{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9915492957746479,
  "eval_steps": 30,
  "global_step": 88,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "logits/chosen": -3.188204765319824,
      "logits/rejected": -2.849832534790039,
      "logps/chosen": -220.16908264160156,
      "logps/rejected": -186.17868041992188,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.02,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "logits/chosen": -3.417233943939209,
      "logits/rejected": -3.385444164276123,
      "logps/chosen": -511.65631103515625,
      "logps/rejected": -303.28558349609375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 2
    },
    {
      "epoch": 0.03,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "logits/chosen": -2.9685182571411133,
      "logits/rejected": -2.9099667072296143,
      "logps/chosen": -131.90631103515625,
      "logps/rejected": -127.09970092773438,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 3
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.127852439880371,
      "learning_rate": 9.999645980833454e-05,
      "logits/chosen": -3.3438305854797363,
      "logits/rejected": -2.7247352600097656,
      "logps/chosen": -212.61917114257812,
      "logps/rejected": -160.17083740234375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 4
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.166139602661133,
      "learning_rate": 9.998583973465646e-05,
      "logits/chosen": -3.382075786590576,
      "logits/rejected": -2.574824094772339,
      "logps/chosen": -284.8162841796875,
      "logps/rejected": -199.35745239257812,
      "loss": 0.5279,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.21959751844406128,
      "rewards/margins": 0.6385258436203003,
      "rewards/rejected": -0.8581234216690063,
      "step": 5
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.322300910949707,
      "learning_rate": 9.99681412828496e-05,
      "logits/chosen": -3.271949291229248,
      "logits/rejected": -2.513960838317871,
      "logps/chosen": -372.6782531738281,
      "logps/rejected": -174.07479858398438,
      "loss": 0.2913,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.6500003933906555,
      "rewards/margins": 2.8001651763916016,
      "rewards/rejected": -2.1501646041870117,
      "step": 6
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.41703987121582,
      "learning_rate": 9.99433669591504e-05,
      "logits/chosen": -3.0961129665374756,
      "logits/rejected": -2.8720593452453613,
      "logps/chosen": -316.4073486328125,
      "logps/rejected": -167.79397583007812,
      "loss": 0.4464,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.1310131549835205,
      "rewards/margins": 1.2248451709747314,
      "rewards/rejected": -2.355858087539673,
      "step": 7
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.9780449867248535,
      "learning_rate": 9.991152027179307e-05,
      "logits/chosen": -2.9302220344543457,
      "logits/rejected": -2.5785322189331055,
      "logps/chosen": -165.38961791992188,
      "logps/rejected": -118.81493377685547,
      "loss": 0.1908,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.9433582425117493,
      "rewards/margins": 3.344132661819458,
      "rewards/rejected": -2.4007744789123535,
      "step": 8
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.300680637359619,
      "learning_rate": 9.987260573051269e-05,
      "logits/chosen": -3.1814284324645996,
      "logits/rejected": -3.0421230792999268,
      "logps/chosen": -258.6530456542969,
      "logps/rejected": -209.901611328125,
      "loss": 0.3466,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.742733895778656,
      "rewards/margins": 1.9031461477279663,
      "rewards/rejected": -2.6458799839019775,
      "step": 9
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.8778839111328125,
      "learning_rate": 9.982662884590662e-05,
      "logits/chosen": -2.9834322929382324,
      "logits/rejected": -2.546109199523926,
      "logps/chosen": -357.8366394042969,
      "logps/rejected": -137.680908203125,
      "loss": 0.4102,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.5062696933746338,
      "rewards/margins": 4.4324493408203125,
      "rewards/rejected": -4.938718795776367,
      "step": 10
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.32342529296875,
      "learning_rate": 9.977359612865423e-05,
      "logits/chosen": -2.992344379425049,
      "logits/rejected": -2.4726979732513428,
      "logps/chosen": -248.93341064453125,
      "logps/rejected": -167.0366668701172,
      "loss": 0.5224,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.211262047290802,
      "rewards/margins": 4.505336284637451,
      "rewards/rejected": -4.294074058532715,
      "step": 11
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.8779878616333,
      "learning_rate": 9.971351508859488e-05,
      "logits/chosen": -3.1347732543945312,
      "logits/rejected": -2.930368423461914,
      "logps/chosen": -286.450439453125,
      "logps/rejected": -202.72720336914062,
      "loss": 0.6846,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.4156379699707031,
      "rewards/margins": 3.582124710083008,
      "rewards/rejected": -4.997762680053711,
      "step": 12
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.25694465637207,
      "learning_rate": 9.964639423366442e-05,
      "logits/chosen": -3.121607780456543,
      "logits/rejected": -2.9769201278686523,
      "logps/chosen": -291.4040832519531,
      "logps/rejected": -219.06231689453125,
      "loss": 0.27,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.20223967730998993,
      "rewards/margins": 4.1018805503845215,
      "rewards/rejected": -3.8996407985687256,
      "step": 13
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.865050792694092,
      "learning_rate": 9.957224306869053e-05,
      "logits/chosen": -3.265730381011963,
      "logits/rejected": -2.9264979362487793,
      "logps/chosen": -269.4535827636719,
      "logps/rejected": -205.49166870117188,
      "loss": 0.1639,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.9554134607315063,
      "rewards/margins": 5.716032028198242,
      "rewards/rejected": -4.760618686676025,
      "step": 14
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.5297553539276123,
      "learning_rate": 9.949107209404665e-05,
      "logits/chosen": -2.461251735687256,
      "logits/rejected": -2.371189832687378,
      "logps/chosen": -259.4444885253906,
      "logps/rejected": -134.9102783203125,
      "loss": 0.2108,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.03092677891254425,
      "rewards/margins": 3.4555163383483887,
      "rewards/rejected": -3.4245896339416504,
      "step": 15
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.433193206787109,
      "learning_rate": 9.940289280416508e-05,
      "logits/chosen": -3.2052440643310547,
      "logits/rejected": -2.8054556846618652,
      "logps/chosen": -265.2724304199219,
      "logps/rejected": -225.47068786621094,
      "loss": 0.1771,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.08068099617958069,
      "rewards/margins": 4.355477333068848,
      "rewards/rejected": -4.274796485900879,
      "step": 16
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.783153533935547,
      "learning_rate": 9.930771768590933e-05,
      "logits/chosen": -3.0025675296783447,
      "logits/rejected": -3.019012451171875,
      "logps/chosen": -195.4888916015625,
      "logps/rejected": -198.30792236328125,
      "loss": 0.6906,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -2.109578847885132,
      "rewards/margins": 0.9626373052597046,
      "rewards/rejected": -3.072216033935547,
      "step": 17
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.201537132263184,
      "learning_rate": 9.92055602168058e-05,
      "logits/chosen": -3.2320845127105713,
      "logits/rejected": -2.9034526348114014,
      "logps/chosen": -226.6787109375,
      "logps/rejected": -132.74095153808594,
      "loss": 0.2037,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.05915778875350952,
      "rewards/margins": 3.9392547607421875,
      "rewards/rejected": -3.8800971508026123,
      "step": 18
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.768068313598633,
      "learning_rate": 9.909643486313533e-05,
      "logits/chosen": -3.178317070007324,
      "logits/rejected": -3.297847270965576,
      "logps/chosen": -412.5939636230469,
      "logps/rejected": -241.15750122070312,
      "loss": 0.5854,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.0589046478271484,
      "rewards/margins": 5.891679286956787,
      "rewards/rejected": -6.9505839347839355,
      "step": 19
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.77907943725586,
      "learning_rate": 9.898035707788463e-05,
      "logits/chosen": -3.237691640853882,
      "logits/rejected": -3.086668014526367,
      "logps/chosen": -311.6419372558594,
      "logps/rejected": -238.47996520996094,
      "loss": 0.4823,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.5970950126647949,
      "rewards/margins": 4.7628984451293945,
      "rewards/rejected": -5.3599934577941895,
      "step": 20
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.0071136951446533,
      "learning_rate": 9.885734329855798e-05,
      "logits/chosen": -3.2764744758605957,
      "logits/rejected": -2.930645704269409,
      "logps/chosen": -345.3895568847656,
      "logps/rejected": -216.34646606445312,
      "loss": 0.0629,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1355853080749512,
      "rewards/margins": 6.265766620635986,
      "rewards/rejected": -5.130181312561035,
      "step": 21
    },
    {
      "epoch": 0.25,
      "grad_norm": 11.331860542297363,
      "learning_rate": 9.872741094484965e-05,
      "logits/chosen": -3.3079967498779297,
      "logits/rejected": -2.7648227214813232,
      "logps/chosen": -351.6337585449219,
      "logps/rejected": -270.9600830078125,
      "loss": 0.369,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.9473640322685242,
      "rewards/margins": 2.920064687728882,
      "rewards/rejected": -3.8674285411834717,
      "step": 22
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.934148073196411,
      "learning_rate": 9.859057841617709e-05,
      "logits/chosen": -2.9965269565582275,
      "logits/rejected": -2.872361183166504,
      "logps/chosen": -424.9866943359375,
      "logps/rejected": -250.78515625,
      "loss": 0.0707,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7661307454109192,
      "rewards/margins": 8.092283248901367,
      "rewards/rejected": -7.326152801513672,
      "step": 23
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.69138765335083,
      "learning_rate": 9.844686508907537e-05,
      "logits/chosen": -3.3082752227783203,
      "logits/rejected": -3.2877731323242188,
      "logps/chosen": -293.59478759765625,
      "logps/rejected": -179.19845581054688,
      "loss": 0.275,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -1.0576655864715576,
      "rewards/margins": 2.9845128059387207,
      "rewards/rejected": -4.042178153991699,
      "step": 24
    },
    {
      "epoch": 0.28,
      "grad_norm": 14.109583854675293,
      "learning_rate": 9.829629131445342e-05,
      "logits/chosen": -2.856363534927368,
      "logits/rejected": -2.133051633834839,
      "logps/chosen": -203.2696533203125,
      "logps/rejected": -100.72183990478516,
      "loss": 0.8885,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.2501509189605713,
      "rewards/margins": 1.4751089811325073,
      "rewards/rejected": -2.725259780883789,
      "step": 25
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.02879810333252,
      "learning_rate": 9.81388784147121e-05,
      "logits/chosen": -3.1515085697174072,
      "logits/rejected": -2.761265754699707,
      "logps/chosen": -230.35438537597656,
      "logps/rejected": -280.5235290527344,
      "loss": 0.3744,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.2650168836116791,
      "rewards/margins": 4.330430507659912,
      "rewards/rejected": -4.065413475036621,
      "step": 26
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.773349761962891,
      "learning_rate": 9.797464868072488e-05,
      "logits/chosen": -3.1555724143981934,
      "logits/rejected": -2.7849271297454834,
      "logps/chosen": -204.43417358398438,
      "logps/rejected": -110.06712341308594,
      "loss": 0.2692,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.146636962890625,
      "rewards/margins": 2.842060089111328,
      "rewards/rejected": -3.988697052001953,
      "step": 27
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.401435375213623,
      "learning_rate": 9.780362536868113e-05,
      "logits/chosen": -2.565051317214966,
      "logits/rejected": -2.8752522468566895,
      "logps/chosen": -158.33001708984375,
      "logps/rejected": -139.02786254882812,
      "loss": 0.3653,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.32712388038635254,
      "rewards/margins": 2.7614259719848633,
      "rewards/rejected": -3.088550090789795,
      "step": 28
    },
    {
      "epoch": 0.33,
      "grad_norm": 14.269856452941895,
      "learning_rate": 9.762583269679303e-05,
      "logits/chosen": -2.4300293922424316,
      "logits/rejected": -2.7993407249450684,
      "logps/chosen": -233.670166015625,
      "logps/rejected": -229.98846435546875,
      "loss": 0.82,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.4308565855026245,
      "rewards/margins": 1.704122543334961,
      "rewards/rejected": -3.134979248046875,
      "step": 29
    },
    {
      "epoch": 0.34,
      "grad_norm": 11.34128189086914,
      "learning_rate": 9.744129584186598e-05,
      "logits/chosen": -2.8739359378814697,
      "logits/rejected": -3.0709590911865234,
      "logps/chosen": -187.5973663330078,
      "logps/rejected": -171.49639892578125,
      "loss": 0.8236,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.5466258525848389,
      "rewards/margins": 1.1274800300598145,
      "rewards/rejected": -2.674105644226074,
      "step": 30
    },
    {
      "epoch": 0.34,
      "eval_logits/chosen": -3.2232699394226074,
      "eval_logits/rejected": -2.919212818145752,
      "eval_logps/chosen": -186.79000854492188,
      "eval_logps/rejected": -195.58921813964844,
      "eval_loss": 0.004889195319265127,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 1.5010696649551392,
      "eval_rewards/margins": 7.495745658874512,
      "eval_rewards/rejected": -5.994676113128662,
      "eval_runtime": 5.0142,
      "eval_samples_per_second": 1.994,
      "eval_steps_per_second": 0.997,
      "step": 30
    },
    {
      "epoch": 0.35,
      "grad_norm": 13.044546127319336,
      "learning_rate": 9.725004093573342e-05,
      "logits/chosen": -2.972625255584717,
      "logits/rejected": -3.1085734367370605,
      "logps/chosen": -234.9746856689453,
      "logps/rejected": -170.3094024658203,
      "loss": 1.0699,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.7236368656158447,
      "rewards/margins": 2.1611523628234863,
      "rewards/rejected": -3.884788990020752,
      "step": 31
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.612191200256348,
      "learning_rate": 9.705209506155634e-05,
      "logits/chosen": -2.7585391998291016,
      "logits/rejected": -2.7289156913757324,
      "logps/chosen": -283.48614501953125,
      "logps/rejected": -195.35939025878906,
      "loss": 0.1887,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.6527029275894165,
      "rewards/margins": 3.757754325866699,
      "rewards/rejected": -4.410456657409668,
      "step": 32
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.19436502456665,
      "learning_rate": 9.68474862499881e-05,
      "logits/chosen": -2.993574857711792,
      "logits/rejected": -2.428236246109009,
      "logps/chosen": -300.1104431152344,
      "logps/rejected": -282.0811462402344,
      "loss": 0.2097,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.1330656111240387,
      "rewards/margins": 3.342900276184082,
      "rewards/rejected": -3.475965738296509,
      "step": 33
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.4402265548706055,
      "learning_rate": 9.663624347520505e-05,
      "logits/chosen": -3.148404359817505,
      "logits/rejected": -3.0182220935821533,
      "logps/chosen": -304.60955810546875,
      "logps/rejected": -140.74435424804688,
      "loss": 0.3689,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.3897354006767273,
      "rewards/margins": 3.62821102142334,
      "rewards/rejected": -4.017946243286133,
      "step": 34
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.259589195251465,
      "learning_rate": 9.641839665080363e-05,
      "logits/chosen": -2.65729022026062,
      "logits/rejected": -2.7510809898376465,
      "logps/chosen": -330.6309814453125,
      "logps/rejected": -256.57421875,
      "loss": 0.156,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.08656883984804153,
      "rewards/margins": 4.7406206130981445,
      "rewards/rejected": -4.654051780700684,
      "step": 35
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.500273704528809,
      "learning_rate": 9.619397662556435e-05,
      "logits/chosen": -2.995790719985962,
      "logits/rejected": -3.0307507514953613,
      "logps/chosen": -308.2181701660156,
      "logps/rejected": -172.92543029785156,
      "loss": 0.2963,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.4932462275028229,
      "rewards/margins": 2.582528829574585,
      "rewards/rejected": -3.075775146484375,
      "step": 36
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.707283973693848,
      "learning_rate": 9.596301517908328e-05,
      "logits/chosen": -3.033268451690674,
      "logits/rejected": -3.0536768436431885,
      "logps/chosen": -245.74594116210938,
      "logps/rejected": -160.86160278320312,
      "loss": 0.202,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.2917206287384033,
      "rewards/margins": 4.046570777893066,
      "rewards/rejected": -4.338291645050049,
      "step": 37
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.139472007751465,
      "learning_rate": 9.572554501727198e-05,
      "logits/chosen": -3.1614038944244385,
      "logits/rejected": -2.7908077239990234,
      "logps/chosen": -297.92730712890625,
      "logps/rejected": -190.27981567382812,
      "loss": 0.1892,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.4447726905345917,
      "rewards/margins": 4.278682708740234,
      "rewards/rejected": -4.723455429077148,
      "step": 38
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.791252851486206,
      "learning_rate": 9.548159976772592e-05,
      "logits/chosen": -3.0960066318511963,
      "logits/rejected": -2.7181763648986816,
      "logps/chosen": -223.69406127929688,
      "logps/rejected": -123.40463256835938,
      "loss": 0.1746,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.49881893396377563,
      "rewards/margins": 3.9385018348693848,
      "rewards/rejected": -4.437320709228516,
      "step": 39
    },
    {
      "epoch": 0.45,
      "grad_norm": 11.180024147033691,
      "learning_rate": 9.523121397496269e-05,
      "logits/chosen": -3.0495779514312744,
      "logits/rejected": -2.918321132659912,
      "logps/chosen": -219.83587646484375,
      "logps/rejected": -188.59054565429688,
      "loss": 0.4327,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.5894317626953125,
      "rewards/margins": 2.6391894817352295,
      "rewards/rejected": -3.228621244430542,
      "step": 40
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.4187703132629395,
      "learning_rate": 9.497442309553016e-05,
      "logits/chosen": -2.807313919067383,
      "logits/rejected": -2.6464860439300537,
      "logps/chosen": -189.6230010986328,
      "logps/rejected": -151.01107788085938,
      "loss": 0.693,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -3.1926302909851074,
      "rewards/margins": 2.0725247859954834,
      "rewards/rejected": -5.26515531539917,
      "step": 41
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.239777565002441,
      "learning_rate": 9.471126349298556e-05,
      "logits/chosen": -2.6991612911224365,
      "logits/rejected": -2.4860870838165283,
      "logps/chosen": -249.05978393554688,
      "logps/rejected": -261.3404541015625,
      "loss": 0.3519,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -2.2070960998535156,
      "rewards/margins": 2.7090442180633545,
      "rewards/rejected": -4.916140079498291,
      "step": 42
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.044631958007812,
      "learning_rate": 9.444177243274618e-05,
      "logits/chosen": -3.069183111190796,
      "logits/rejected": -3.239974021911621,
      "logps/chosen": -191.36724853515625,
      "logps/rejected": -203.04444885253906,
      "loss": 0.8501,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -2.2000980377197266,
      "rewards/margins": 3.8068671226501465,
      "rewards/rejected": -6.006965160369873,
      "step": 43
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.961316108703613,
      "learning_rate": 9.41659880768122e-05,
      "logits/chosen": -3.117356538772583,
      "logits/rejected": -3.1276466846466064,
      "logps/chosen": -241.28115844726562,
      "logps/rejected": -217.63143920898438,
      "loss": 0.5688,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -1.914351463317871,
      "rewards/margins": 3.1862049102783203,
      "rewards/rejected": -5.100556373596191,
      "step": 44
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.185661315917969,
      "learning_rate": 9.388394947836279e-05,
      "logits/chosen": -3.105802536010742,
      "logits/rejected": -2.782811164855957,
      "logps/chosen": -346.5468444824219,
      "logps/rejected": -281.4388732910156,
      "loss": 0.5225,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.3682477474212646,
      "rewards/margins": 1.8964588642120361,
      "rewards/rejected": -3.264706611633301,
      "step": 45
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.120816230773926,
      "learning_rate": 9.359569657622574e-05,
      "logits/chosen": -3.119126319885254,
      "logits/rejected": -3.2095906734466553,
      "logps/chosen": -529.1534423828125,
      "logps/rejected": -279.0389404296875,
      "loss": 0.3314,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.4428637623786926,
      "rewards/margins": 6.021182060241699,
      "rewards/rejected": -5.578318119049072,
      "step": 46
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.270225524902344,
      "learning_rate": 9.330127018922194e-05,
      "logits/chosen": -2.903927803039551,
      "logits/rejected": -2.7225542068481445,
      "logps/chosen": -275.26318359375,
      "logps/rejected": -169.46505737304688,
      "loss": 0.6252,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -1.2819722890853882,
      "rewards/margins": 2.14062237739563,
      "rewards/rejected": -3.4225947856903076,
      "step": 47
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.7271859645843506,
      "learning_rate": 9.300071201038503e-05,
      "logits/chosen": -2.98264741897583,
      "logits/rejected": -2.486337184906006,
      "logps/chosen": -257.7012023925781,
      "logps/rejected": -152.14598083496094,
      "loss": 0.1543,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.4061692953109741,
      "rewards/margins": 4.309848785400391,
      "rewards/rejected": -4.716017723083496,
      "step": 48
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.926673889160156,
      "learning_rate": 9.26940646010574e-05,
      "logits/chosen": -3.2901933193206787,
      "logits/rejected": -2.8755884170532227,
      "logps/chosen": -191.0556640625,
      "logps/rejected": -104.35969543457031,
      "loss": 0.1687,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.4329254627227783,
      "rewards/margins": 4.122668266296387,
      "rewards/rejected": -4.555593490600586,
      "step": 49
    },
    {
      "epoch": 0.56,
      "grad_norm": NaN,
      "learning_rate": 9.26940646010574e-05,
      "logits/chosen": -3.4009902477264404,
      "logits/rejected": -3.2163825035095215,
      "logps/chosen": -228.1378936767578,
      "logps/rejected": -145.276123046875,
      "loss": 0.3099,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.2526423931121826,
      "rewards/margins": 4.154683589935303,
      "rewards/rejected": -3.90204119682312,
      "step": 50
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.389315128326416,
      "learning_rate": 9.238137138486318e-05,
      "logits/chosen": -3.3087496757507324,
      "logits/rejected": -3.0156049728393555,
      "logps/chosen": -367.84161376953125,
      "logps/rejected": -198.76544189453125,
      "loss": 0.1843,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.8194153308868408,
      "rewards/margins": 4.259294033050537,
      "rewards/rejected": -3.439878463745117,
      "step": 51
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.2384748458862305,
      "learning_rate": 9.206267664155907e-05,
      "logits/chosen": -3.412761926651001,
      "logits/rejected": -3.3606858253479004,
      "logps/chosen": -394.595947265625,
      "logps/rejected": -234.29969787597656,
      "loss": 0.0957,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6140856742858887,
      "rewards/margins": 3.623807191848755,
      "rewards/rejected": -3.009721517562866,
      "step": 52
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.3274507522583,
      "learning_rate": 9.173802550076401e-05,
      "logits/chosen": -2.9212708473205566,
      "logits/rejected": -3.2253193855285645,
      "logps/chosen": -309.66741943359375,
      "logps/rejected": -194.5745849609375,
      "loss": 0.3459,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.11235330998897552,
      "rewards/margins": 2.6622633934020996,
      "rewards/rejected": -2.549910068511963,
      "step": 53
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.517582893371582,
      "learning_rate": 9.140746393556854e-05,
      "logits/chosen": -3.4288511276245117,
      "logits/rejected": -3.304213285446167,
      "logps/chosen": -376.1202392578125,
      "logps/rejected": -233.58489990234375,
      "loss": 0.5965,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.555878758430481,
      "rewards/margins": 4.904500961303711,
      "rewards/rejected": -4.3486223220825195,
      "step": 54
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8991745710372925,
      "learning_rate": 9.107103875602459e-05,
      "logits/chosen": -2.924043655395508,
      "logits/rejected": -3.185393810272217,
      "logps/chosen": -341.2203063964844,
      "logps/rejected": -234.63758850097656,
      "loss": 0.0427,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3599395751953125,
      "rewards/margins": 3.722670555114746,
      "rewards/rejected": -3.3627307415008545,
      "step": 55
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.762215614318848,
      "learning_rate": 9.072879760251679e-05,
      "logits/chosen": -3.2768988609313965,
      "logits/rejected": -3.0862388610839844,
      "logps/chosen": -224.249755859375,
      "logps/rejected": -180.57443237304688,
      "loss": 0.3701,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.7867403626441956,
      "rewards/margins": 2.9776391983032227,
      "rewards/rejected": -3.7643797397613525,
      "step": 56
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.915589332580566,
      "learning_rate": 9.038078893901634e-05,
      "logits/chosen": -3.2617459297180176,
      "logits/rejected": -3.202209711074829,
      "logps/chosen": -89.90280151367188,
      "logps/rejected": -147.65475463867188,
      "loss": 0.4777,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.5283554792404175,
      "rewards/margins": 2.442681312561035,
      "rewards/rejected": -2.971036911010742,
      "step": 57
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.555606842041016,
      "learning_rate": 9.002706204621803e-05,
      "logits/chosen": -3.001934766769409,
      "logits/rejected": -3.068469285964966,
      "logps/chosen": -344.8009948730469,
      "logps/rejected": -205.77664184570312,
      "loss": 0.3874,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.012376070022583008,
      "rewards/margins": 2.2519307136535645,
      "rewards/rejected": -2.2395546436309814,
      "step": 58
    },
    {
      "epoch": 0.66,
      "grad_norm": 14.009621620178223,
      "learning_rate": 8.966766701456177e-05,
      "logits/chosen": -2.8611111640930176,
      "logits/rejected": -3.4576072692871094,
      "logps/chosen": -338.19970703125,
      "logps/rejected": -260.2259521484375,
      "loss": 0.7432,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.9131150841712952,
      "rewards/margins": 2.4023218154907227,
      "rewards/rejected": -3.315437078475952,
      "step": 59
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.895538330078125,
      "learning_rate": 8.930265473713938e-05,
      "logits/chosen": -3.3844141960144043,
      "logits/rejected": -3.3034396171569824,
      "logps/chosen": -369.8033447265625,
      "logps/rejected": -244.80897521972656,
      "loss": 0.2099,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.4544011354446411,
      "rewards/margins": 2.975295305252075,
      "rewards/rejected": -2.5208940505981445,
      "step": 60
    },
    {
      "epoch": 0.68,
      "eval_logits/chosen": -3.299164295196533,
      "eval_logits/rejected": -3.039053440093994,
      "eval_logps/chosen": -182.6027069091797,
      "eval_logps/rejected": -194.1580810546875,
      "eval_loss": 0.0008067694725468755,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 2.757256269454956,
      "eval_rewards/margins": 8.322589874267578,
      "eval_rewards/rejected": -5.565333366394043,
      "eval_runtime": 4.7427,
      "eval_samples_per_second": 2.108,
      "eval_steps_per_second": 1.054,
      "step": 60
    },
    {
      "epoch": 0.69,
      "grad_norm": 13.20186710357666,
      "learning_rate": 8.893207690248776e-05,
      "logits/chosen": -3.109037399291992,
      "logits/rejected": -3.326498508453369,
      "logps/chosen": -275.9548034667969,
      "logps/rejected": -253.3433380126953,
      "loss": 0.6256,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.315821647644043,
      "rewards/margins": 1.796335220336914,
      "rewards/rejected": -3.112156867980957,
      "step": 61
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.376776933670044,
      "learning_rate": 8.855598598726939e-05,
      "logits/chosen": -3.083179473876953,
      "logits/rejected": -3.2115023136138916,
      "logps/chosen": -234.42640686035156,
      "logps/rejected": -150.53904724121094,
      "loss": 0.0802,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.07728518545627594,
      "rewards/margins": 4.584873199462891,
      "rewards/rejected": -4.507587909698486,
      "step": 62
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.479050636291504,
      "learning_rate": 8.817443524884119e-05,
      "logits/chosen": -2.5378470420837402,
      "logits/rejected": -2.813725471496582,
      "logps/chosen": -321.4371337890625,
      "logps/rejected": -244.43296813964844,
      "loss": 0.2551,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.725128173828125,
      "rewards/margins": 6.416964530944824,
      "rewards/rejected": -5.691835880279541,
      "step": 63
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.595841407775879,
      "learning_rate": 8.778747871771292e-05,
      "logits/chosen": -3.3628735542297363,
      "logits/rejected": -2.6477882862091064,
      "logps/chosen": -491.04852294921875,
      "logps/rejected": -248.0843505859375,
      "loss": 0.4921,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.3514632284641266,
      "rewards/margins": 7.0789079666137695,
      "rewards/rejected": -6.727444648742676,
      "step": 64
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.204965591430664,
      "learning_rate": 8.739517118989605e-05,
      "logits/chosen": -3.0609679222106934,
      "logits/rejected": -3.0792436599731445,
      "logps/chosen": -297.3428649902344,
      "logps/rejected": -190.6544647216797,
      "loss": 0.5732,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.3494699001312256,
      "rewards/margins": 2.159775733947754,
      "rewards/rejected": -3.5092456340789795,
      "step": 65
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.17355728149414,
      "learning_rate": 8.69975682191442e-05,
      "logits/chosen": -3.373033046722412,
      "logits/rejected": -3.043926239013672,
      "logps/chosen": -578.4942626953125,
      "logps/rejected": -265.8677978515625,
      "loss": 0.3497,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.6642241477966309,
      "rewards/margins": 6.849948883056641,
      "rewards/rejected": -7.5141730308532715,
      "step": 66
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.397715091705322,
      "learning_rate": 8.659472610908627e-05,
      "logits/chosen": -2.711515426635742,
      "logits/rejected": -2.395803928375244,
      "logps/chosen": -203.78604125976562,
      "logps/rejected": -129.5730438232422,
      "loss": 0.1589,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -1.117239236831665,
      "rewards/margins": 4.5233540534973145,
      "rewards/rejected": -5.640593528747559,
      "step": 67
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.30370500683784485,
      "learning_rate": 8.618670190525352e-05,
      "logits/chosen": -3.0820484161376953,
      "logits/rejected": -3.0193471908569336,
      "logps/chosen": -378.7370300292969,
      "logps/rejected": -222.33255004882812,
      "loss": 0.0083,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.245200514793396,
      "rewards/margins": 7.6438798904418945,
      "rewards/rejected": -7.398679733276367,
      "step": 68
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.54161548614502,
      "learning_rate": 8.577355338700132e-05,
      "logits/chosen": -3.2389307022094727,
      "logits/rejected": -3.0821778774261475,
      "logps/chosen": -346.2233581542969,
      "logps/rejected": -192.50180053710938,
      "loss": 0.7383,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -1.195668339729309,
      "rewards/margins": 5.02830696105957,
      "rewards/rejected": -6.22397518157959,
      "step": 69
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.390268087387085,
      "learning_rate": 8.535533905932738e-05,
      "logits/chosen": -3.3620431423187256,
      "logits/rejected": -2.399602174758911,
      "logps/chosen": -221.27833557128906,
      "logps/rejected": -118.21566009521484,
      "loss": 0.0245,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3484119176864624,
      "rewards/margins": 8.018238067626953,
      "rewards/rejected": -7.669825553894043,
      "step": 70
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.928914070129395,
      "learning_rate": 8.493211814458673e-05,
      "logits/chosen": -2.9451489448547363,
      "logits/rejected": -3.133699893951416,
      "logps/chosen": -201.42678833007812,
      "logps/rejected": -202.67453002929688,
      "loss": 0.3648,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -2.156554698944092,
      "rewards/margins": 4.664039611816406,
      "rewards/rejected": -6.820594310760498,
      "step": 71
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.7106189131736755,
      "learning_rate": 8.450395057410561e-05,
      "logits/chosen": -3.4210917949676514,
      "logits/rejected": -2.9528050422668457,
      "logps/chosen": -317.3839416503906,
      "logps/rejected": -170.9514617919922,
      "loss": 0.0172,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.5772140026092529,
      "rewards/margins": 6.574632167816162,
      "rewards/rejected": -7.151845455169678,
      "step": 72
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.109986305236816,
      "learning_rate": 8.407089697969457e-05,
      "logits/chosen": -3.1691198348999023,
      "logits/rejected": -2.9826912879943848,
      "logps/chosen": -377.9816589355469,
      "logps/rejected": -288.35528564453125,
      "loss": 0.4652,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.14444704353809357,
      "rewards/margins": 4.671531677246094,
      "rewards/rejected": -4.815978527069092,
      "step": 73
    },
    {
      "epoch": 0.83,
      "grad_norm": 16.743736267089844,
      "learning_rate": 8.363301868506264e-05,
      "logits/chosen": -3.06179141998291,
      "logits/rejected": -2.977768898010254,
      "logps/chosen": -336.5064392089844,
      "logps/rejected": -273.6090393066406,
      "loss": 0.9035,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.7004997730255127,
      "rewards/margins": 3.5526602268218994,
      "rewards/rejected": -5.253159999847412,
      "step": 74
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.297181606292725,
      "learning_rate": 8.319037769713338e-05,
      "logits/chosen": -3.1873130798339844,
      "logits/rejected": -3.2429358959198,
      "logps/chosen": -295.90167236328125,
      "logps/rejected": -252.4221954345703,
      "loss": 0.0913,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.35775870084762573,
      "rewards/margins": 7.1074934005737305,
      "rewards/rejected": -7.46525239944458,
      "step": 75
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.5024635791778564,
      "learning_rate": 8.274303669726426e-05,
      "logits/chosen": -3.2978920936584473,
      "logits/rejected": -3.056913375854492,
      "logps/chosen": -209.08226013183594,
      "logps/rejected": -216.75137329101562,
      "loss": 0.0595,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.7533246278762817,
      "rewards/margins": 6.2144622802734375,
      "rewards/rejected": -6.96778678894043,
      "step": 76
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.813451766967773,
      "learning_rate": 8.229105903237044e-05,
      "logits/chosen": -3.1069507598876953,
      "logits/rejected": -2.643913507461548,
      "logps/chosen": -379.9039611816406,
      "logps/rejected": -199.5642547607422,
      "loss": 0.2955,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -2.039405345916748,
      "rewards/margins": 6.617141246795654,
      "rewards/rejected": -8.656547546386719,
      "step": 77
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.086407661437988,
      "learning_rate": 8.183450870595441e-05,
      "logits/chosen": -3.06406307220459,
      "logits/rejected": -3.0317187309265137,
      "logps/chosen": -369.6795654296875,
      "logps/rejected": -303.2950744628906,
      "loss": 0.4151,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.9069956541061401,
      "rewards/margins": 4.3820600509643555,
      "rewards/rejected": -6.289055824279785,
      "step": 78
    },
    {
      "epoch": 0.89,
      "grad_norm": 3.3662607669830322,
      "learning_rate": 8.13734503690426e-05,
      "logits/chosen": -3.289384126663208,
      "logits/rejected": -3.1391568183898926,
      "logps/chosen": -345.8055114746094,
      "logps/rejected": -193.66909790039062,
      "loss": 0.1584,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.599603533744812,
      "rewards/margins": 6.25246524810791,
      "rewards/rejected": -6.852068901062012,
      "step": 79
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.315209865570068,
      "learning_rate": 8.090794931103026e-05,
      "logits/chosen": -3.228783369064331,
      "logits/rejected": -3.0682406425476074,
      "logps/chosen": -270.1308898925781,
      "logps/rejected": -199.65390014648438,
      "loss": 0.197,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -1.2517391443252563,
      "rewards/margins": 5.22841215133667,
      "rewards/rejected": -6.480151176452637,
      "step": 80
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.3422465324401855,
      "learning_rate": 8.043807145043604e-05,
      "logits/chosen": -3.3883039951324463,
      "logits/rejected": -2.8610763549804688,
      "logps/chosen": -375.4300231933594,
      "logps/rejected": -212.55474853515625,
      "loss": 0.0891,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.205045461654663,
      "rewards/margins": 7.316597938537598,
      "rewards/rejected": -8.521642684936523,
      "step": 81
    },
    {
      "epoch": 0.92,
      "grad_norm": 14.285599708557129,
      "learning_rate": 7.996388332556735e-05,
      "logits/chosen": -3.384727954864502,
      "logits/rejected": -3.3384907245635986,
      "logps/chosen": -403.5943603515625,
      "logps/rejected": -331.4810791015625,
      "loss": 0.6671,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -1.1636757850646973,
      "rewards/margins": 2.9779176712036133,
      "rewards/rejected": -4.1415934562683105,
      "step": 82
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.26110827922821045,
      "learning_rate": 7.94854520850981e-05,
      "logits/chosen": -3.1430768966674805,
      "logits/rejected": -3.018350601196289,
      "logps/chosen": -502.6695556640625,
      "logps/rejected": -263.3882751464844,
      "loss": 0.0062,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0688759088516235,
      "rewards/margins": 9.059045791625977,
      "rewards/rejected": -7.990170001983643,
      "step": 83
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.293683052062988,
      "learning_rate": 7.900284547855991e-05,
      "logits/chosen": -3.26918363571167,
      "logits/rejected": -2.9317312240600586,
      "logps/chosen": -287.8534851074219,
      "logps/rejected": -225.27447509765625,
      "loss": 0.1985,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.16245371103286743,
      "rewards/margins": 6.57786750793457,
      "rewards/rejected": -6.415413856506348,
      "step": 84
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.314470648765564,
      "learning_rate": 7.85161318467482e-05,
      "logits/chosen": -3.219449043273926,
      "logits/rejected": -2.8287854194641113,
      "logps/chosen": -366.5806884765625,
      "logps/rejected": -191.7723388671875,
      "loss": 0.0161,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.41258037090301514,
      "rewards/margins": 6.966887474060059,
      "rewards/rejected": -6.554306983947754,
      "step": 85
    },
    {
      "epoch": 0.97,
      "grad_norm": 20.908584594726562,
      "learning_rate": 7.80253801120447e-05,
      "logits/chosen": -3.160271406173706,
      "logits/rejected": -3.0260157585144043,
      "logps/chosen": -382.6925048828125,
      "logps/rejected": -233.9979248046875,
      "loss": 0.7179,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.4525451958179474,
      "rewards/margins": 2.8419482707977295,
      "rewards/rejected": -3.2944936752319336,
      "step": 86
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.2079492807388306,
      "learning_rate": 7.753065976865744e-05,
      "logits/chosen": -3.191283702850342,
      "logits/rejected": -3.0339455604553223,
      "logps/chosen": -285.8114013671875,
      "logps/rejected": -169.56741333007812,
      "loss": 0.0272,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.33831024169921875,
      "rewards/margins": 6.575672149658203,
      "rewards/rejected": -6.237361907958984,
      "step": 87
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.654447615146637,
      "learning_rate": 7.703204087277988e-05,
      "logits/chosen": -3.2184348106384277,
      "logits/rejected": -3.2598187923431396,
      "logps/chosen": -163.65548706054688,
      "logps/rejected": -143.7698974609375,
      "loss": 0.0114,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 2.0235941410064697,
      "rewards/margins": 7.350224494934082,
      "rewards/rejected": -5.326630592346191,
      "step": 88
    }
  ],
  "logging_steps": 1,
  "max_steps": 264,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}