{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 50,
  "global_step": 375,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.3157894736842104e-08,
      "logits/generated": -2.9681174755096436,
      "logits/real": -2.6839590072631836,
      "logps/generated": -114.33634185791016,
      "logps/real": -86.82533264160156,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/generated": 0.0,
      "rewards/margins": 0.0,
      "rewards/real": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.3157894736842104e-07,
      "logits/generated": -3.07245135307312,
      "logits/real": -2.7284328937530518,
      "logps/generated": -137.1116943359375,
      "logps/real": -131.3474884033203,
      "loss": 0.6592,
      "rewards/accuracies": 0.8333333134651184,
      "rewards/generated": -0.030464060604572296,
      "rewards/margins": 0.07382892817258835,
      "rewards/real": 0.04336486756801605,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.631578947368421e-07,
      "logits/generated": -2.764019012451172,
      "logits/real": -2.412205219268799,
      "logps/generated": -272.6034240722656,
      "logps/real": -128.99154663085938,
      "loss": 0.3405,
      "rewards/accuracies": 1.0,
      "rewards/generated": -1.3848954439163208,
      "rewards/margins": 1.49906325340271,
      "rewards/real": 0.1141679435968399,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.9473684210526315e-07,
      "logits/generated": -2.10380482673645,
      "logits/real": -1.7264606952667236,
      "logps/generated": -1245.616943359375,
      "logps/real": -437.305908203125,
      "loss": 0.0957,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/generated": -11.244729995727539,
      "rewards/margins": 8.229094505310059,
      "rewards/real": -3.0156354904174805,
      "step": 30
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.970326409495548e-07,
      "logits/generated": -1.417828917503357,
      "logits/real": -1.4325724840164185,
      "logps/generated": -1411.8931884765625,
      "logps/real": -254.04269409179688,
      "loss": 0.0944,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/generated": -12.817289352416992,
      "rewards/margins": 11.621607780456543,
      "rewards/real": -1.1956826448440552,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.821958456973294e-07,
      "logits/generated": 2.231259822845459,
      "logits/real": 0.8822120428085327,
      "logps/generated": -3168.895751953125,
      "logps/real": -463.5333557128906,
      "loss": 0.0717,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/generated": -30.413869857788086,
      "rewards/margins": 27.164615631103516,
      "rewards/real": -3.2492573261260986,
      "step": 50
    },
    {
      "epoch": 0.13,
      "eval_logits/generated": 3.4830620288848877,
      "eval_logits/real": 1.9303022623062134,
      "eval_logps/generated": -3859.57080078125,
      "eval_logps/real": -449.6531677246094,
      "eval_loss": 0.049043506383895874,
      "eval_rewards/accuracies": 0.9906914830207825,
      "eval_rewards/generated": -37.34311294555664,
      "eval_rewards/margins": 34.217308044433594,
      "eval_rewards/real": -3.125804901123047,
      "eval_runtime": 215.9205,
      "eval_samples_per_second": 13.894,
      "eval_steps_per_second": 0.435,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.673590504451038e-07,
      "logits/generated": 1.6689307689666748,
      "logits/real": 0.8687565922737122,
      "logps/generated": -3099.0478515625,
      "logps/real": -587.6290283203125,
      "loss": 0.0921,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/generated": -29.83612060546875,
      "rewards/margins": 25.382980346679688,
      "rewards/real": -4.453139305114746,
      "step": 60
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.5252225519287835e-07,
      "logits/generated": -0.5513351559638977,
      "logits/real": -0.6214720010757446,
      "logps/generated": -2553.958740234375,
      "logps/real": -673.9515991210938,
      "loss": 0.081,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -24.296937942504883,
      "rewards/margins": 19.031129837036133,
      "rewards/real": -5.265807628631592,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.376854599406528e-07,
      "logits/generated": -1.0486136674880981,
      "logits/real": -1.0468746423721313,
      "logps/generated": -3602.500732421875,
      "logps/real": -834.3878784179688,
      "loss": 0.0477,
      "rewards/accuracies": 1.0,
      "rewards/generated": -34.76307678222656,
      "rewards/margins": 27.884380340576172,
      "rewards/real": -6.8786940574646,
      "step": 80
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.228486646884273e-07,
      "logits/generated": -0.8733774423599243,
      "logits/real": -0.8482550382614136,
      "logps/generated": -3528.38330078125,
      "logps/real": -824.9230346679688,
      "loss": 0.0551,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/generated": -34.10634994506836,
      "rewards/margins": 27.168109893798828,
      "rewards/real": -6.938241004943848,
      "step": 90
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.0801186943620176e-07,
      "logits/generated": -0.21600039303302765,
      "logits/real": -0.1288779377937317,
      "logps/generated": -3730.033203125,
      "logps/real": -664.6776733398438,
      "loss": 0.0323,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/generated": -36.04426574707031,
      "rewards/margins": 30.667240142822266,
      "rewards/real": -5.377036094665527,
      "step": 100
    },
    {
      "epoch": 0.27,
      "eval_logits/generated": -0.015528609976172447,
      "eval_logits/real": -0.16258487105369568,
      "eval_logps/generated": -4009.05517578125,
      "eval_logps/real": -536.6591796875,
      "eval_loss": 0.03004666045308113,
      "eval_rewards/accuracies": 0.9973404407501221,
      "eval_rewards/generated": -38.83795928955078,
      "eval_rewards/margins": 34.84209442138672,
      "eval_rewards/real": -3.995864152908325,
      "eval_runtime": 213.8643,
      "eval_samples_per_second": 14.028,
      "eval_steps_per_second": 0.44,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 3.931750741839762e-07,
      "logits/generated": 0.13914000988006592,
      "logits/real": -0.10056102275848389,
      "logps/generated": -3989.37890625,
      "logps/real": -550.2565307617188,
      "loss": 0.0283,
      "rewards/accuracies": 1.0,
      "rewards/generated": -38.63182830810547,
      "rewards/margins": 34.458984375,
      "rewards/real": -4.172848224639893,
      "step": 110
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.7833827893175073e-07,
      "logits/generated": 0.34012800455093384,
      "logits/real": -0.4745260775089264,
      "logps/generated": -4985.052734375,
      "logps/real": -664.0899047851562,
      "loss": 0.0182,
      "rewards/accuracies": 1.0,
      "rewards/generated": -48.502471923828125,
      "rewards/margins": 43.33242416381836,
      "rewards/real": -5.170046806335449,
      "step": 120
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.635014836795252e-07,
      "logits/generated": 0.1598968654870987,
      "logits/real": 0.478419691324234,
      "logps/generated": -4668.8681640625,
      "logps/real": -993.2604370117188,
      "loss": 0.0371,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -45.391544342041016,
      "rewards/margins": 36.9045524597168,
      "rewards/real": -8.48698616027832,
      "step": 130
    },
    {
      "epoch": 0.37,
      "learning_rate": 3.486646884272997e-07,
      "logits/generated": 0.4115122854709625,
      "logits/real": 0.367294043302536,
      "logps/generated": -4831.67919921875,
      "logps/real": -1011.8640747070312,
      "loss": 0.0204,
      "rewards/accuracies": 1.0,
      "rewards/generated": -46.988433837890625,
      "rewards/margins": 38.47046661376953,
      "rewards/real": -8.517963409423828,
      "step": 140
    },
    {
      "epoch": 0.4,
      "learning_rate": 3.3382789317507414e-07,
      "logits/generated": 0.7945320010185242,
      "logits/real": 0.49422749876976013,
      "logps/generated": -4842.3310546875,
      "logps/real": -903.1175537109375,
      "loss": 0.026,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/generated": -47.182762145996094,
      "rewards/margins": 39.56975555419922,
      "rewards/real": -7.613003730773926,
      "step": 150
    },
    {
      "epoch": 0.4,
      "eval_logits/generated": 1.0207222700119019,
      "eval_logits/real": 1.0071420669555664,
      "eval_logps/generated": -5130.18798828125,
      "eval_logps/real": -958.144287109375,
      "eval_loss": 0.01575618050992489,
      "eval_rewards/accuracies": 0.9946808218955994,
      "eval_rewards/generated": -50.04927444458008,
      "eval_rewards/margins": 41.83856201171875,
      "eval_rewards/real": -8.210715293884277,
      "eval_runtime": 213.3884,
      "eval_samples_per_second": 14.059,
      "eval_steps_per_second": 0.441,
      "step": 150
    },
    {
      "epoch": 0.43,
      "learning_rate": 3.189910979228487e-07,
      "logits/generated": 0.9495289921760559,
      "logits/real": 0.9829273223876953,
      "logps/generated": -4704.9228515625,
      "logps/real": -946.2503051757812,
      "loss": 0.0139,
      "rewards/accuracies": 1.0,
      "rewards/generated": -45.76738739013672,
      "rewards/margins": 37.696834564208984,
      "rewards/real": -8.070551872253418,
      "step": 160
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.0415430267062316e-07,
      "logits/generated": 1.2480947971343994,
      "logits/real": 1.0694479942321777,
      "logps/generated": -4743.8857421875,
      "logps/real": -894.25341796875,
      "loss": 0.0234,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -46.242591857910156,
      "rewards/margins": 38.66545104980469,
      "rewards/real": -7.577139854431152,
      "step": 170
    },
    {
      "epoch": 0.48,
      "learning_rate": 2.893175074183976e-07,
      "logits/generated": 1.5751049518585205,
      "logits/real": 0.7654831409454346,
      "logps/generated": -4563.38671875,
      "logps/real": -743.2201538085938,
      "loss": 0.0227,
      "rewards/accuracies": 1.0,
      "rewards/generated": -44.37525177001953,
      "rewards/margins": 38.28398895263672,
      "rewards/real": -6.0912652015686035,
      "step": 180
    },
    {
      "epoch": 0.51,
      "learning_rate": 2.744807121661721e-07,
      "logits/generated": 1.8286101818084717,
      "logits/real": 1.0053759813308716,
      "logps/generated": -5664.7138671875,
      "logps/real": -959.2091064453125,
      "loss": 0.0254,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -55.386558532714844,
      "rewards/margins": 47.158206939697266,
      "rewards/real": -8.228350639343262,
      "step": 190
    },
    {
      "epoch": 0.53,
      "learning_rate": 2.596439169139466e-07,
      "logits/generated": 2.440897226333618,
      "logits/real": 1.7533397674560547,
      "logps/generated": -6381.81005859375,
      "logps/real": -920.97509765625,
      "loss": 0.0106,
      "rewards/accuracies": 1.0,
      "rewards/generated": -62.468101501464844,
      "rewards/margins": 54.67470169067383,
      "rewards/real": -7.793404579162598,
      "step": 200
    },
    {
      "epoch": 0.53,
      "eval_logits/generated": 2.2348735332489014,
      "eval_logits/real": 1.2992360591888428,
      "eval_logps/generated": -6298.50927734375,
      "eval_logps/real": -1062.12646484375,
      "eval_loss": 0.008726131170988083,
      "eval_rewards/accuracies": 0.9960106611251831,
      "eval_rewards/generated": -61.73249435424805,
      "eval_rewards/margins": 52.48196029663086,
      "eval_rewards/real": -9.250536918640137,
      "eval_runtime": 212.9976,
      "eval_samples_per_second": 14.085,
      "eval_steps_per_second": 0.441,
      "step": 200
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.4480712166172106e-07,
      "logits/generated": 2.323308229446411,
      "logits/real": 1.4621614217758179,
      "logps/generated": -6335.68310546875,
      "logps/real": -1120.7493896484375,
      "loss": 0.0035,
      "rewards/accuracies": 1.0,
      "rewards/generated": -62.093727111816406,
      "rewards/margins": 52.304664611816406,
      "rewards/real": -9.789057731628418,
      "step": 210
    },
    {
      "epoch": 0.59,
      "learning_rate": 2.2997032640949554e-07,
      "logits/generated": 2.158539295196533,
      "logits/real": 1.8379024267196655,
      "logps/generated": -4581.6611328125,
      "logps/real": -948.8102416992188,
      "loss": 0.0202,
      "rewards/accuracies": 1.0,
      "rewards/generated": -44.635250091552734,
      "rewards/margins": 36.491943359375,
      "rewards/real": -8.143308639526367,
      "step": 220
    },
    {
      "epoch": 0.61,
      "learning_rate": 2.1513353115727e-07,
      "logits/generated": 2.464277982711792,
      "logits/real": 2.7233450412750244,
      "logps/generated": -4033.82421875,
      "logps/real": -1097.900390625,
      "loss": 0.0053,
      "rewards/accuracies": 1.0,
      "rewards/generated": -39.125518798828125,
      "rewards/margins": 29.4420166015625,
      "rewards/real": -9.683503150939941,
      "step": 230
    },
    {
      "epoch": 0.64,
      "learning_rate": 2.0029673590504451e-07,
      "logits/generated": 2.664623260498047,
      "logits/real": 2.550762176513672,
      "logps/generated": -4957.3818359375,
      "logps/real": -1175.33740234375,
      "loss": 0.0067,
      "rewards/accuracies": 1.0,
      "rewards/generated": -48.34945297241211,
      "rewards/margins": 38.046714782714844,
      "rewards/real": -10.302745819091797,
      "step": 240
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.8545994065281897e-07,
      "logits/generated": 2.886002779006958,
      "logits/real": 2.470453977584839,
      "logps/generated": -5135.99755859375,
      "logps/real": -1098.706787109375,
      "loss": 0.0071,
      "rewards/accuracies": 1.0,
      "rewards/generated": -50.12360382080078,
      "rewards/margins": 40.536460876464844,
      "rewards/real": -9.587137222290039,
      "step": 250
    },
    {
      "epoch": 0.67,
      "eval_logits/generated": 2.8798305988311768,
      "eval_logits/real": 3.2924983501434326,
      "eval_logps/generated": -5056.44091796875,
      "eval_logps/real": -1277.58740234375,
      "eval_loss": 0.010579616762697697,
      "eval_rewards/accuracies": 0.998670220375061,
      "eval_rewards/generated": -49.31182098388672,
      "eval_rewards/margins": 37.90666961669922,
      "eval_rewards/real": -11.405147552490234,
      "eval_runtime": 213.7924,
      "eval_samples_per_second": 14.032,
      "eval_steps_per_second": 0.44,
      "step": 250
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.7062314540059346e-07,
      "logits/generated": 3.001654863357544,
      "logits/real": 3.0144619941711426,
      "logps/generated": -5104.75,
      "logps/real": -1167.21826171875,
      "loss": 0.0035,
      "rewards/accuracies": 1.0,
      "rewards/generated": -49.85370635986328,
      "rewards/margins": 39.53137969970703,
      "rewards/real": -10.322334289550781,
      "step": 260
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.5578635014836795e-07,
      "logits/generated": 2.4701550006866455,
      "logits/real": 2.8096556663513184,
      "logps/generated": -4720.4482421875,
      "logps/real": -1196.321044921875,
      "loss": 0.0055,
      "rewards/accuracies": 1.0,
      "rewards/generated": -45.905601501464844,
      "rewards/margins": 35.247032165527344,
      "rewards/real": -10.6585693359375,
      "step": 270
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.4094955489614243e-07,
      "logits/generated": 2.977241277694702,
      "logits/real": 2.97806715965271,
      "logps/generated": -5751.66748046875,
      "logps/real": -1152.299072265625,
      "loss": 0.0151,
      "rewards/accuracies": 1.0,
      "rewards/generated": -56.2369384765625,
      "rewards/margins": 46.072731018066406,
      "rewards/real": -10.164213180541992,
      "step": 280
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.261127596439169e-07,
      "logits/generated": 2.786933183670044,
      "logits/real": 2.2626030445098877,
      "logps/generated": -5328.9072265625,
      "logps/real": -1131.6319580078125,
      "loss": 0.0074,
      "rewards/accuracies": 1.0,
      "rewards/generated": -51.968414306640625,
      "rewards/margins": 41.99260711669922,
      "rewards/real": -9.975804328918457,
      "step": 290
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.1127596439169139e-07,
      "logits/generated": 2.6825292110443115,
      "logits/real": 2.87423038482666,
      "logps/generated": -4691.046875,
      "logps/real": -1226.396728515625,
      "loss": 0.0121,
      "rewards/accuracies": 1.0,
      "rewards/generated": -45.6788330078125,
      "rewards/margins": 34.744598388671875,
      "rewards/real": -10.934233665466309,
      "step": 300
    },
    {
      "epoch": 0.8,
      "eval_logits/generated": 2.8713479042053223,
      "eval_logits/real": 2.6792426109313965,
      "eval_logps/generated": -5036.78271484375,
      "eval_logps/real": -1039.31103515625,
      "eval_loss": 0.007437660358846188,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -49.115230560302734,
      "eval_rewards/margins": 40.09284591674805,
      "eval_rewards/real": -9.022383689880371,
      "eval_runtime": 213.3741,
      "eval_samples_per_second": 14.06,
      "eval_steps_per_second": 0.441,
      "step": 300
    },
    {
      "epoch": 0.83,
      "learning_rate": 9.643916913946587e-08,
      "logits/generated": 2.8469080924987793,
      "logits/real": 2.4686405658721924,
      "logps/generated": -5047.57080078125,
      "logps/real": -1106.5921630859375,
      "loss": 0.0032,
      "rewards/accuracies": 1.0,
      "rewards/generated": -49.19374465942383,
      "rewards/margins": 39.55516815185547,
      "rewards/real": -9.638582229614258,
      "step": 310
    },
    {
      "epoch": 0.85,
      "learning_rate": 8.160237388724035e-08,
      "logits/generated": 3.1090855598449707,
      "logits/real": 2.914607048034668,
      "logps/generated": -6005.77587890625,
      "logps/real": -1140.301025390625,
      "loss": 0.0123,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -58.757904052734375,
      "rewards/margins": 48.729026794433594,
      "rewards/real": -10.028875350952148,
      "step": 320
    },
    {
      "epoch": 0.88,
      "learning_rate": 6.676557863501484e-08,
      "logits/generated": 2.659290313720703,
      "logits/real": 2.3654632568359375,
      "logps/generated": -5298.484375,
      "logps/real": -1117.4388427734375,
      "loss": 0.0108,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -51.67462158203125,
      "rewards/margins": 41.78734588623047,
      "rewards/real": -9.887277603149414,
      "step": 330
    },
    {
      "epoch": 0.91,
      "learning_rate": 5.192878338278932e-08,
      "logits/generated": 2.7824113368988037,
      "logits/real": 1.4837300777435303,
      "logps/generated": -5636.43017578125,
      "logps/real": -1132.7353515625,
      "loss": 0.0045,
      "rewards/accuracies": 1.0,
      "rewards/generated": -55.110389709472656,
      "rewards/margins": 45.19716262817383,
      "rewards/real": -9.913228034973145,
      "step": 340
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.709198813056379e-08,
      "logits/generated": 2.721921443939209,
      "logits/real": 1.7522449493408203,
      "logps/generated": -5212.6552734375,
      "logps/real": -1117.9945068359375,
      "loss": 0.0013,
      "rewards/accuracies": 1.0,
      "rewards/generated": -50.842464447021484,
      "rewards/margins": 41.09899139404297,
      "rewards/real": -9.743471145629883,
      "step": 350
    },
    {
      "epoch": 0.93,
      "eval_logits/generated": 3.074373722076416,
      "eval_logits/real": 1.9176675081253052,
      "eval_logps/generated": -5640.6728515625,
      "eval_logps/real": -1151.2216796875,
      "eval_loss": 0.006871914956718683,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -55.15412521362305,
      "eval_rewards/margins": 45.012638092041016,
      "eval_rewards/real": -10.141489028930664,
      "eval_runtime": 213.3069,
      "eval_samples_per_second": 14.064,
      "eval_steps_per_second": 0.441,
      "step": 350
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.225519287833828e-08,
      "logits/generated": 3.0481696128845215,
      "logits/real": 1.9120832681655884,
      "logps/generated": -5828.06787109375,
      "logps/real": -1140.780517578125,
      "loss": 0.0043,
      "rewards/accuracies": 1.0,
      "rewards/generated": -56.97248077392578,
      "rewards/margins": 47.068485260009766,
      "rewards/real": -9.903992652893066,
      "step": 360
    },
    {
      "epoch": 0.99,
      "learning_rate": 7.418397626112759e-09,
      "logits/generated": 3.2567169666290283,
      "logits/real": 2.3705618381500244,
      "logps/generated": -6292.123046875,
      "logps/real": -1168.8006591796875,
      "loss": 0.0035,
      "rewards/accuracies": 1.0,
      "rewards/generated": -61.6358642578125,
      "rewards/margins": 51.312286376953125,
      "rewards/real": -10.323575973510742,
      "step": 370
    },
    {
      "epoch": 1.0,
      "step": 375,
      "total_flos": 0.0,
      "train_loss": 0.051339938134110225,
      "train_runtime": 3437.1804,
      "train_samples_per_second": 3.491,
      "train_steps_per_second": 0.109
    }
  ],
  "logging_steps": 10,
  "max_steps": 375,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}