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.806519508361816,
      "rewards/margins": 15.082513809204102,
      "rewards/real": 10.275992393493652,
      "step": 280
    },
    {
      "epoch": 0.1855406269993602,
      "grad_norm": 3.11519890787563,
      "learning_rate": 4.5270270270270264e-07,
      "logits/generated": -2.757876396179199,
      "logits/real": -2.795422077178955,
      "logps/generated": -128.11813354492188,
      "logps/real": -102.88358306884766,
      "loss": 0.233,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -4.833106994628906,
      "rewards/margins": 14.29919147491455,
      "rewards/real": 9.466082572937012,
      "step": 290
    },
    {
      "epoch": 0.19193857965451055,
      "grad_norm": 3.487329720349146,
      "learning_rate": 4.491465149359886e-07,
      "logits/generated": -2.8109488487243652,
      "logits/real": -2.7747206687927246,
      "logps/generated": -129.84274291992188,
      "logps/real": -106.18394470214844,
      "loss": 0.2286,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.749394416809082,
      "rewards/margins": 14.446981430053711,
      "rewards/real": 9.697588920593262,
      "step": 300
    },
    {
      "epoch": 0.19833653230966092,
      "grad_norm": 4.236344907363098,
      "learning_rate": 4.4559032716927454e-07,
      "logits/generated": -2.7962634563446045,
      "logits/real": -2.86857533454895,
      "logps/generated": -137.8242645263672,
      "logps/real": -126.29652404785156,
      "loss": 0.2454,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.464461803436279,
      "rewards/margins": 16.495359420776367,
      "rewards/real": 11.030899047851562,
      "step": 310
    },
    {
      "epoch": 0.20473448496481125,
      "grad_norm": 2.0886040022985206,
      "learning_rate": 4.420341394025605e-07,
      "logits/generated": -2.780796766281128,
      "logits/real": -2.830658197402954,
      "logps/generated": -128.07977294921875,
      "logps/real": -102.74151611328125,
      "loss": 0.2217,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -4.719244956970215,
      "rewards/margins": 13.948654174804688,
      "rewards/real": 9.229410171508789,
      "step": 320
    },
    {
      "epoch": 0.21113243761996162,
      "grad_norm": 2.7732179589224737,
      "learning_rate": 4.384779516358463e-07,
      "logits/generated": -2.771653175354004,
      "logits/real": -2.719320058822632,
      "logps/generated": -128.0609130859375,
      "logps/real": -110.12910461425781,
      "loss": 0.2298,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.775327682495117,
      "rewards/margins": 14.936683654785156,
      "rewards/real": 10.161355972290039,
      "step": 330
    },
    {
      "epoch": 0.21753039027511195,
      "grad_norm": 2.4432792911747327,
      "learning_rate": 4.3492176386913227e-07,
      "logits/generated": -2.7587437629699707,
      "logits/real": -2.802103042602539,
      "logps/generated": -152.65858459472656,
      "logps/real": -112.84379577636719,
      "loss": 0.2312,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.072720527648926,
      "rewards/margins": 16.01877212524414,
      "rewards/real": 9.946051597595215,
      "step": 340
    },
    {
      "epoch": 0.22392834293026231,
      "grad_norm": 4.160225941183556,
      "learning_rate": 4.313655761024182e-07,
      "logits/generated": -2.7678041458129883,
      "logits/real": -2.7790327072143555,
      "logps/generated": -120.19953918457031,
      "logps/real": -113.22270202636719,
      "loss": 0.2164,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.654033660888672,
      "rewards/margins": 14.861625671386719,
      "rewards/real": 10.207592010498047,
      "step": 350
    },
    {
      "epoch": 0.23032629558541268,
      "grad_norm": 2.2403711981241106,
      "learning_rate": 4.278093883357041e-07,
      "logits/generated": -2.836555004119873,
      "logits/real": -2.6989388465881348,
      "logps/generated": -133.00775146484375,
      "logps/real": -108.97412109375,
      "loss": 0.2254,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/generated": -4.826836109161377,
      "rewards/margins": 14.98741340637207,
      "rewards/real": 10.160576820373535,
      "step": 360
    },
    {
      "epoch": 0.236724248240563,
      "grad_norm": 3.08941181412325,
      "learning_rate": 4.2425320056899e-07,
      "logits/generated": -2.7016754150390625,
      "logits/real": -2.7019612789154053,
      "logps/generated": -125.4333724975586,
      "logps/real": -107.02604675292969,
      "loss": 0.2184,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.93725061416626,
      "rewards/margins": 14.800321578979492,
      "rewards/real": 9.86307144165039,
      "step": 370
    },
    {
      "epoch": 0.24312220089571338,
      "grad_norm": 8.811415551257076,
      "learning_rate": 4.2069701280227595e-07,
      "logits/generated": -2.7188172340393066,
      "logits/real": -2.777225971221924,
      "logps/generated": -131.77468872070312,
      "logps/real": -97.81922912597656,
      "loss": 0.2132,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.891592979431152,
      "rewards/margins": 14.796887397766113,
      "rewards/real": 9.905295372009277,
      "step": 380
    },
    {
      "epoch": 0.2495201535508637,
      "grad_norm": 2.1656082848995992,
      "learning_rate": 4.1714082503556185e-07,
      "logits/generated": -2.81782865524292,
      "logits/real": -2.733973264694214,
      "logps/generated": -125.90080261230469,
      "logps/real": -108.9613037109375,
      "loss": 0.208,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.474228858947754,
      "rewards/margins": 14.704130172729492,
      "rewards/real": 10.229901313781738,
      "step": 390
    },
    {
      "epoch": 0.2559181062060141,
      "grad_norm": 2.1913448139958307,
      "learning_rate": 4.135846372688478e-07,
      "logits/generated": -2.741250514984131,
      "logits/real": -2.797882080078125,
      "logps/generated": -140.27984619140625,
      "logps/real": -113.54410552978516,
      "loss": 0.2141,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.6347832679748535,
      "rewards/margins": 16.515918731689453,
      "rewards/real": 10.881135940551758,
      "step": 400
    },
    {
      "epoch": 0.2559181062060141,
      "eval_logits/generated": -2.7752091884613037,
      "eval_logits/real": -2.76131010055542,
      "eval_logps/generated": -128.1855010986328,
      "eval_logps/real": -113.70217895507812,
      "eval_loss": 0.23566634953022003,
      "eval_rewards/accuracies": 0.9921875,
      "eval_rewards/generated": -4.946849822998047,
      "eval_rewards/margins": 14.844742774963379,
      "eval_rewards/real": 9.897892951965332,
      "eval_runtime": 70.5415,
      "eval_samples_per_second": 7.088,
      "eval_steps_per_second": 0.227,
      "step": 400
    },
    {
      "epoch": 0.26231605886116444,
      "grad_norm": 3.5382231043936563,
      "learning_rate": 4.100284495021337e-07,
      "logits/generated": -2.7606406211853027,
      "logits/real": -2.8516087532043457,
      "logps/generated": -150.7864990234375,
      "logps/real": -107.80538177490234,
      "loss": 0.222,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.30859899520874,
      "rewards/margins": 16.771678924560547,
      "rewards/real": 10.463083267211914,
      "step": 410
    },
    {
      "epoch": 0.2687140115163148,
      "grad_norm": 2.180636822570479,
      "learning_rate": 4.064722617354196e-07,
      "logits/generated": -2.777144193649292,
      "logits/real": -2.7791004180908203,
      "logps/generated": -141.62130737304688,
      "logps/real": -111.6278305053711,
      "loss": 0.2179,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.845305442810059,
      "rewards/margins": 16.107940673828125,
      "rewards/real": 10.262636184692383,
      "step": 420
    },
    {
      "epoch": 0.2751119641714651,
      "grad_norm": 7.826230059811001,
      "learning_rate": 4.0291607396870553e-07,
      "logits/generated": -2.7701034545898438,
      "logits/real": -2.835702657699585,
      "logps/generated": -134.1055450439453,
      "logps/real": -115.15937805175781,
      "loss": 0.2277,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.376823425292969,
      "rewards/margins": 16.15007972717285,
      "rewards/real": 10.773252487182617,
      "step": 430
    },
    {
      "epoch": 0.28150991682661547,
      "grad_norm": 4.574173821690705,
      "learning_rate": 3.993598862019915e-07,
      "logits/generated": -2.7749392986297607,
      "logits/real": -2.82127046585083,
      "logps/generated": -133.71063232421875,
      "logps/real": -103.6177978515625,
      "loss": 0.2214,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.1843953132629395,
      "rewards/margins": 14.904150009155273,
      "rewards/real": 9.719755172729492,
      "step": 440
    },
    {
      "epoch": 0.28790786948176583,
      "grad_norm": 2.774800684163752,
      "learning_rate": 3.9580369843527737e-07,
      "logits/generated": -2.744696617126465,
      "logits/real": -2.7654058933258057,
      "logps/generated": -157.8831024169922,
      "logps/real": -103.58988952636719,
      "loss": 0.2138,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.751193046569824,
      "rewards/margins": 17.019277572631836,
      "rewards/real": 10.268084526062012,
      "step": 450
    },
    {
      "epoch": 0.2943058221369162,
      "grad_norm": 5.160429480281925,
      "learning_rate": 3.9224751066856327e-07,
      "logits/generated": -2.7339484691619873,
      "logits/real": -2.685930013656616,
      "logps/generated": -135.38714599609375,
      "logps/real": -114.84906005859375,
      "loss": 0.2204,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.209853649139404,
      "rewards/margins": 15.920175552368164,
      "rewards/real": 10.710321426391602,
      "step": 460
    },
    {
      "epoch": 0.30070377479206656,
      "grad_norm": 4.706717543122211,
      "learning_rate": 3.886913229018492e-07,
      "logits/generated": -2.7286899089813232,
      "logits/real": -2.782683849334717,
      "logps/generated": -146.5428924560547,
      "logps/real": -119.84122467041016,
      "loss": 0.2176,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.839329719543457,
      "rewards/margins": 16.770212173461914,
      "rewards/real": 10.930883407592773,
      "step": 470
    },
    {
      "epoch": 0.30710172744721687,
      "grad_norm": 2.3058413282374715,
      "learning_rate": 3.851351351351351e-07,
      "logits/generated": -2.7684266567230225,
      "logits/real": -2.802767276763916,
      "logps/generated": -138.41696166992188,
      "logps/real": -113.2447738647461,
      "loss": 0.2321,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.6439032554626465,
      "rewards/margins": 15.897249221801758,
      "rewards/real": 10.253345489501953,
      "step": 480
    },
    {
      "epoch": 0.31349968010236723,
      "grad_norm": 3.3424369618748706,
      "learning_rate": 3.8157894736842105e-07,
      "logits/generated": -2.766950845718384,
      "logits/real": -2.794663906097412,
      "logps/generated": -123.19486236572266,
      "logps/real": -98.98766326904297,
      "loss": 0.2199,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.739457130432129,
      "rewards/margins": 14.55401611328125,
      "rewards/real": 9.814558029174805,
      "step": 490
    },
    {
      "epoch": 0.3198976327575176,
      "grad_norm": 2.169410849214599,
      "learning_rate": 3.7802275960170695e-07,
      "logits/generated": -2.715301513671875,
      "logits/real": -2.8804736137390137,
      "logps/generated": -140.2711181640625,
      "logps/real": -100.17481994628906,
      "loss": 0.2348,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -5.455787658691406,
      "rewards/margins": 15.202703475952148,
      "rewards/real": 9.746915817260742,
      "step": 500
    },
    {
      "epoch": 0.32629558541266795,
      "grad_norm": 2.04238252540844,
      "learning_rate": 3.7446657183499284e-07,
      "logits/generated": -2.7904624938964844,
      "logits/real": -2.7905654907226562,
      "logps/generated": -145.4995574951172,
      "logps/real": -121.39151763916016,
      "loss": 0.2262,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.898383140563965,
      "rewards/margins": 17.675220489501953,
      "rewards/real": 11.776837348937988,
      "step": 510
    },
    {
      "epoch": 0.3326935380678183,
      "grad_norm": 2.4330841754647823,
      "learning_rate": 3.709103840682788e-07,
      "logits/generated": -2.728691577911377,
      "logits/real": -2.8865561485290527,
      "logps/generated": -156.2779998779297,
      "logps/real": -108.7221450805664,
      "loss": 0.2081,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.528653144836426,
      "rewards/margins": 16.937610626220703,
      "rewards/real": 10.408956527709961,
      "step": 520
    },
    {
      "epoch": 0.3390914907229686,
      "grad_norm": 3.3742934939891174,
      "learning_rate": 3.6735419630156474e-07,
      "logits/generated": -2.819417953491211,
      "logits/real": -2.816087245941162,
      "logps/generated": -153.2522430419922,
      "logps/real": -133.01565551757812,
      "loss": 0.2314,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.505625247955322,
      "rewards/margins": 18.493694305419922,
      "rewards/real": 11.988069534301758,
      "step": 530
    },
    {
      "epoch": 0.345489443378119,
      "grad_norm": 2.5651527495210242,
      "learning_rate": 3.637980085348506e-07,
      "logits/generated": -2.7839221954345703,
      "logits/real": -2.799818515777588,
      "logps/generated": -126.00923156738281,
      "logps/real": -103.7330093383789,
      "loss": 0.2194,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.9756646156311035,
      "rewards/margins": 15.352749824523926,
      "rewards/real": 10.377084732055664,
      "step": 540
    },
    {
      "epoch": 0.35188739603326935,
      "grad_norm": 2.096110953513035,
      "learning_rate": 3.602418207681365e-07,
      "logits/generated": -2.7390735149383545,
      "logits/real": -2.818681240081787,
      "logps/generated": -123.09765625,
      "logps/real": -97.98001861572266,
      "loss": 0.211,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -4.679843902587891,
      "rewards/margins": 14.732965469360352,
      "rewards/real": 10.053121566772461,
      "step": 550
    },
    {
      "epoch": 0.3582853486884197,
      "grad_norm": 2.3805224123873767,
      "learning_rate": 3.5668563300142247e-07,
      "logits/generated": -2.74855899810791,
      "logits/real": -2.8764424324035645,
      "logps/generated": -138.00978088378906,
      "logps/real": -104.4416275024414,
      "loss": 0.2092,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.503247261047363,
      "rewards/margins": 15.430496215820312,
      "rewards/real": 9.927248001098633,
      "step": 560
    },
    {
      "epoch": 0.3646833013435701,
      "grad_norm": 2.9036557884699996,
      "learning_rate": 3.5312944523470837e-07,
      "logits/generated": -2.8142824172973633,
      "logits/real": -2.8175225257873535,
      "logps/generated": -138.83673095703125,
      "logps/real": -105.87809753417969,
      "loss": 0.2188,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.465325355529785,
      "rewards/margins": 15.622749328613281,
      "rewards/real": 10.157423973083496,
      "step": 570
    },
    {
      "epoch": 0.3710812539987204,
      "grad_norm": 2.1880847814800948,
      "learning_rate": 3.495732574679943e-07,
      "logits/generated": -2.8173575401306152,
      "logits/real": -2.8300251960754395,
      "logps/generated": -149.59030151367188,
      "logps/real": -106.2435302734375,
      "loss": 0.2234,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.7458953857421875,
      "rewards/margins": 16.33285140991211,
      "rewards/real": 10.586955070495605,
      "step": 580
    },
    {
      "epoch": 0.37747920665387075,
      "grad_norm": 2.192659377942886,
      "learning_rate": 3.460170697012802e-07,
      "logits/generated": -2.777217388153076,
      "logits/real": -2.863020181655884,
      "logps/generated": -129.68231201171875,
      "logps/real": -108.95207214355469,
      "loss": 0.2224,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.835152626037598,
      "rewards/margins": 15.670806884765625,
      "rewards/real": 10.835652351379395,
      "step": 590
    },
    {
      "epoch": 0.3838771593090211,
      "grad_norm": 12.154950352869962,
      "learning_rate": 3.424608819345661e-07,
      "logits/generated": -2.823272228240967,
      "logits/real": -2.7826805114746094,
      "logps/generated": -117.91121673583984,
      "logps/real": -111.25467681884766,
      "loss": 0.2089,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.435954570770264,
      "rewards/margins": 14.716769218444824,
      "rewards/real": 10.280814170837402,
      "step": 600
    },
    {
      "epoch": 0.3838771593090211,
      "eval_logits/generated": -2.7913894653320312,
      "eval_logits/real": -2.798403024673462,
      "eval_logps/generated": -127.67301940917969,
      "eval_logps/real": -112.43653106689453,
      "eval_loss": 0.23413488268852234,
      "eval_rewards/accuracies": 0.9921875,
      "eval_rewards/generated": -4.895601272583008,
      "eval_rewards/margins": 14.920058250427246,
      "eval_rewards/real": 10.024457931518555,
      "eval_runtime": 70.7535,
      "eval_samples_per_second": 7.067,
      "eval_steps_per_second": 0.226,
      "step": 600
    },
    {
      "epoch": 0.3902751119641715,
      "grad_norm": 2.0157880028714548,
      "learning_rate": 3.3890469416785205e-07,
      "logits/generated": -2.7878739833831787,
      "logits/real": -2.886143445968628,
      "logps/generated": -132.29232788085938,
      "logps/real": -107.32408142089844,
      "loss": 0.2185,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.14271354675293,
      "rewards/margins": 15.367979049682617,
      "rewards/real": 10.225264549255371,
      "step": 610
    },
    {
      "epoch": 0.39667306461932184,
      "grad_norm": 3.4761506983717623,
      "learning_rate": 3.35348506401138e-07,
      "logits/generated": -2.800938129425049,
      "logits/real": -2.9011588096618652,
      "logps/generated": -152.92681884765625,
      "logps/real": -90.30223083496094,
      "loss": 0.2051,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.535437107086182,
      "rewards/margins": 15.690652847290039,
      "rewards/real": 9.155214309692383,
      "step": 620
    },
    {
      "epoch": 0.40307101727447214,
      "grad_norm": 2.066415328730325,
      "learning_rate": 3.3179231863442384e-07,
      "logits/generated": -2.7363858222961426,
      "logits/real": -2.9205002784729004,
      "logps/generated": -139.44544982910156,
      "logps/real": -93.95768737792969,
      "loss": 0.2054,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -5.3508195877075195,
      "rewards/margins": 14.589759826660156,
      "rewards/real": 9.23893928527832,
      "step": 630
    },
    {
      "epoch": 0.4094689699296225,
      "grad_norm": 2.720280779674223,
      "learning_rate": 3.282361308677098e-07,
      "logits/generated": -2.717585325241089,
      "logits/real": -2.8081445693969727,
      "logps/generated": -139.2246551513672,
      "logps/real": -95.35455322265625,
      "loss": 0.2105,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.649844169616699,
      "rewards/margins": 15.598958015441895,
      "rewards/real": 9.949111938476562,
      "step": 640
    },
    {
      "epoch": 0.41586692258477287,
      "grad_norm": 2.4467415453795534,
      "learning_rate": 3.2467994310099573e-07,
      "logits/generated": -2.746037244796753,
      "logits/real": -2.7760324478149414,
      "logps/generated": -151.04168701171875,
      "logps/real": -117.99039459228516,
      "loss": 0.228,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.320367336273193,
      "rewards/margins": 17.100696563720703,
      "rewards/real": 10.780328750610352,
      "step": 650
    },
    {
      "epoch": 0.42226487523992323,
      "grad_norm": 2.5369736094086752,
      "learning_rate": 3.211237553342817e-07,
      "logits/generated": -2.7743470668792725,
      "logits/real": -2.9016330242156982,
      "logps/generated": -169.43215942382812,
      "logps/real": -110.16139221191406,
      "loss": 0.2213,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.090481758117676,
      "rewards/margins": 17.460203170776367,
      "rewards/real": 10.369722366333008,
      "step": 660
    },
    {
      "epoch": 0.4286628278950736,
      "grad_norm": 3.940886446614688,
      "learning_rate": 3.175675675675675e-07,
      "logits/generated": -2.7578988075256348,
      "logits/real": -2.7422776222229004,
      "logps/generated": -142.6200408935547,
      "logps/real": -109.90318298339844,
      "loss": 0.2226,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.739054203033447,
      "rewards/margins": 16.534833908081055,
      "rewards/real": 10.79577922821045,
      "step": 670
    },
    {
      "epoch": 0.4350607805502239,
      "grad_norm": 1.9885549537383231,
      "learning_rate": 3.1401137980085347e-07,
      "logits/generated": -2.753906011581421,
      "logits/real": -2.8121540546417236,
      "logps/generated": -142.36412048339844,
      "logps/real": -95.9502182006836,
      "loss": 0.2135,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.0002851486206055,
      "rewards/margins": 15.877609252929688,
      "rewards/real": 9.877325057983398,
      "step": 680
    },
    {
      "epoch": 0.44145873320537427,
      "grad_norm": 2.360303864176274,
      "learning_rate": 3.104551920341394e-07,
      "logits/generated": -2.7917466163635254,
      "logits/real": -2.8809592723846436,
      "logps/generated": -128.33108520507812,
      "logps/real": -109.88875579833984,
      "loss": 0.2091,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.061340808868408,
      "rewards/margins": 15.903111457824707,
      "rewards/real": 10.841771125793457,
      "step": 690
    },
    {
      "epoch": 0.44785668586052463,
      "grad_norm": 2.2232483398700884,
      "learning_rate": 3.068990042674253e-07,
      "logits/generated": -2.7540788650512695,
      "logits/real": -2.8440656661987305,
      "logps/generated": -137.61160278320312,
      "logps/real": -110.4582290649414,
      "loss": 0.2235,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.629530906677246,
      "rewards/margins": 16.784801483154297,
      "rewards/real": 11.155269622802734,
      "step": 700
    },
    {
      "epoch": 0.454254638515675,
      "grad_norm": 2.420484904502962,
      "learning_rate": 3.033428165007112e-07,
      "logits/generated": -2.7791481018066406,
      "logits/real": -2.823207378387451,
      "logps/generated": -130.23995971679688,
      "logps/real": -104.56422424316406,
      "loss": 0.2045,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.046719551086426,
      "rewards/margins": 15.610448837280273,
      "rewards/real": 10.563729286193848,
      "step": 710
    },
    {
      "epoch": 0.46065259117082535,
      "grad_norm": 1.8814105476658205,
      "learning_rate": 2.9978662873399715e-07,
      "logits/generated": -2.754664897918701,
      "logits/real": -2.868396043777466,
      "logps/generated": -138.9653778076172,
      "logps/real": -104.86256408691406,
      "loss": 0.2139,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.563291549682617,
      "rewards/margins": 16.40620994567871,
      "rewards/real": 10.842918395996094,
      "step": 720
    },
    {
      "epoch": 0.46705054382597566,
      "grad_norm": 1.9123445490996223,
      "learning_rate": 2.9623044096728305e-07,
      "logits/generated": -2.7599892616271973,
      "logits/real": -2.836871862411499,
      "logps/generated": -143.76820373535156,
      "logps/real": -101.16252136230469,
      "loss": 0.2134,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.618950843811035,
      "rewards/margins": 16.395917892456055,
      "rewards/real": 10.776965141296387,
      "step": 730
    },
    {
      "epoch": 0.473448496481126,
      "grad_norm": 1.9685905302445135,
      "learning_rate": 2.92674253200569e-07,
      "logits/generated": -2.7747621536254883,
      "logits/real": -2.7902274131774902,
      "logps/generated": -118.72608947753906,
      "logps/real": -107.44352722167969,
      "loss": 0.2035,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.259701728820801,
      "rewards/margins": 15.15577220916748,
      "rewards/real": 10.89607048034668,
      "step": 740
    },
    {
      "epoch": 0.4798464491362764,
      "grad_norm": 2.185389542016207,
      "learning_rate": 2.8911806543385494e-07,
      "logits/generated": -2.694744348526001,
      "logits/real": -2.7872262001037598,
      "logps/generated": -117.7671890258789,
      "logps/real": -92.30455017089844,
      "loss": 0.2175,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -4.346129894256592,
      "rewards/margins": 13.698951721191406,
      "rewards/real": 9.352822303771973,
      "step": 750
    },
    {
      "epoch": 0.48624440179142675,
      "grad_norm": 2.0382414250965732,
      "learning_rate": 2.855618776671408e-07,
      "logits/generated": -2.7974352836608887,
      "logits/real": -2.8290600776672363,
      "logps/generated": -132.22039794921875,
      "logps/real": -104.40059661865234,
      "loss": 0.2157,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -5.511082649230957,
      "rewards/margins": 15.783149719238281,
      "rewards/real": 10.272066116333008,
      "step": 760
    },
    {
      "epoch": 0.4926423544465771,
      "grad_norm": 2.1388688577679016,
      "learning_rate": 2.8200568990042673e-07,
      "logits/generated": -2.7909514904022217,
      "logits/real": -2.808551073074341,
      "logps/generated": -151.1422119140625,
      "logps/real": -116.09953308105469,
      "loss": 0.2271,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.148834228515625,
      "rewards/margins": 17.4340877532959,
      "rewards/real": 11.28525447845459,
      "step": 770
    },
    {
      "epoch": 0.4990403071017274,
      "grad_norm": 2.2656673226538717,
      "learning_rate": 2.784495021337127e-07,
      "logits/generated": -2.7903616428375244,
      "logits/real": -2.8122215270996094,
      "logps/generated": -138.20358276367188,
      "logps/real": -104.3212661743164,
      "loss": 0.2234,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -5.353945732116699,
      "rewards/margins": 16.125553131103516,
      "rewards/real": 10.771608352661133,
      "step": 780
    },
    {
      "epoch": 0.5054382597568778,
      "grad_norm": 2.1885026864804624,
      "learning_rate": 2.7489331436699857e-07,
      "logits/generated": -2.8039913177490234,
      "logits/real": -2.7904744148254395,
      "logps/generated": -129.17857360839844,
      "logps/real": -115.23606872558594,
      "loss": 0.2302,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.161337852478027,
      "rewards/margins": 16.69219398498535,
      "rewards/real": 11.530855178833008,
      "step": 790
    },
    {
      "epoch": 0.5118362124120281,
      "grad_norm": 2.869890153550306,
      "learning_rate": 2.7133712660028446e-07,
      "logits/generated": -2.773207426071167,
      "logits/real": -2.7846627235412598,
      "logps/generated": -119.42828369140625,
      "logps/real": -115.4027099609375,
      "loss": 0.2148,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.340595722198486,
      "rewards/margins": 15.677358627319336,
      "rewards/real": 11.336763381958008,
      "step": 800
    },
    {
      "epoch": 0.5118362124120281,
      "eval_logits/generated": -2.8194947242736816,
      "eval_logits/real": -2.823777437210083,
      "eval_logps/generated": -129.62098693847656,
      "eval_logps/real": -112.27095031738281,
      "eval_loss": 0.23086096346378326,
      "eval_rewards/accuracies": 0.9921875,
      "eval_rewards/generated": -5.09039831161499,
      "eval_rewards/margins": 15.131412506103516,
      "eval_rewards/real": 10.041014671325684,
      "eval_runtime": 69.2574,
      "eval_samples_per_second": 7.219,
      "eval_steps_per_second": 0.231,
      "step": 800
    },
    {
      "epoch": 0.5182341650671785,
      "grad_norm": 2.356636174793149,
      "learning_rate": 2.677809388335704e-07,
      "logits/generated": -2.8063623905181885,
      "logits/real": -2.863706111907959,
      "logps/generated": -147.79519653320312,
      "logps/real": -103.06642150878906,
      "loss": 0.2038,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.835212707519531,
      "rewards/margins": 16.04536247253418,
      "rewards/real": 10.210151672363281,
      "step": 810
    },
    {
      "epoch": 0.5246321177223289,
      "grad_norm": 2.4379730238495885,
      "learning_rate": 2.642247510668563e-07,
      "logits/generated": -2.8116323947906494,
      "logits/real": -2.8085341453552246,
      "logps/generated": -135.0247039794922,
      "logps/real": -88.69074249267578,
      "loss": 0.2247,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.124009132385254,
      "rewards/margins": 14.631864547729492,
      "rewards/real": 9.507856369018555,
      "step": 820
    },
    {
      "epoch": 0.5310300703774792,
      "grad_norm": 2.4726280945236083,
      "learning_rate": 2.6066856330014225e-07,
      "logits/generated": -2.7714428901672363,
      "logits/real": -2.796095132827759,
      "logps/generated": -138.78175354003906,
      "logps/real": -103.1611099243164,
      "loss": 0.2042,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.705328941345215,
      "rewards/margins": 16.654027938842773,
      "rewards/real": 10.948700904846191,
      "step": 830
    },
    {
      "epoch": 0.5374280230326296,
      "grad_norm": 2.0930312705666965,
      "learning_rate": 2.5711237553342815e-07,
      "logits/generated": -2.796297311782837,
      "logits/real": -2.761157512664795,
      "logps/generated": -132.76963806152344,
      "logps/real": -101.53390502929688,
      "loss": 0.2154,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.048023223876953,
      "rewards/margins": 16.279930114746094,
      "rewards/real": 11.231904983520508,
      "step": 840
    },
    {
      "epoch": 0.5438259756877799,
      "grad_norm": 1.995146822930404,
      "learning_rate": 2.5355618776671404e-07,
      "logits/generated": -2.74373722076416,
      "logits/real": -2.8401942253112793,
      "logps/generated": -134.88999938964844,
      "logps/real": -108.44051361083984,
      "loss": 0.2085,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.090542793273926,
      "rewards/margins": 16.1528377532959,
      "rewards/real": 11.06229305267334,
      "step": 850
    },
    {
      "epoch": 0.5502239283429302,
      "grad_norm": 2.383171664852739,
      "learning_rate": 2.5e-07,
      "logits/generated": -2.734670639038086,
      "logits/real": -2.836339235305786,
      "logps/generated": -138.05905151367188,
      "logps/real": -96.95460510253906,
      "loss": 0.2,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.488851070404053,
      "rewards/margins": 15.819448471069336,
      "rewards/real": 10.330598831176758,
      "step": 860
    },
    {
      "epoch": 0.5566218809980806,
      "grad_norm": 2.20520607110751,
      "learning_rate": 2.4644381223328594e-07,
      "logits/generated": -2.7709481716156006,
      "logits/real": -2.7831263542175293,
      "logps/generated": -137.1595916748047,
      "logps/real": -114.11800384521484,
      "loss": 0.2189,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.659204959869385,
      "rewards/margins": 17.742206573486328,
      "rewards/real": 12.083002090454102,
      "step": 870
    },
    {
      "epoch": 0.5630198336532309,
      "grad_norm": 2.6353918937836847,
      "learning_rate": 2.4288762446657183e-07,
      "logits/generated": -2.708103656768799,
      "logits/real": -2.895563840866089,
      "logps/generated": -141.7598876953125,
      "logps/real": -99.23255920410156,
      "loss": 0.2069,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.729107856750488,
      "rewards/margins": 15.888574600219727,
      "rewards/real": 10.159466743469238,
      "step": 880
    },
    {
      "epoch": 0.5694177863083814,
      "grad_norm": 2.159987018302931,
      "learning_rate": 2.393314366998578e-07,
      "logits/generated": -2.8029890060424805,
      "logits/real": -2.899193048477173,
      "logps/generated": -142.26734924316406,
      "logps/real": -89.63228607177734,
      "loss": 0.2086,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.678102493286133,
      "rewards/margins": 15.533955574035645,
      "rewards/real": 9.855853080749512,
      "step": 890
    },
    {
      "epoch": 0.5758157389635317,
      "grad_norm": 2.829485679186667,
      "learning_rate": 2.3577524893314365e-07,
      "logits/generated": -2.8933000564575195,
      "logits/real": -2.8898868560791016,
      "logps/generated": -140.3587646484375,
      "logps/real": -119.44731140136719,
      "loss": 0.2126,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.4967241287231445,
      "rewards/margins": 17.632753372192383,
      "rewards/real": 12.136028289794922,
      "step": 900
    },
    {
      "epoch": 0.582213691618682,
      "grad_norm": 2.076889704737332,
      "learning_rate": 2.322190611664296e-07,
      "logits/generated": -2.783379077911377,
      "logits/real": -2.837111473083496,
      "logps/generated": -159.20602416992188,
      "logps/real": -101.35649108886719,
      "loss": 0.2152,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.505612850189209,
      "rewards/margins": 16.421142578125,
      "rewards/real": 9.91553020477295,
      "step": 910
    },
    {
      "epoch": 0.5886116442738324,
      "grad_norm": 2.9561242755966917,
      "learning_rate": 2.2866287339971549e-07,
      "logits/generated": -2.7626872062683105,
      "logits/real": -2.775630235671997,
      "logps/generated": -142.0345916748047,
      "logps/real": -105.5779037475586,
      "loss": 0.2106,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.930171012878418,
      "rewards/margins": 16.78856086730957,
      "rewards/real": 10.858389854431152,
      "step": 920
    },
    {
      "epoch": 0.5950095969289827,
      "grad_norm": 5.343288980733804,
      "learning_rate": 2.251066856330014e-07,
      "logits/generated": -2.863882064819336,
      "logits/real": -2.7736942768096924,
      "logps/generated": -146.3633575439453,
      "logps/real": -116.1565170288086,
      "loss": 0.2372,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.034173965454102,
      "rewards/margins": 17.2874755859375,
      "rewards/real": 11.253301620483398,
      "step": 930
    },
    {
      "epoch": 0.6014075495841331,
      "grad_norm": 2.1700141539005418,
      "learning_rate": 2.2155049786628733e-07,
      "logits/generated": -2.7782211303710938,
      "logits/real": -2.784595012664795,
      "logps/generated": -126.44002532958984,
      "logps/real": -98.55274200439453,
      "loss": 0.2121,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.89090633392334,
      "rewards/margins": 15.318072319030762,
      "rewards/real": 10.427166938781738,
      "step": 940
    },
    {
      "epoch": 0.6078055022392834,
      "grad_norm": 2.086474199547169,
      "learning_rate": 2.1799431009957325e-07,
      "logits/generated": -2.7844066619873047,
      "logits/real": -2.87266206741333,
      "logps/generated": -154.29930114746094,
      "logps/real": -118.48374938964844,
      "loss": 0.2242,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.924124717712402,
      "rewards/margins": 17.47787857055664,
      "rewards/real": 11.553754806518555,
      "step": 950
    },
    {
      "epoch": 0.6142034548944337,
      "grad_norm": 5.893270475169604,
      "learning_rate": 2.1443812233285914e-07,
      "logits/generated": -2.7291111946105957,
      "logits/real": -2.7862040996551514,
      "logps/generated": -128.32461547851562,
      "logps/real": -88.90164947509766,
      "loss": 0.2049,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -5.07694149017334,
      "rewards/margins": 14.397714614868164,
      "rewards/real": 9.320773124694824,
      "step": 960
    },
    {
      "epoch": 0.6206014075495841,
      "grad_norm": 3.4745504874973854,
      "learning_rate": 2.108819345661451e-07,
      "logits/generated": -2.7670676708221436,
      "logits/real": -2.906858205795288,
      "logps/generated": -142.1813507080078,
      "logps/real": -99.36964416503906,
      "loss": 0.2152,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.543498992919922,
      "rewards/margins": 15.908276557922363,
      "rewards/real": 10.364778518676758,
      "step": 970
    },
    {
      "epoch": 0.6269993602047345,
      "grad_norm": 1.9536154527758394,
      "learning_rate": 2.0732574679943098e-07,
      "logits/generated": -2.7182888984680176,
      "logits/real": -2.795628070831299,
      "logps/generated": -123.13966369628906,
      "logps/real": -92.99507141113281,
      "loss": 0.2048,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.866164207458496,
      "rewards/margins": 14.955586433410645,
      "rewards/real": 10.089422225952148,
      "step": 980
    },
    {
      "epoch": 0.6333973128598849,
      "grad_norm": 2.3702688907183425,
      "learning_rate": 2.0376955903271693e-07,
      "logits/generated": -2.792490243911743,
      "logits/real": -2.814042568206787,
      "logps/generated": -122.5984878540039,
      "logps/real": -119.72834777832031,
      "loss": 0.2154,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.933380603790283,
      "rewards/margins": 16.830997467041016,
      "rewards/real": 11.897615432739258,
      "step": 990
    },
    {
      "epoch": 0.6397952655150352,
      "grad_norm": 2.710438187714284,
      "learning_rate": 2.0021337126600283e-07,
      "logits/generated": -2.7410390377044678,
      "logits/real": -2.817781925201416,
      "logps/generated": -161.62539672851562,
      "logps/real": -91.64434051513672,
      "loss": 0.1994,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.895718574523926,
      "rewards/margins": 17.067646026611328,
      "rewards/real": 10.171927452087402,
      "step": 1000
    },
    {
      "epoch": 0.6397952655150352,
      "eval_logits/generated": -2.74424147605896,
      "eval_logits/real": -2.7461187839508057,
      "eval_logps/generated": -130.59327697753906,
      "eval_logps/real": -111.54972839355469,
      "eval_loss": 0.23031392693519592,
      "eval_rewards/accuracies": 0.9921875,
      "eval_rewards/generated": -5.187626838684082,
      "eval_rewards/margins": 15.300764083862305,
      "eval_rewards/real": 10.113137245178223,
      "eval_runtime": 69.5849,
      "eval_samples_per_second": 7.185,
      "eval_steps_per_second": 0.23,
      "step": 1000
    },
    {
      "epoch": 0.6461932181701855,
      "grad_norm": 2.2354079100751845,
      "learning_rate": 1.9665718349928875e-07,
      "logits/generated": -2.7386083602905273,
      "logits/real": -2.7921862602233887,
      "logps/generated": -145.14804077148438,
      "logps/real": -100.52366638183594,
      "loss": 0.1945,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.059700012207031,
      "rewards/margins": 16.424558639526367,
      "rewards/real": 10.364858627319336,
      "step": 1010
    },
    {
      "epoch": 0.6525911708253359,
      "grad_norm": 2.1844813037214124,
      "learning_rate": 1.931009957325747e-07,
      "logits/generated": -2.7453203201293945,
      "logits/real": -2.812004327774048,
      "logps/generated": -147.8553466796875,
      "logps/real": -100.11280822753906,
      "loss": 0.2096,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.994841575622559,
      "rewards/margins": 15.893636703491211,
      "rewards/real": 9.898796081542969,
      "step": 1020
    },
    {
      "epoch": 0.6589891234804862,
      "grad_norm": 2.046398564222724,
      "learning_rate": 1.895448079658606e-07,
      "logits/generated": -2.7719972133636475,
      "logits/real": -2.837329387664795,
      "logps/generated": -141.85464477539062,
      "logps/real": -95.72025299072266,
      "loss": 0.1927,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.702435493469238,
      "rewards/margins": 15.670855522155762,
      "rewards/real": 9.968420028686523,
      "step": 1030
    },
    {
      "epoch": 0.6653870761356366,
      "grad_norm": 2.2879001250497297,
      "learning_rate": 1.859886201991465e-07,
      "logits/generated": -2.708482265472412,
      "logits/real": -2.903043270111084,
      "logps/generated": -139.8241729736328,
      "logps/real": -103.22698974609375,
      "loss": 0.2013,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.350104331970215,
      "rewards/margins": 16.140804290771484,
      "rewards/real": 10.790700912475586,
      "step": 1040
    },
    {
      "epoch": 0.6717850287907869,
      "grad_norm": 4.628698290056213,
      "learning_rate": 1.8243243243243243e-07,
      "logits/generated": -2.748556137084961,
      "logits/real": -2.8339285850524902,
      "logps/generated": -148.6436004638672,
      "logps/real": -114.01179504394531,
      "loss": 0.2241,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.570124626159668,
      "rewards/margins": 17.3154354095459,
      "rewards/real": 11.745311737060547,
      "step": 1050
    },
    {
      "epoch": 0.6781829814459372,
      "grad_norm": 3.6247485539345417,
      "learning_rate": 1.7887624466571835e-07,
      "logits/generated": -2.8411951065063477,
      "logits/real": -2.7839694023132324,
      "logps/generated": -123.47579193115234,
      "logps/real": -112.1775131225586,
      "loss": 0.2218,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.763791561126709,
      "rewards/margins": 16.458126068115234,
      "rewards/real": 11.694334030151367,
      "step": 1060
    },
    {
      "epoch": 0.6845809341010877,
      "grad_norm": 2.2522203971012487,
      "learning_rate": 1.7532005689900424e-07,
      "logits/generated": -2.7332136631011963,
      "logits/real": -2.804201602935791,
      "logps/generated": -142.97860717773438,
      "logps/real": -85.4603271484375,
      "loss": 0.1947,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.868603706359863,
      "rewards/margins": 15.239709854125977,
      "rewards/real": 9.37110424041748,
      "step": 1070
    },
    {
      "epoch": 0.690978886756238,
      "grad_norm": 2.2382250321484225,
      "learning_rate": 1.717638691322902e-07,
      "logits/generated": -2.7342700958251953,
      "logits/real": -2.8574981689453125,
      "logps/generated": -151.73580932617188,
      "logps/real": -103.33231353759766,
      "loss": 0.2102,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.019370079040527,
      "rewards/margins": 16.717998504638672,
      "rewards/real": 10.698628425598145,
      "step": 1080
    },
    {
      "epoch": 0.6973768394113884,
      "grad_norm": 4.541077368968895,
      "learning_rate": 1.6820768136557609e-07,
      "logits/generated": -2.6836509704589844,
      "logits/real": -2.7782435417175293,
      "logps/generated": -140.19631958007812,
      "logps/real": -103.10017395019531,
      "loss": 0.2103,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.432912349700928,
      "rewards/margins": 16.41040802001953,
      "rewards/real": 10.977496147155762,
      "step": 1090
    },
    {
      "epoch": 0.7037747920665387,
      "grad_norm": 2.134995135769877,
      "learning_rate": 1.64651493598862e-07,
      "logits/generated": -2.700814962387085,
      "logits/real": -2.7614195346832275,
      "logps/generated": -139.90142822265625,
      "logps/real": -101.66877746582031,
      "loss": 0.2131,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -5.645236015319824,
      "rewards/margins": 16.219711303710938,
      "rewards/real": 10.57447624206543,
      "step": 1100
    },
    {
      "epoch": 0.710172744721689,
      "grad_norm": 2.1362568113809277,
      "learning_rate": 1.6109530583214793e-07,
      "logits/generated": -2.7405338287353516,
      "logits/real": -2.801772117614746,
      "logps/generated": -135.68980407714844,
      "logps/real": -115.05977630615234,
      "loss": 0.2147,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.29919958114624,
      "rewards/margins": 17.55141258239746,
      "rewards/real": 12.252215385437012,
      "step": 1110
    },
    {
      "epoch": 0.7165706973768394,
      "grad_norm": 2.6687487627386965,
      "learning_rate": 1.5753911806543385e-07,
      "logits/generated": -2.655336380004883,
      "logits/real": -2.716346502304077,
      "logps/generated": -138.2339630126953,
      "logps/real": -106.83113098144531,
      "loss": 0.2236,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.799440860748291,
      "rewards/margins": 16.477214813232422,
      "rewards/real": 10.677773475646973,
      "step": 1120
    },
    {
      "epoch": 0.7229686500319897,
      "grad_norm": 1.9796488466759263,
      "learning_rate": 1.5398293029871974e-07,
      "logits/generated": -2.7700233459472656,
      "logits/real": -2.719053268432617,
      "logps/generated": -138.2145233154297,
      "logps/real": -101.98434448242188,
      "loss": 0.1982,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.574553966522217,
      "rewards/margins": 15.94153118133545,
      "rewards/real": 10.366975784301758,
      "step": 1130
    },
    {
      "epoch": 0.7293666026871402,
      "grad_norm": 1.97652173855478,
      "learning_rate": 1.504267425320057e-07,
      "logits/generated": -2.749962091445923,
      "logits/real": -2.7098100185394287,
      "logps/generated": -128.74168395996094,
      "logps/real": -103.87724304199219,
      "loss": 0.203,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.09400749206543,
      "rewards/margins": 16.894054412841797,
      "rewards/real": 11.800048828125,
      "step": 1140
    },
    {
      "epoch": 0.7357645553422905,
      "grad_norm": 1.9276505695865507,
      "learning_rate": 1.4687055476529158e-07,
      "logits/generated": -2.7056586742401123,
      "logits/real": -2.84588885307312,
      "logps/generated": -132.11521911621094,
      "logps/real": -103.04747009277344,
      "loss": 0.2,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.32986307144165,
      "rewards/margins": 16.5743408203125,
      "rewards/real": 11.244477272033691,
      "step": 1150
    },
    {
      "epoch": 0.7421625079974408,
      "grad_norm": 2.4003697136809587,
      "learning_rate": 1.4331436699857753e-07,
      "logits/generated": -2.7689154148101807,
      "logits/real": -2.7181622982025146,
      "logps/generated": -128.201171875,
      "logps/real": -111.31385803222656,
      "loss": 0.2257,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.305639266967773,
      "rewards/margins": 16.98516273498535,
      "rewards/real": 11.679524421691895,
      "step": 1160
    },
    {
      "epoch": 0.7485604606525912,
      "grad_norm": 2.1662792343990898,
      "learning_rate": 1.3975817923186345e-07,
      "logits/generated": -2.7471747398376465,
      "logits/real": -2.745327949523926,
      "logps/generated": -131.89920043945312,
      "logps/real": -101.08728790283203,
      "loss": 0.2126,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -5.3999738693237305,
      "rewards/margins": 16.271684646606445,
      "rewards/real": 10.871710777282715,
      "step": 1170
    },
    {
      "epoch": 0.7549584133077415,
      "grad_norm": 1.9618990548812567,
      "learning_rate": 1.3620199146514935e-07,
      "logits/generated": -2.73884654045105,
      "logits/real": -2.792114734649658,
      "logps/generated": -139.20669555664062,
      "logps/real": -111.1391830444336,
      "loss": 0.2157,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.806328296661377,
      "rewards/margins": 17.363449096679688,
      "rewards/real": 11.557123184204102,
      "step": 1180
    },
    {
      "epoch": 0.7613563659628919,
      "grad_norm": 1.8565236814707453,
      "learning_rate": 1.326458036984353e-07,
      "logits/generated": -2.708221673965454,
      "logits/real": -2.8019871711730957,
      "logps/generated": -141.60244750976562,
      "logps/real": -103.365966796875,
      "loss": 0.2066,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -5.380032539367676,
      "rewards/margins": 15.911615371704102,
      "rewards/real": 10.53158187866211,
      "step": 1190
    },
    {
      "epoch": 0.7677543186180422,
      "grad_norm": 5.832771744417194,
      "learning_rate": 1.290896159317212e-07,
      "logits/generated": -2.7190277576446533,
      "logits/real": -2.782357692718506,
      "logps/generated": -133.46572875976562,
      "logps/real": -102.03229522705078,
      "loss": 0.2075,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.16973876953125,
      "rewards/margins": 16.539508819580078,
      "rewards/real": 11.369768142700195,
      "step": 1200
    },
    {
      "epoch": 0.7677543186180422,
      "eval_logits/generated": -2.7359671592712402,
      "eval_logits/real": -2.7372217178344727,
      "eval_logps/generated": -128.395751953125,
      "eval_logps/real": -111.5260009765625,
      "eval_loss": 0.23038950562477112,
      "eval_rewards/accuracies": 0.9921875,
      "eval_rewards/generated": -4.9678730964660645,
      "eval_rewards/margins": 15.083383560180664,
      "eval_rewards/real": 10.115509986877441,
      "eval_runtime": 69.835,
      "eval_samples_per_second": 7.16,
      "eval_steps_per_second": 0.229,
      "step": 1200
    },
    {
      "epoch": 0.7741522712731925,
      "grad_norm": 2.3644426695343914,
      "learning_rate": 1.255334281650071e-07,
      "logits/generated": -2.702448606491089,
      "logits/real": -2.849297046661377,
      "logps/generated": -133.25411987304688,
      "logps/real": -94.89964294433594,
      "loss": 0.2192,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.127745151519775,
      "rewards/margins": 15.297076225280762,
      "rewards/real": 10.169332504272461,
      "step": 1210
    },
    {
      "epoch": 0.780550223928343,
      "grad_norm": 4.2920815806287695,
      "learning_rate": 1.2197724039829303e-07,
      "logits/generated": -2.8339786529541016,
      "logits/real": -2.757995128631592,
      "logps/generated": -148.0692138671875,
      "logps/real": -121.7199478149414,
      "loss": 0.2159,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.0756330490112305,
      "rewards/margins": 18.304834365844727,
      "rewards/real": 12.229202270507812,
      "step": 1220
    },
    {
      "epoch": 0.7869481765834933,
      "grad_norm": 2.624742549146953,
      "learning_rate": 1.1842105263157894e-07,
      "logits/generated": -2.7471704483032227,
      "logits/real": -2.8416290283203125,
      "logps/generated": -147.3047637939453,
      "logps/real": -111.00262451171875,
      "loss": 0.2119,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.8865203857421875,
      "rewards/margins": 17.451772689819336,
      "rewards/real": 11.565254211425781,
      "step": 1230
    },
    {
      "epoch": 0.7933461292386437,
      "grad_norm": 2.1307989909566256,
      "learning_rate": 1.1486486486486487e-07,
      "logits/generated": -2.731872797012329,
      "logits/real": -2.8878135681152344,
      "logps/generated": -172.3035430908203,
      "logps/real": -110.19361877441406,
      "loss": 0.2202,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.301609039306641,
      "rewards/margins": 18.403263092041016,
      "rewards/real": 11.101654052734375,
      "step": 1240
    },
    {
      "epoch": 0.799744081893794,
      "grad_norm": 2.18810616018797,
      "learning_rate": 1.1130867709815078e-07,
      "logits/generated": -2.635159969329834,
      "logits/real": -2.8312065601348877,
      "logps/generated": -154.04550170898438,
      "logps/real": -104.9063949584961,
      "loss": 0.195,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.582805633544922,
      "rewards/margins": 17.949687957763672,
      "rewards/real": 11.366880416870117,
      "step": 1250
    },
    {
      "epoch": 0.8061420345489443,
      "grad_norm": 2.3243034253667,
      "learning_rate": 1.077524893314367e-07,
      "logits/generated": -2.7570877075195312,
      "logits/real": -2.676670789718628,
      "logps/generated": -132.24571228027344,
      "logps/real": -110.49552154541016,
      "loss": 0.2048,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.188769340515137,
      "rewards/margins": 16.681943893432617,
      "rewards/real": 11.49317455291748,
      "step": 1260
    },
    {
      "epoch": 0.8125399872040947,
      "grad_norm": 2.22786920352928,
      "learning_rate": 1.0419630156472262e-07,
      "logits/generated": -2.7040412425994873,
      "logits/real": -2.7337164878845215,
      "logps/generated": -129.7920379638672,
      "logps/real": -102.24781799316406,
      "loss": 0.2098,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.8722944259643555,
      "rewards/margins": 16.266225814819336,
      "rewards/real": 11.393930435180664,
      "step": 1270
    },
    {
      "epoch": 0.818937939859245,
      "grad_norm": 2.000001121079601,
      "learning_rate": 1.0064011379800854e-07,
      "logits/generated": -2.724949359893799,
      "logits/real": -2.7215514183044434,
      "logps/generated": -146.55990600585938,
      "logps/real": -114.5284423828125,
      "loss": 0.2211,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.995248794555664,
      "rewards/margins": 17.621685028076172,
      "rewards/real": 11.626437187194824,
      "step": 1280
    },
    {
      "epoch": 0.8253358925143954,
      "grad_norm": 2.220536708583519,
      "learning_rate": 9.708392603129445e-08,
      "logits/generated": -2.6592986583709717,
      "logits/real": -2.766054391860962,
      "logps/generated": -141.36831665039062,
      "logps/real": -96.46757507324219,
      "loss": 0.2112,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.908102989196777,
      "rewards/margins": 16.13393783569336,
      "rewards/real": 10.225834846496582,
      "step": 1290
    },
    {
      "epoch": 0.8317338451695457,
      "grad_norm": 1.8458396373240298,
      "learning_rate": 9.352773826458037e-08,
      "logits/generated": -2.7686662673950195,
      "logits/real": -2.6808652877807617,
      "logps/generated": -147.9801788330078,
      "logps/real": -109.27278900146484,
      "loss": 0.2098,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.193856716156006,
      "rewards/margins": 17.859333038330078,
      "rewards/real": 11.66547679901123,
      "step": 1300
    },
    {
      "epoch": 0.838131797824696,
      "grad_norm": 2.2735605764303304,
      "learning_rate": 8.997155049786629e-08,
      "logits/generated": -2.791935443878174,
      "logits/real": -2.7119717597961426,
      "logps/generated": -136.2453155517578,
      "logps/real": -101.1696548461914,
      "loss": 0.2048,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.072253227233887,
      "rewards/margins": 15.385726928710938,
      "rewards/real": 10.313471794128418,
      "step": 1310
    },
    {
      "epoch": 0.8445297504798465,
      "grad_norm": 7.320467978107467,
      "learning_rate": 8.64153627311522e-08,
      "logits/generated": -2.716169834136963,
      "logits/real": -2.7271907329559326,
      "logps/generated": -142.0037384033203,
      "logps/real": -96.90984344482422,
      "loss": 0.2039,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.719296455383301,
      "rewards/margins": 16.5465087890625,
      "rewards/real": 10.827211380004883,
      "step": 1320
    },
    {
      "epoch": 0.8509277031349968,
      "grad_norm": 2.1090442333698927,
      "learning_rate": 8.285917496443812e-08,
      "logits/generated": -2.7019193172454834,
      "logits/real": -2.7879014015197754,
      "logps/generated": -131.5638885498047,
      "logps/real": -103.13008880615234,
      "loss": 0.2036,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.254150390625,
      "rewards/margins": 15.576313972473145,
      "rewards/real": 10.322164535522461,
      "step": 1330
    },
    {
      "epoch": 0.8573256557901472,
      "grad_norm": 1.9764553854622655,
      "learning_rate": 7.930298719772404e-08,
      "logits/generated": -2.7741990089416504,
      "logits/real": -2.786623954772949,
      "logps/generated": -128.60775756835938,
      "logps/real": -95.55816650390625,
      "loss": 0.2071,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.881311893463135,
      "rewards/margins": 15.2656888961792,
      "rewards/real": 10.384378433227539,
      "step": 1340
    },
    {
      "epoch": 0.8637236084452975,
      "grad_norm": 2.4449760593366765,
      "learning_rate": 7.574679943100994e-08,
      "logits/generated": -2.7581117153167725,
      "logits/real": -2.8711161613464355,
      "logps/generated": -159.32579040527344,
      "logps/real": -127.8974838256836,
      "loss": 0.2075,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.493968963623047,
      "rewards/margins": 19.491052627563477,
      "rewards/real": 12.997082710266113,
      "step": 1350
    },
    {
      "epoch": 0.8701215611004478,
      "grad_norm": 1.8244220811357807,
      "learning_rate": 7.219061166429587e-08,
      "logits/generated": -2.740255832672119,
      "logits/real": -2.8236019611358643,
      "logps/generated": -140.54568481445312,
      "logps/real": -108.64134216308594,
      "loss": 0.1994,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.692780494689941,
      "rewards/margins": 16.945453643798828,
      "rewards/real": 11.25267219543457,
      "step": 1360
    },
    {
      "epoch": 0.8765195137555982,
      "grad_norm": 2.9020057472557528,
      "learning_rate": 6.863442389758179e-08,
      "logits/generated": -2.748126745223999,
      "logits/real": -2.682790517807007,
      "logps/generated": -129.32846069335938,
      "logps/real": -92.95919036865234,
      "loss": 0.2034,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.053739547729492,
      "rewards/margins": 15.391085624694824,
      "rewards/real": 10.337347030639648,
      "step": 1370
    },
    {
      "epoch": 0.8829174664107485,
      "grad_norm": 2.342066981273815,
      "learning_rate": 6.507823613086771e-08,
      "logits/generated": -2.734999179840088,
      "logits/real": -2.705516815185547,
      "logps/generated": -126.626708984375,
      "logps/real": -108.821044921875,
      "loss": 0.2083,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.249380111694336,
      "rewards/margins": 17.1163387298584,
      "rewards/real": 11.866960525512695,
      "step": 1380
    },
    {
      "epoch": 0.889315419065899,
      "grad_norm": 2.2461978230592665,
      "learning_rate": 6.152204836415363e-08,
      "logits/generated": -2.715066432952881,
      "logits/real": -2.795833110809326,
      "logps/generated": -136.6764373779297,
      "logps/real": -112.36259460449219,
      "loss": 0.219,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.541347026824951,
      "rewards/margins": 17.154422760009766,
      "rewards/real": 11.613073348999023,
      "step": 1390
    },
    {
      "epoch": 0.8957133717210493,
      "grad_norm": 1.9643094015838993,
      "learning_rate": 5.796586059743954e-08,
      "logits/generated": -2.7037057876586914,
      "logits/real": -2.892324447631836,
      "logps/generated": -144.58990478515625,
      "logps/real": -93.6194839477539,
      "loss": 0.1961,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.912665843963623,
      "rewards/margins": 16.394804000854492,
      "rewards/real": 10.482137680053711,
      "step": 1400
    },
    {
      "epoch": 0.8957133717210493,
      "eval_logits/generated": -2.731992244720459,
      "eval_logits/real": -2.7331748008728027,
      "eval_logps/generated": -128.72311401367188,
      "eval_logps/real": -111.41729736328125,
      "eval_loss": 0.22951382398605347,
      "eval_rewards/accuracies": 0.9921875,
      "eval_rewards/generated": -5.000611305236816,
      "eval_rewards/margins": 15.126992225646973,
      "eval_rewards/real": 10.12637996673584,
      "eval_runtime": 68.2753,
      "eval_samples_per_second": 7.323,
      "eval_steps_per_second": 0.234,
      "step": 1400
    },
    {
      "epoch": 0.9021113243761996,
      "grad_norm": 1.9508599183671798,
      "learning_rate": 5.4409672830725456e-08,
      "logits/generated": -2.771930694580078,
      "logits/real": -2.9047772884368896,
      "logps/generated": -157.674072265625,
      "logps/real": -95.12882232666016,
      "loss": 0.2091,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.542889595031738,
      "rewards/margins": 16.641141891479492,
      "rewards/real": 10.098252296447754,
      "step": 1410
    },
    {
      "epoch": 0.90850927703135,
      "grad_norm": 8.200787208344204,
      "learning_rate": 5.0853485064011376e-08,
      "logits/generated": -2.743286609649658,
      "logits/real": -2.764103412628174,
      "logps/generated": -124.41139221191406,
      "logps/real": -113.6160659790039,
      "loss": 0.2089,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.869805812835693,
      "rewards/margins": 16.73537826538086,
      "rewards/real": 11.865571975708008,
      "step": 1420
    },
    {
      "epoch": 0.9149072296865003,
      "grad_norm": 4.459379575662199,
      "learning_rate": 4.72972972972973e-08,
      "logits/generated": -2.746967077255249,
      "logits/real": -2.726330280303955,
      "logps/generated": -127.21881103515625,
      "logps/real": -105.5604248046875,
      "loss": 0.2142,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.854151248931885,
      "rewards/margins": 16.145641326904297,
      "rewards/real": 11.291491508483887,
      "step": 1430
    },
    {
      "epoch": 0.9213051823416507,
      "grad_norm": 2.74872128438021,
      "learning_rate": 4.374110953058322e-08,
      "logits/generated": -2.7054824829101562,
      "logits/real": -2.7788712978363037,
      "logps/generated": -141.21192932128906,
      "logps/real": -107.06086730957031,
      "loss": 0.2159,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.830003261566162,
      "rewards/margins": 17.112892150878906,
      "rewards/real": 11.282888412475586,
      "step": 1440
    },
    {
      "epoch": 0.927703134996801,
      "grad_norm": 2.4246159341403537,
      "learning_rate": 4.018492176386913e-08,
      "logits/generated": -2.6539907455444336,
      "logits/real": -2.773902177810669,
      "logps/generated": -130.4512939453125,
      "logps/real": -101.6545181274414,
      "loss": 0.2101,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.960906028747559,
      "rewards/margins": 15.910059928894043,
      "rewards/real": 10.949153900146484,
      "step": 1450
    },
    {
      "epoch": 0.9341010876519513,
      "grad_norm": 1.8673176828484437,
      "learning_rate": 3.6628733997155046e-08,
      "logits/generated": -2.7114219665527344,
      "logits/real": -2.8677616119384766,
      "logps/generated": -130.7792510986328,
      "logps/real": -88.69828796386719,
      "loss": 0.1885,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.1922407150268555,
      "rewards/margins": 14.514605522155762,
      "rewards/real": 9.32236385345459,
      "step": 1460
    },
    {
      "epoch": 0.9404990403071017,
      "grad_norm": 2.167805338937868,
      "learning_rate": 3.3072546230440967e-08,
      "logits/generated": -2.709365129470825,
      "logits/real": -2.7279558181762695,
      "logps/generated": -129.8989715576172,
      "logps/real": -96.73854064941406,
      "loss": 0.1966,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.115813255310059,
      "rewards/margins": 15.31916332244873,
      "rewards/real": 10.203348159790039,
      "step": 1470
    },
    {
      "epoch": 0.946896992962252,
      "grad_norm": 2.343424293796162,
      "learning_rate": 2.9516358463726884e-08,
      "logits/generated": -2.6958348751068115,
      "logits/real": -2.7627506256103516,
      "logps/generated": -132.286376953125,
      "logps/real": -85.90814971923828,
      "loss": 0.2066,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.348180294036865,
      "rewards/margins": 14.917572021484375,
      "rewards/real": 9.569391250610352,
      "step": 1480
    },
    {
      "epoch": 0.9532949456174025,
      "grad_norm": 2.443526667812465,
      "learning_rate": 2.59601706970128e-08,
      "logits/generated": -2.731065273284912,
      "logits/real": -2.722964286804199,
      "logps/generated": -144.53329467773438,
      "logps/real": -111.11639404296875,
      "loss": 0.2125,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.713242530822754,
      "rewards/margins": 17.423105239868164,
      "rewards/real": 11.70986270904541,
      "step": 1490
    },
    {
      "epoch": 0.9596928982725528,
      "grad_norm": 2.5802622540203726,
      "learning_rate": 2.240398293029872e-08,
      "logits/generated": -2.7666499614715576,
      "logits/real": -2.7990810871124268,
      "logps/generated": -151.97225952148438,
      "logps/real": -102.18070983886719,
      "loss": 0.2062,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.25309419631958,
      "rewards/margins": 17.280641555786133,
      "rewards/real": 11.027546882629395,
      "step": 1500
    },
    {
      "epoch": 0.9660908509277031,
      "grad_norm": 4.900620511000788,
      "learning_rate": 1.8847795163584636e-08,
      "logits/generated": -2.7653520107269287,
      "logits/real": -2.737842082977295,
      "logps/generated": -132.54611206054688,
      "logps/real": -105.3329086303711,
      "loss": 0.2044,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.18030309677124,
      "rewards/margins": 16.7701358795166,
      "rewards/real": 11.58983325958252,
      "step": 1510
    },
    {
      "epoch": 0.9724888035828535,
      "grad_norm": 2.2165972838393935,
      "learning_rate": 1.5291607396870554e-08,
      "logits/generated": -2.709165334701538,
      "logits/real": -2.7248358726501465,
      "logps/generated": -131.63613891601562,
      "logps/real": -111.87278747558594,
      "loss": 0.2149,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.262539863586426,
      "rewards/margins": 17.233844757080078,
      "rewards/real": 11.971303939819336,
      "step": 1520
    },
    {
      "epoch": 0.9788867562380038,
      "grad_norm": 4.970086199198337,
      "learning_rate": 1.1735419630156473e-08,
      "logits/generated": -2.7099928855895996,
      "logits/real": -2.7440731525421143,
      "logps/generated": -127.85652160644531,
      "logps/real": -98.89935302734375,
      "loss": 0.1979,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.844382286071777,
      "rewards/margins": 15.250204086303711,
      "rewards/real": 10.4058198928833,
      "step": 1530
    },
    {
      "epoch": 0.9852847088931542,
      "grad_norm": 2.031754041409757,
      "learning_rate": 8.179231863442388e-09,
      "logits/generated": -2.6747050285339355,
      "logits/real": -2.6660659313201904,
      "logps/generated": -139.36756896972656,
      "logps/real": -103.53269958496094,
      "loss": 0.2106,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.578815460205078,
      "rewards/margins": 16.89048194885254,
      "rewards/real": 11.311666488647461,
      "step": 1540
    },
    {
      "epoch": 0.9916826615483045,
      "grad_norm": 2.2877158724102564,
      "learning_rate": 4.623044096728307e-09,
      "logits/generated": -2.660313606262207,
      "logits/real": -2.855464220046997,
      "logps/generated": -145.4078826904297,
      "logps/real": -106.03314208984375,
      "loss": 0.2083,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.917939186096191,
      "rewards/margins": 17.319114685058594,
      "rewards/real": 11.40117359161377,
      "step": 1550
    },
    {
      "epoch": 0.9980806142034548,
      "grad_norm": 2.427910151920255,
      "learning_rate": 1.0668563300142248e-09,
      "logits/generated": -2.697697401046753,
      "logits/real": -2.7648138999938965,
      "logps/generated": -132.71572875976562,
      "logps/real": -100.33382415771484,
      "loss": 0.2061,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.841822624206543,
      "rewards/margins": 15.7789306640625,
      "rewards/real": 10.937108039855957,
      "step": 1560
    },
    {
      "epoch": 1.0,
      "step": 1563,
      "total_flos": 0.0,
      "train_loss": 0.2369839982275618,
      "train_runtime": 14627.5004,
      "train_samples_per_second": 3.418,
      "train_steps_per_second": 0.107
    }
  ],
  "logging_steps": 10,
  "max_steps": 1563,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "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
}