{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.6,
  "eval_steps": 5,
  "global_step": 130,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.024615384615384615,
      "grad_norm": 51.137206373851456,
      "learning_rate": 5.88235294117647e-08,
      "logits/chosen": -0.5265347361564636,
      "logits/rejected": -0.5229614973068237,
      "logps/chosen": -40.1543083190918,
      "logps/rejected": -40.267845153808594,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 2
    },
    {
      "epoch": 0.04923076923076923,
      "grad_norm": 53.426708668382766,
      "learning_rate": 1.176470588235294e-07,
      "logits/chosen": -0.5090410113334656,
      "logits/rejected": -0.502047061920166,
      "logps/chosen": -45.320037841796875,
      "logps/rejected": -42.6551399230957,
      "loss": 0.6918,
      "rewards/accuracies": 0.453125,
      "rewards/chosen": 0.014461740851402283,
      "rewards/margins": 0.0072987377643585205,
      "rewards/rejected": 0.007163003087043762,
      "step": 4
    },
    {
      "epoch": 0.06153846153846154,
      "eval_logits/chosen": -0.49676236510276794,
      "eval_logits/rejected": -0.4936961829662323,
      "eval_logps/chosen": -45.20378875732422,
      "eval_logps/rejected": -42.88367462158203,
      "eval_loss": 0.6931462287902832,
      "eval_rewards/accuracies": 0.4623287618160248,
      "eval_rewards/chosen": 0.0008219562005251646,
      "eval_rewards/margins": 0.0035426486283540726,
      "eval_rewards/rejected": -0.002720692427828908,
      "eval_runtime": 400.2137,
      "eval_samples_per_second": 4.333,
      "eval_steps_per_second": 0.182,
      "step": 5
    },
    {
      "epoch": 0.07384615384615385,
      "grad_norm": 47.768106379014085,
      "learning_rate": 1.764705882352941e-07,
      "logits/chosen": -0.4896703064441681,
      "logits/rejected": -0.48678484559059143,
      "logps/chosen": -47.16059875488281,
      "logps/rejected": -35.45697021484375,
      "loss": 0.6958,
      "rewards/accuracies": 0.484375,
      "rewards/chosen": 0.007428005337715149,
      "rewards/margins": 0.012338653206825256,
      "rewards/rejected": -0.004910647869110107,
      "step": 6
    },
    {
      "epoch": 0.09846153846153846,
      "grad_norm": 39.51012410260543,
      "learning_rate": 2.352941176470588e-07,
      "logits/chosen": -0.4996076822280884,
      "logits/rejected": -0.49669766426086426,
      "logps/chosen": -47.94187545776367,
      "logps/rejected": -41.556766510009766,
      "loss": 0.6914,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.018062740564346313,
      "rewards/margins": 0.025705993175506592,
      "rewards/rejected": -0.007643252611160278,
      "step": 8
    },
    {
      "epoch": 0.12307692307692308,
      "grad_norm": 39.08254813329663,
      "learning_rate": 2.941176470588235e-07,
      "logits/chosen": -0.49092355370521545,
      "logits/rejected": -0.48874035477638245,
      "logps/chosen": -50.65516662597656,
      "logps/rejected": -30.806886672973633,
      "loss": 0.6842,
      "rewards/accuracies": 0.515625,
      "rewards/chosen": 0.016676604747772217,
      "rewards/margins": 0.011630743741989136,
      "rewards/rejected": 0.005045861005783081,
      "step": 10
    },
    {
      "epoch": 0.12307692307692308,
      "eval_logits/chosen": -0.49970191717147827,
      "eval_logits/rejected": -0.4963047206401825,
      "eval_logps/chosen": -45.12245178222656,
      "eval_logps/rejected": -42.913116455078125,
      "eval_loss": 0.6740441918373108,
      "eval_rewards/accuracies": 0.5410959124565125,
      "eval_rewards/chosen": 0.041489191353321075,
      "eval_rewards/margins": 0.058929312974214554,
      "eval_rewards/rejected": -0.017440123483538628,
      "eval_runtime": 398.3897,
      "eval_samples_per_second": 4.353,
      "eval_steps_per_second": 0.183,
      "step": 10
    },
    {
      "epoch": 0.1476923076923077,
      "grad_norm": 38.09131172735928,
      "learning_rate": 3.529411764705882e-07,
      "logits/chosen": -0.48407822847366333,
      "logits/rejected": -0.4791830778121948,
      "logps/chosen": -45.49330139160156,
      "logps/rejected": -42.8103141784668,
      "loss": 0.6725,
      "rewards/accuracies": 0.609375,
      "rewards/chosen": 0.04486130177974701,
      "rewards/margins": 0.06311255693435669,
      "rewards/rejected": -0.01825125515460968,
      "step": 12
    },
    {
      "epoch": 0.1723076923076923,
      "grad_norm": 35.41525379746354,
      "learning_rate": 4.117647058823529e-07,
      "logits/chosen": -0.5208038091659546,
      "logits/rejected": -0.5175592303276062,
      "logps/chosen": -40.418609619140625,
      "logps/rejected": -34.53902816772461,
      "loss": 0.6501,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": 0.11583159863948822,
      "rewards/margins": 0.07396885752677917,
      "rewards/rejected": 0.041862741112709045,
      "step": 14
    },
    {
      "epoch": 0.18461538461538463,
      "eval_logits/chosen": -0.4988827705383301,
      "eval_logits/rejected": -0.4956616461277008,
      "eval_logps/chosen": -44.772987365722656,
      "eval_logps/rejected": -42.95915985107422,
      "eval_loss": 0.6234598159790039,
      "eval_rewards/accuracies": 0.6404109597206116,
      "eval_rewards/chosen": 0.2162197083234787,
      "eval_rewards/margins": 0.25668132305145264,
      "eval_rewards/rejected": -0.040461596101522446,
      "eval_runtime": 398.8168,
      "eval_samples_per_second": 4.348,
      "eval_steps_per_second": 0.183,
      "step": 15
    },
    {
      "epoch": 0.19692307692307692,
      "grad_norm": 33.20243289262748,
      "learning_rate": 4.705882352941176e-07,
      "logits/chosen": -0.49464067816734314,
      "logits/rejected": -0.4926251173019409,
      "logps/chosen": -41.81456756591797,
      "logps/rejected": -34.730552673339844,
      "loss": 0.6139,
      "rewards/accuracies": 0.734375,
      "rewards/chosen": 0.22246630489826202,
      "rewards/margins": 0.1789122372865677,
      "rewards/rejected": 0.043554067611694336,
      "step": 16
    },
    {
      "epoch": 0.22153846153846155,
      "grad_norm": 31.414401498072234,
      "learning_rate": 4.99941324504621e-07,
      "logits/chosen": -0.5083720684051514,
      "logits/rejected": -0.5054486393928528,
      "logps/chosen": -40.7899169921875,
      "logps/rejected": -41.80813217163086,
      "loss": 0.6038,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.22346775233745575,
      "rewards/margins": 0.36676231026649475,
      "rewards/rejected": -0.143294557929039,
      "step": 18
    },
    {
      "epoch": 0.24615384615384617,
      "grad_norm": 26.94475877791274,
      "learning_rate": 4.99472085783721e-07,
      "logits/chosen": -0.5102497935295105,
      "logits/rejected": -0.5054219961166382,
      "logps/chosen": -38.036109924316406,
      "logps/rejected": -41.01247787475586,
      "loss": 0.5437,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.5671411752700806,
      "rewards/margins": 0.6098884344100952,
      "rewards/rejected": -0.04274718463420868,
      "step": 20
    },
    {
      "epoch": 0.24615384615384617,
      "eval_logits/chosen": -0.4962932765483856,
      "eval_logits/rejected": -0.49319401383399963,
      "eval_logps/chosen": -43.80841827392578,
      "eval_logps/rejected": -42.96743392944336,
      "eval_loss": 0.5607677698135376,
      "eval_rewards/accuracies": 0.6952054500579834,
      "eval_rewards/chosen": 0.6985027194023132,
      "eval_rewards/margins": 0.7430992722511292,
      "eval_rewards/rejected": -0.04459657892584801,
      "eval_runtime": 398.3125,
      "eval_samples_per_second": 4.353,
      "eval_steps_per_second": 0.183,
      "step": 20
    },
    {
      "epoch": 0.27076923076923076,
      "grad_norm": 24.12367725777412,
      "learning_rate": 4.985344892885899e-07,
      "logits/chosen": -0.5035347938537598,
      "logits/rejected": -0.5017133951187134,
      "logps/chosen": -42.291107177734375,
      "logps/rejected": -43.623016357421875,
      "loss": 0.5756,
      "rewards/accuracies": 0.734375,
      "rewards/chosen": 0.6317887306213379,
      "rewards/margins": 0.7202781438827515,
      "rewards/rejected": -0.08848930895328522,
      "step": 22
    },
    {
      "epoch": 0.2953846153846154,
      "grad_norm": 24.794972937394068,
      "learning_rate": 4.971302952586796e-07,
      "logits/chosen": -0.5023949146270752,
      "logits/rejected": -0.49643266201019287,
      "logps/chosen": -39.85929489135742,
      "logps/rejected": -46.9812126159668,
      "loss": 0.5219,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.8194392919540405,
      "rewards/margins": 1.2944234609603882,
      "rewards/rejected": -0.4749842882156372,
      "step": 24
    },
    {
      "epoch": 0.3076923076923077,
      "eval_logits/chosen": -0.4914375841617584,
      "eval_logits/rejected": -0.4883214831352234,
      "eval_logps/chosen": -43.38737869262695,
      "eval_logps/rejected": -43.16010284423828,
      "eval_loss": 0.516257107257843,
      "eval_rewards/accuracies": 0.7397260069847107,
      "eval_rewards/chosen": 0.9090243577957153,
      "eval_rewards/margins": 1.0499579906463623,
      "eval_rewards/rejected": -0.14093351364135742,
      "eval_runtime": 397.9655,
      "eval_samples_per_second": 4.357,
      "eval_steps_per_second": 0.183,
      "step": 25
    },
    {
      "epoch": 0.32,
      "grad_norm": 25.854991033804026,
      "learning_rate": 4.952621399215597e-07,
      "logits/chosen": -0.49028468132019043,
      "logits/rejected": -0.48677563667297363,
      "logps/chosen": -39.38325500488281,
      "logps/rejected": -30.141273498535156,
      "loss": 0.5432,
      "rewards/accuracies": 0.765625,
      "rewards/chosen": 0.9173061847686768,
      "rewards/margins": 0.6880545020103455,
      "rewards/rejected": 0.2292517125606537,
      "step": 26
    },
    {
      "epoch": 0.3446153846153846,
      "grad_norm": 22.546249636399228,
      "learning_rate": 4.929335305436764e-07,
      "logits/chosen": -0.4758816063404083,
      "logits/rejected": -0.47193384170532227,
      "logps/chosen": -37.941322326660156,
      "logps/rejected": -36.9672966003418,
      "loss": 0.4989,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 0.9205312132835388,
      "rewards/margins": 0.96444171667099,
      "rewards/rejected": -0.0439104288816452,
      "step": 28
    },
    {
      "epoch": 0.36923076923076925,
      "grad_norm": 25.277496467379798,
      "learning_rate": 4.901488388458247e-07,
      "logits/chosen": -0.47848933935165405,
      "logits/rejected": -0.4744369089603424,
      "logps/chosen": -42.54041290283203,
      "logps/rejected": -32.181236267089844,
      "loss": 0.5002,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.949561595916748,
      "rewards/margins": 0.7569789886474609,
      "rewards/rejected": 0.19258266687393188,
      "step": 30
    },
    {
      "epoch": 0.36923076923076925,
      "eval_logits/chosen": -0.48240911960601807,
      "eval_logits/rejected": -0.47941771149635315,
      "eval_logps/chosen": -42.97114944458008,
      "eval_logps/rejected": -43.50387191772461,
      "eval_loss": 0.4467584788799286,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": 1.1171396970748901,
      "eval_rewards/margins": 1.429957389831543,
      "eval_rewards/rejected": -0.3128177523612976,
      "eval_runtime": 399.6945,
      "eval_samples_per_second": 4.338,
      "eval_steps_per_second": 0.183,
      "step": 30
    },
    {
      "epoch": 0.39384615384615385,
      "grad_norm": 21.934080715012698,
      "learning_rate": 4.869132927957006e-07,
      "logits/chosen": -0.47030550241470337,
      "logits/rejected": -0.4669092893600464,
      "logps/chosen": -39.564884185791016,
      "logps/rejected": -46.268707275390625,
      "loss": 0.4561,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.1370558738708496,
      "rewards/margins": 1.5885730981826782,
      "rewards/rejected": -0.45151710510253906,
      "step": 32
    },
    {
      "epoch": 0.41846153846153844,
      "grad_norm": 17.9530062267105,
      "learning_rate": 4.832329667929376e-07,
      "logits/chosen": -0.48381492495536804,
      "logits/rejected": -0.47740432620048523,
      "logps/chosen": -37.79485321044922,
      "logps/rejected": -41.405555725097656,
      "loss": 0.3974,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 1.0800038576126099,
      "rewards/margins": 1.8046053647994995,
      "rewards/rejected": -0.7246013879776001,
      "step": 34
    },
    {
      "epoch": 0.4307692307692308,
      "eval_logits/chosen": -0.4790378212928772,
      "eval_logits/rejected": -0.4759667217731476,
      "eval_logps/chosen": -42.852088928222656,
      "eval_logps/rejected": -43.939449310302734,
      "eval_loss": 0.3961867392063141,
      "eval_rewards/accuracies": 0.7602739930152893,
      "eval_rewards/chosen": 1.1766709089279175,
      "eval_rewards/margins": 1.7072778940200806,
      "eval_rewards/rejected": -0.5306068062782288,
      "eval_runtime": 402.2584,
      "eval_samples_per_second": 4.311,
      "eval_steps_per_second": 0.181,
      "step": 35
    },
    {
      "epoch": 0.4430769230769231,
      "grad_norm": 18.01450637894319,
      "learning_rate": 4.791147702650565e-07,
      "logits/chosen": -0.4862886071205139,
      "logits/rejected": -0.48092207312583923,
      "logps/chosen": -40.004295349121094,
      "logps/rejected": -39.387386322021484,
      "loss": 0.4067,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.3134092092514038,
      "rewards/margins": 1.8237063884735107,
      "rewards/rejected": -0.5102972984313965,
      "step": 36
    },
    {
      "epoch": 0.4676923076923077,
      "grad_norm": 18.54804560370934,
      "learning_rate": 4.745664346957361e-07,
      "logits/chosen": -0.46789872646331787,
      "logits/rejected": -0.4679996967315674,
      "logps/chosen": -42.09858322143555,
      "logps/rejected": -31.129802703857422,
      "loss": 0.3702,
      "rewards/accuracies": 0.796875,
      "rewards/chosen": 1.3026843070983887,
      "rewards/margins": 1.3334376811981201,
      "rewards/rejected": -0.030753374099731445,
      "step": 38
    },
    {
      "epoch": 0.49230769230769234,
      "grad_norm": 13.439566896592863,
      "learning_rate": 4.695964991097616e-07,
      "logits/chosen": -0.449634313583374,
      "logits/rejected": -0.4447551965713501,
      "logps/chosen": -36.52521514892578,
      "logps/rejected": -42.37124252319336,
      "loss": 0.3514,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 1.1417609453201294,
      "rewards/margins": 1.7916193008422852,
      "rewards/rejected": -0.6498584747314453,
      "step": 40
    },
    {
      "epoch": 0.49230769230769234,
      "eval_logits/chosen": -0.47975781559944153,
      "eval_logits/rejected": -0.47623777389526367,
      "eval_logps/chosen": -42.85760498046875,
      "eval_logps/rejected": -44.426246643066406,
      "eval_loss": 0.35519111156463623,
      "eval_rewards/accuracies": 0.7876712083816528,
      "eval_rewards/chosen": 1.1739110946655273,
      "eval_rewards/margins": 1.9479175806045532,
      "eval_rewards/rejected": -0.7740064859390259,
      "eval_runtime": 398.0514,
      "eval_samples_per_second": 4.356,
      "eval_steps_per_second": 0.183,
      "step": 40
    },
    {
      "epoch": 0.5169230769230769,
      "grad_norm": 14.688902217061264,
      "learning_rate": 4.642142940418973e-07,
      "logits/chosen": -0.5011200904846191,
      "logits/rejected": -0.49806085228919983,
      "logps/chosen": -37.53756332397461,
      "logps/rejected": -43.31938171386719,
      "loss": 0.3629,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 1.0989573001861572,
      "rewards/margins": 1.9275836944580078,
      "rewards/rejected": -0.8286263346672058,
      "step": 42
    },
    {
      "epoch": 0.5415384615384615,
      "grad_norm": 14.193796961040897,
      "learning_rate": 4.5842992401978256e-07,
      "logits/chosen": -0.4911167323589325,
      "logits/rejected": -0.4909789264202118,
      "logps/chosen": -46.16688537597656,
      "logps/rejected": -35.1140251159668,
      "loss": 0.316,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 1.0063791275024414,
      "rewards/margins": 1.4864487648010254,
      "rewards/rejected": -0.48006969690322876,
      "step": 44
    },
    {
      "epoch": 0.5538461538461539,
      "eval_logits/chosen": -0.472987562417984,
      "eval_logits/rejected": -0.46926581859588623,
      "eval_logps/chosen": -42.71507263183594,
      "eval_logps/rejected": -45.037986755371094,
      "eval_loss": 0.31506600975990295,
      "eval_rewards/accuracies": 0.801369845867157,
      "eval_rewards/chosen": 1.2451775074005127,
      "eval_rewards/margins": 2.3250503540039062,
      "eval_rewards/rejected": -1.0798726081848145,
      "eval_runtime": 398.07,
      "eval_samples_per_second": 4.356,
      "eval_steps_per_second": 0.183,
      "step": 45
    },
    {
      "epoch": 0.5661538461538461,
      "grad_norm": 11.94835713175047,
      "learning_rate": 4.5225424859373684e-07,
      "logits/chosen": -0.48655006289482117,
      "logits/rejected": -0.4802987575531006,
      "logps/chosen": -45.382694244384766,
      "logps/rejected": -34.198219299316406,
      "loss": 0.2964,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 1.3039376735687256,
      "rewards/margins": 1.7199229001998901,
      "rewards/rejected": -0.4159852862358093,
      "step": 46
    },
    {
      "epoch": 0.5907692307692308,
      "grad_norm": 11.665118422994077,
      "learning_rate": 4.456988619490889e-07,
      "logits/chosen": -0.4525223672389984,
      "logits/rejected": -0.4487767517566681,
      "logps/chosen": -42.471214294433594,
      "logps/rejected": -37.61225128173828,
      "loss": 0.2878,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 1.1848539113998413,
      "rewards/margins": 1.884321928024292,
      "rewards/rejected": -0.6994677782058716,
      "step": 48
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 11.208056217338184,
      "learning_rate": 4.3877607113930516e-07,
      "logits/chosen": -0.4693542420864105,
      "logits/rejected": -0.46645450592041016,
      "logps/chosen": -38.541446685791016,
      "logps/rejected": -28.433433532714844,
      "loss": 0.2786,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.3120430707931519,
      "rewards/margins": 1.601030945777893,
      "rewards/rejected": -0.2889878749847412,
      "step": 50
    },
    {
      "epoch": 0.6153846153846154,
      "eval_logits/chosen": -0.46347275376319885,
      "eval_logits/rejected": -0.45981672406196594,
      "eval_logps/chosen": -42.53354263305664,
      "eval_logps/rejected": -45.56825256347656,
      "eval_loss": 0.28788629174232483,
      "eval_rewards/accuracies": 0.818493127822876,
      "eval_rewards/chosen": 1.335942029953003,
      "eval_rewards/margins": 2.6809515953063965,
      "eval_rewards/rejected": -1.345009684562683,
      "eval_runtime": 398.8881,
      "eval_samples_per_second": 4.347,
      "eval_steps_per_second": 0.183,
      "step": 50
    },
    {
      "epoch": 0.64,
      "grad_norm": 12.527059053107184,
      "learning_rate": 4.314988729807827e-07,
      "logits/chosen": -0.46812835335731506,
      "logits/rejected": -0.4670564532279968,
      "logps/chosen": -39.73173904418945,
      "logps/rejected": -31.038818359375,
      "loss": 0.277,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.1246474981307983,
      "rewards/margins": 1.668458342552185,
      "rewards/rejected": -0.5438107252120972,
      "step": 52
    },
    {
      "epoch": 0.6646153846153846,
      "grad_norm": 8.553726428740159,
      "learning_rate": 4.238809296526846e-07,
      "logits/chosen": -0.4466942548751831,
      "logits/rejected": -0.4432934820652008,
      "logps/chosen": -40.79918670654297,
      "logps/rejected": -43.690181732177734,
      "loss": 0.2475,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 1.295045256614685,
      "rewards/margins": 2.5195388793945312,
      "rewards/rejected": -1.2244938611984253,
      "step": 54
    },
    {
      "epoch": 0.676923076923077,
      "eval_logits/chosen": -0.4640098512172699,
      "eval_logits/rejected": -0.4600638151168823,
      "eval_logps/chosen": -42.25896453857422,
      "eval_logps/rejected": -45.97068786621094,
      "eval_loss": 0.26564928889274597,
      "eval_rewards/accuracies": 0.818493127822876,
      "eval_rewards/chosen": 1.47323477268219,
      "eval_rewards/margins": 3.019458293914795,
      "eval_rewards/rejected": -1.5462236404418945,
      "eval_runtime": 397.9539,
      "eval_samples_per_second": 4.357,
      "eval_steps_per_second": 0.183,
      "step": 55
    },
    {
      "epoch": 0.6892307692307692,
      "grad_norm": 10.956104922328452,
      "learning_rate": 4.159365430476261e-07,
      "logits/chosen": -0.45851805806159973,
      "logits/rejected": -0.4554572105407715,
      "logps/chosen": -39.361473083496094,
      "logps/rejected": -40.478187561035156,
      "loss": 0.2571,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.2780680656433105,
      "rewards/margins": 2.710892915725708,
      "rewards/rejected": -1.4328248500823975,
      "step": 56
    },
    {
      "epoch": 0.7138461538461538,
      "grad_norm": 8.998450490495424,
      "learning_rate": 4.076806279213655e-07,
      "logits/chosen": -0.44970858097076416,
      "logits/rejected": -0.4487365484237671,
      "logps/chosen": -39.6558723449707,
      "logps/rejected": -30.797224044799805,
      "loss": 0.2377,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 1.423360824584961,
      "rewards/margins": 2.117220163345337,
      "rewards/rejected": -0.6938591003417969,
      "step": 58
    },
    {
      "epoch": 0.7384615384615385,
      "grad_norm": 7.997921172873259,
      "learning_rate": 3.991286838919086e-07,
      "logits/chosen": -0.45197734236717224,
      "logits/rejected": -0.4494423270225525,
      "logps/chosen": -39.821556091308594,
      "logps/rejected": -30.172922134399414,
      "loss": 0.2308,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.3537344932556152,
      "rewards/margins": 2.035463571548462,
      "rewards/rejected": -0.6817290186882019,
      "step": 60
    },
    {
      "epoch": 0.7384615384615385,
      "eval_logits/chosen": -0.4599209725856781,
      "eval_logits/rejected": -0.45611828565597534,
      "eval_logps/chosen": -41.97780990600586,
      "eval_logps/rejected": -46.275150299072266,
      "eval_loss": 0.25006040930747986,
      "eval_rewards/accuracies": 0.8219178318977356,
      "eval_rewards/chosen": 1.6138089895248413,
      "eval_rewards/margins": 3.3122646808624268,
      "eval_rewards/rejected": -1.698456048965454,
      "eval_runtime": 400.364,
      "eval_samples_per_second": 4.331,
      "eval_steps_per_second": 0.182,
      "step": 60
    },
    {
      "epoch": 0.7630769230769231,
      "grad_norm": 8.82980837299946,
      "learning_rate": 3.902967663405956e-07,
      "logits/chosen": -0.4859851002693176,
      "logits/rejected": -0.48067888617515564,
      "logps/chosen": -41.49687194824219,
      "logps/rejected": -52.49277114868164,
      "loss": 0.2392,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.802024483680725,
      "rewards/margins": 3.978426456451416,
      "rewards/rejected": -2.1764023303985596,
      "step": 62
    },
    {
      "epoch": 0.7876923076923077,
      "grad_norm": 9.426495089553443,
      "learning_rate": 3.8120145626980015e-07,
      "logits/chosen": -0.4271140694618225,
      "logits/rejected": -0.42757901549339294,
      "logps/chosen": -42.58659744262695,
      "logps/rejected": -34.0533561706543,
      "loss": 0.2168,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 1.4099901914596558,
      "rewards/margins": 2.444286346435547,
      "rewards/rejected": -1.0342963933944702,
      "step": 64
    },
    {
      "epoch": 0.8,
      "eval_logits/chosen": -0.45980003476142883,
      "eval_logits/rejected": -0.4559204876422882,
      "eval_logps/chosen": -41.840850830078125,
      "eval_logps/rejected": -46.55259323120117,
      "eval_loss": 0.24060946702957153,
      "eval_rewards/accuracies": 0.8287671208381653,
      "eval_rewards/chosen": 1.6822888851165771,
      "eval_rewards/margins": 3.519465208053589,
      "eval_rewards/rejected": -1.8371765613555908,
      "eval_runtime": 400.9472,
      "eval_samples_per_second": 4.325,
      "eval_steps_per_second": 0.182,
      "step": 65
    },
    {
      "epoch": 0.8123076923076923,
      "grad_norm": 7.070039356412563,
      "learning_rate": 3.718598291738298e-07,
      "logits/chosen": -0.44447261095046997,
      "logits/rejected": -0.44296732544898987,
      "logps/chosen": -37.90412902832031,
      "logps/rejected": -36.56926727294922,
      "loss": 0.2129,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 1.5699365139007568,
      "rewards/margins": 2.4697201251983643,
      "rewards/rejected": -0.899783730506897,
      "step": 66
    },
    {
      "epoch": 0.8369230769230769,
      "grad_norm": 7.560682128141536,
      "learning_rate": 3.622894229814698e-07,
      "logits/chosen": -0.4488967955112457,
      "logits/rejected": -0.44460177421569824,
      "logps/chosen": -38.6946907043457,
      "logps/rejected": -46.51150894165039,
      "loss": 0.2148,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.4161229133605957,
      "rewards/margins": 3.171121835708618,
      "rewards/rejected": -1.754999041557312,
      "step": 68
    },
    {
      "epoch": 0.8615384615384616,
      "grad_norm": 8.217469156292612,
      "learning_rate": 3.52508205130354e-07,
      "logits/chosen": -0.45258235931396484,
      "logits/rejected": -0.44472819566726685,
      "logps/chosen": -41.79392623901367,
      "logps/rejected": -49.990150451660156,
      "loss": 0.2015,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.6353414058685303,
      "rewards/margins": 4.414577960968018,
      "rewards/rejected": -2.779236316680908,
      "step": 70
    },
    {
      "epoch": 0.8615384615384616,
      "eval_logits/chosen": -0.45598042011260986,
      "eval_logits/rejected": -0.4519526958465576,
      "eval_logps/chosen": -41.811824798583984,
      "eval_logps/rejected": -46.88448715209961,
      "eval_loss": 0.23113983869552612,
      "eval_rewards/accuracies": 0.835616409778595,
      "eval_rewards/chosen": 1.696802020072937,
      "eval_rewards/margins": 3.699925661087036,
      "eval_rewards/rejected": -2.0031237602233887,
      "eval_runtime": 397.814,
      "eval_samples_per_second": 4.359,
      "eval_steps_per_second": 0.184,
      "step": 70
    },
    {
      "epoch": 0.8861538461538462,
      "grad_norm": 9.912800042727127,
      "learning_rate": 3.4253453883497864e-07,
      "logits/chosen": -0.46372586488723755,
      "logits/rejected": -0.45979729294776917,
      "logps/chosen": -38.839900970458984,
      "logps/rejected": -39.225364685058594,
      "loss": 0.1988,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 1.1858875751495361,
      "rewards/margins": 3.046271562576294,
      "rewards/rejected": -1.8603839874267578,
      "step": 72
    },
    {
      "epoch": 0.9107692307692308,
      "grad_norm": 8.369939820458308,
      "learning_rate": 3.323871486116851e-07,
      "logits/chosen": -0.4712873697280884,
      "logits/rejected": -0.4673464894294739,
      "logps/chosen": -38.399444580078125,
      "logps/rejected": -31.575439453125,
      "loss": 0.2334,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 1.5223883390426636,
      "rewards/margins": 2.529682159423828,
      "rewards/rejected": -1.0072938203811646,
      "step": 74
    },
    {
      "epoch": 0.9230769230769231,
      "eval_logits/chosen": -0.4527950584888458,
      "eval_logits/rejected": -0.44881972670555115,
      "eval_logps/chosen": -41.899085998535156,
      "eval_logps/rejected": -47.179649353027344,
      "eval_loss": 0.2244771271944046,
      "eval_rewards/accuracies": 0.8424657583236694,
      "eval_rewards/chosen": 1.6531713008880615,
      "eval_rewards/margins": 3.803877830505371,
      "eval_rewards/rejected": -2.1507065296173096,
      "eval_runtime": 398.9461,
      "eval_samples_per_second": 4.346,
      "eval_steps_per_second": 0.183,
      "step": 75
    },
    {
      "epoch": 0.9353846153846154,
      "grad_norm": 9.041284945416068,
      "learning_rate": 3.220850851253377e-07,
      "logits/chosen": -0.48120999336242676,
      "logits/rejected": -0.476688027381897,
      "logps/chosen": -41.46022415161133,
      "logps/rejected": -46.435829162597656,
      "loss": 0.232,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.5242944955825806,
      "rewards/margins": 3.8246002197265625,
      "rewards/rejected": -2.3003056049346924,
      "step": 76
    },
    {
      "epoch": 0.96,
      "grad_norm": 5.93400233545759,
      "learning_rate": 3.1164768942369053e-07,
      "logits/chosen": -0.4919479787349701,
      "logits/rejected": -0.4866538941860199,
      "logps/chosen": -35.453460693359375,
      "logps/rejected": -63.363746643066406,
      "loss": 0.2211,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.4180715084075928,
      "rewards/margins": 4.756035804748535,
      "rewards/rejected": -3.3379647731781006,
      "step": 78
    },
    {
      "epoch": 0.9846153846153847,
      "grad_norm": 6.808355501392862,
      "learning_rate": 3.010945566265912e-07,
      "logits/chosen": -0.4714447855949402,
      "logits/rejected": -0.46636244654655457,
      "logps/chosen": -44.541507720947266,
      "logps/rejected": -33.366180419921875,
      "loss": 0.2075,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.5286719799041748,
      "rewards/margins": 2.761920928955078,
      "rewards/rejected": -1.233249306678772,
      "step": 80
    },
    {
      "epoch": 0.9846153846153847,
      "eval_logits/chosen": -0.45233336091041565,
      "eval_logits/rejected": -0.44827964901924133,
      "eval_logps/chosen": -41.91486740112305,
      "eval_logps/rejected": -47.406375885009766,
      "eval_loss": 0.2178829461336136,
      "eval_rewards/accuracies": 0.8458904027938843,
      "eval_rewards/chosen": 1.645279884338379,
      "eval_rewards/margins": 3.909348964691162,
      "eval_rewards/rejected": -2.2640695571899414,
      "eval_runtime": 398.4802,
      "eval_samples_per_second": 4.352,
      "eval_steps_per_second": 0.183,
      "step": 80
    },
    {
      "epoch": 1.0092307692307692,
      "grad_norm": 7.772486195689839,
      "learning_rate": 2.9044549913819124e-07,
      "logits/chosen": -0.4541618227958679,
      "logits/rejected": -0.4516042470932007,
      "logps/chosen": -42.46653747558594,
      "logps/rejected": -42.07926940917969,
      "loss": 0.1903,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 1.5551599264144897,
      "rewards/margins": 3.8793962001800537,
      "rewards/rejected": -2.3242363929748535,
      "step": 82
    },
    {
      "epoch": 1.0338461538461539,
      "grad_norm": 5.782441199522606,
      "learning_rate": 2.797205094512266e-07,
      "logits/chosen": -0.43424034118652344,
      "logits/rejected": -0.42985162138938904,
      "logps/chosen": -41.127593994140625,
      "logps/rejected": -45.52932357788086,
      "loss": 0.1885,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 1.399495005607605,
      "rewards/margins": 3.8588943481445312,
      "rewards/rejected": -2.4593992233276367,
      "step": 84
    },
    {
      "epoch": 1.0461538461538462,
      "eval_logits/chosen": -0.45614686608314514,
      "eval_logits/rejected": -0.45181941986083984,
      "eval_logps/chosen": -41.860015869140625,
      "eval_logps/rejected": -47.55991744995117,
      "eval_loss": 0.21262364089488983,
      "eval_rewards/accuracies": 0.8493150472640991,
      "eval_rewards/chosen": 1.6727051734924316,
      "eval_rewards/margins": 4.013547897338867,
      "eval_rewards/rejected": -2.3408427238464355,
      "eval_runtime": 401.4695,
      "eval_samples_per_second": 4.319,
      "eval_steps_per_second": 0.182,
      "step": 85
    },
    {
      "epoch": 1.0584615384615386,
      "grad_norm": 4.97735971019092,
      "learning_rate": 2.6893972261320264e-07,
      "logits/chosen": -0.42422690987586975,
      "logits/rejected": -0.4213142693042755,
      "logps/chosen": -39.9366340637207,
      "logps/rejected": -40.46057891845703,
      "loss": 0.1749,
      "rewards/accuracies": 0.953125,
      "rewards/chosen": 1.7523412704467773,
      "rewards/margins": 3.447618007659912,
      "rewards/rejected": -1.6952769756317139,
      "step": 86
    },
    {
      "epoch": 1.083076923076923,
      "grad_norm": 5.4006356704238465,
      "learning_rate": 2.5812337842494516e-07,
      "logits/chosen": -0.4536130726337433,
      "logits/rejected": -0.45070070028305054,
      "logps/chosen": -41.13835906982422,
      "logps/rejected": -36.39931106567383,
      "loss": 0.1732,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.473576545715332,
      "rewards/margins": 3.01824688911438,
      "rewards/rejected": -1.5446703433990479,
      "step": 88
    },
    {
      "epoch": 1.1076923076923078,
      "grad_norm": 5.007052371554692,
      "learning_rate": 2.4729178344249006e-07,
      "logits/chosen": -0.45735257863998413,
      "logits/rejected": -0.45181047916412354,
      "logps/chosen": -39.739837646484375,
      "logps/rejected": -40.88286590576172,
      "loss": 0.1838,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 1.5848816633224487,
      "rewards/margins": 3.2226974964141846,
      "rewards/rejected": -1.6378157138824463,
      "step": 90
    },
    {
      "epoch": 1.1076923076923078,
      "eval_logits/chosen": -0.44746431708335876,
      "eval_logits/rejected": -0.44344425201416016,
      "eval_logps/chosen": -41.768062591552734,
      "eval_logps/rejected": -47.707096099853516,
      "eval_loss": 0.20802097022533417,
      "eval_rewards/accuracies": 0.8527397513389587,
      "eval_rewards/chosen": 1.7186800241470337,
      "eval_rewards/margins": 4.133108615875244,
      "eval_rewards/rejected": -2.414428472518921,
      "eval_runtime": 398.368,
      "eval_samples_per_second": 4.353,
      "eval_steps_per_second": 0.183,
      "step": 90
    },
    {
      "epoch": 1.1323076923076922,
      "grad_norm": 5.422550680740592,
      "learning_rate": 2.3646527285364563e-07,
      "logits/chosen": -0.45245465636253357,
      "logits/rejected": -0.4461210072040558,
      "logps/chosen": -40.16954040527344,
      "logps/rejected": -40.06535339355469,
      "loss": 0.1908,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.9867323637008667,
      "rewards/margins": 3.7591819763183594,
      "rewards/rejected": -1.7724494934082031,
      "step": 92
    },
    {
      "epoch": 1.156923076923077,
      "grad_norm": 5.242542300523203,
      "learning_rate": 2.256641723008026e-07,
      "logits/chosen": -0.4320615530014038,
      "logits/rejected": -0.4263082444667816,
      "logps/chosen": -37.380733489990234,
      "logps/rejected": -34.89535903930664,
      "loss": 0.2062,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 1.6694892644882202,
      "rewards/margins": 3.161728858947754,
      "rewards/rejected": -1.4922394752502441,
      "step": 94
    },
    {
      "epoch": 1.1692307692307693,
      "eval_logits/chosen": -0.4461810886859894,
      "eval_logits/rejected": -0.44209808111190796,
      "eval_logps/chosen": -41.719635009765625,
      "eval_logps/rejected": -47.83554458618164,
      "eval_loss": 0.20408523082733154,
      "eval_rewards/accuracies": 0.8595890402793884,
      "eval_rewards/chosen": 1.7428948879241943,
      "eval_rewards/margins": 4.221549987792969,
      "eval_rewards/rejected": -2.4786548614501953,
      "eval_runtime": 399.3763,
      "eval_samples_per_second": 4.342,
      "eval_steps_per_second": 0.183,
      "step": 95
    },
    {
      "epoch": 1.1815384615384614,
      "grad_norm": 5.9451928875167805,
      "learning_rate": 2.1490875972166393e-07,
      "logits/chosen": -0.471175879240036,
      "logits/rejected": -0.46583351492881775,
      "logps/chosen": -34.772342681884766,
      "logps/rejected": -39.63105392456055,
      "loss": 0.1776,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.5616707801818848,
      "rewards/margins": 3.4220144748687744,
      "rewards/rejected": -1.8603436946868896,
      "step": 96
    },
    {
      "epoch": 1.2061538461538461,
      "grad_norm": 5.4858414240825075,
      "learning_rate": 2.0421922727953595e-07,
      "logits/chosen": -0.4286724925041199,
      "logits/rejected": -0.4263175427913666,
      "logps/chosen": -36.57586669921875,
      "logps/rejected": -31.930763244628906,
      "loss": 0.1655,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.753572940826416,
      "rewards/margins": 2.9586939811706543,
      "rewards/rejected": -1.2051211595535278,
      "step": 98
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 5.720592020809117,
      "learning_rate": 1.9361564345465145e-07,
      "logits/chosen": -0.4527323842048645,
      "logits/rejected": -0.44913169741630554,
      "logps/chosen": -34.954219818115234,
      "logps/rejected": -37.62247085571289,
      "loss": 0.1794,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.6976581811904907,
      "rewards/margins": 3.37191104888916,
      "rewards/rejected": -1.674253225326538,
      "step": 100
    },
    {
      "epoch": 1.2307692307692308,
      "eval_logits/chosen": -0.446783185005188,
      "eval_logits/rejected": -0.44245585799217224,
      "eval_logps/chosen": -41.71480178833008,
      "eval_logps/rejected": -47.994571685791016,
      "eval_loss": 0.20066353678703308,
      "eval_rewards/accuracies": 0.8561643958091736,
      "eval_rewards/chosen": 1.745313048362732,
      "eval_rewards/margins": 4.30348014831543,
      "eval_rewards/rejected": -2.558166742324829,
      "eval_runtime": 400.8371,
      "eval_samples_per_second": 4.326,
      "eval_steps_per_second": 0.182,
      "step": 100
    },
    {
      "epoch": 1.2553846153846153,
      "grad_norm": 5.495318529151337,
      "learning_rate": 1.8311791536769483e-07,
      "logits/chosen": -0.4154907763004303,
      "logits/rejected": -0.41147375106811523,
      "logps/chosen": -37.79768753051758,
      "logps/rejected": -48.32915496826172,
      "loss": 0.1734,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.4379969835281372,
      "rewards/margins": 4.379321098327637,
      "rewards/rejected": -2.94132399559021,
      "step": 102
    },
    {
      "epoch": 1.28,
      "grad_norm": 5.371136709550833,
      "learning_rate": 1.7274575140626315e-07,
      "logits/chosen": -0.44362062215805054,
      "logits/rejected": -0.4401937425136566,
      "logps/chosen": -38.68498992919922,
      "logps/rejected": -41.304710388183594,
      "loss": 0.1857,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.9046180248260498,
      "rewards/margins": 3.721491813659668,
      "rewards/rejected": -1.8168736696243286,
      "step": 104
    },
    {
      "epoch": 1.2923076923076924,
      "eval_logits/chosen": -0.44231978058815,
      "eval_logits/rejected": -0.43791159987449646,
      "eval_logps/chosen": -41.76345443725586,
      "eval_logps/rejected": -48.174468994140625,
      "eval_loss": 0.1971779763698578,
      "eval_rewards/accuracies": 0.8595890402793884,
      "eval_rewards/chosen": 1.7209877967834473,
      "eval_rewards/margins": 4.369105339050293,
      "eval_rewards/rejected": -2.6481170654296875,
      "eval_runtime": 400.535,
      "eval_samples_per_second": 4.329,
      "eval_steps_per_second": 0.182,
      "step": 105
    },
    {
      "epoch": 1.3046153846153845,
      "grad_norm": 5.888788259237849,
      "learning_rate": 1.6251862422442788e-07,
      "logits/chosen": -0.4513101577758789,
      "logits/rejected": -0.4436649978160858,
      "logps/chosen": -40.830169677734375,
      "logps/rejected": -56.550262451171875,
      "loss": 0.1686,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 1.6578915119171143,
      "rewards/margins": 4.946527481079102,
      "rewards/rejected": -3.2886364459991455,
      "step": 106
    },
    {
      "epoch": 1.3292307692307692,
      "grad_norm": 6.612486006770989,
      "learning_rate": 1.5245573418486135e-07,
      "logits/chosen": -0.44994574785232544,
      "logits/rejected": -0.4477950930595398,
      "logps/chosen": -38.80488586425781,
      "logps/rejected": -41.393211364746094,
      "loss": 0.1875,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.4876614809036255,
      "rewards/margins": 3.6893627643585205,
      "rewards/rejected": -2.2017014026641846,
      "step": 108
    },
    {
      "epoch": 1.353846153846154,
      "grad_norm": 4.656660129291058,
      "learning_rate": 1.4257597331216208e-07,
      "logits/chosen": -0.4406875669956207,
      "logits/rejected": -0.43555599451065063,
      "logps/chosen": -42.75905990600586,
      "logps/rejected": -54.37929153442383,
      "loss": 0.1647,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.7431519031524658,
      "rewards/margins": 5.2682318687438965,
      "rewards/rejected": -3.5250799655914307,
      "step": 110
    },
    {
      "epoch": 1.353846153846154,
      "eval_logits/chosen": -0.44183987379074097,
      "eval_logits/rejected": -0.4375387728214264,
      "eval_logps/chosen": -41.78536605834961,
      "eval_logps/rejected": -48.30013656616211,
      "eval_loss": 0.19461464881896973,
      "eval_rewards/accuracies": 0.8595890402793884,
      "eval_rewards/chosen": 1.710031509399414,
      "eval_rewards/margins": 4.420979976654053,
      "eval_rewards/rejected": -2.7109484672546387,
      "eval_runtime": 398.8533,
      "eval_samples_per_second": 4.347,
      "eval_steps_per_second": 0.183,
      "step": 110
    },
    {
      "epoch": 1.3784615384615384,
      "grad_norm": 4.429475200922326,
      "learning_rate": 1.328978898250525e-07,
      "logits/chosen": -0.45591893792152405,
      "logits/rejected": -0.45284152030944824,
      "logps/chosen": -37.928653717041016,
      "logps/rejected": -38.99214172363281,
      "loss": 0.1596,
      "rewards/accuracies": 0.953125,
      "rewards/chosen": 1.702134609222412,
      "rewards/margins": 3.827345371246338,
      "rewards/rejected": -2.125210762023926,
      "step": 112
    },
    {
      "epoch": 1.403076923076923,
      "grad_norm": 4.278591337865169,
      "learning_rate": 1.234396533140365e-07,
      "logits/chosen": -0.42062410712242126,
      "logits/rejected": -0.41781002283096313,
      "logps/chosen": -41.77610778808594,
      "logps/rejected": -48.5487060546875,
      "loss": 0.1524,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 1.599095106124878,
      "rewards/margins": 4.177494049072266,
      "rewards/rejected": -2.578399181365967,
      "step": 114
    },
    {
      "epoch": 1.4153846153846155,
      "eval_logits/chosen": -0.44146084785461426,
      "eval_logits/rejected": -0.437191367149353,
      "eval_logps/chosen": -41.83814239501953,
      "eval_logps/rejected": -48.426334381103516,
      "eval_loss": 0.1936241090297699,
      "eval_rewards/accuracies": 0.8561643958091736,
      "eval_rewards/chosen": 1.6836414337158203,
      "eval_rewards/margins": 4.457688331604004,
      "eval_rewards/rejected": -2.7740468978881836,
      "eval_runtime": 398.1677,
      "eval_samples_per_second": 4.355,
      "eval_steps_per_second": 0.183,
      "step": 115
    },
    {
      "epoch": 1.4276923076923076,
      "grad_norm": 4.8647599997437805,
      "learning_rate": 1.1421902062989178e-07,
      "logits/chosen": -0.44749611616134644,
      "logits/rejected": -0.4411754608154297,
      "logps/chosen": -43.03426742553711,
      "logps/rejected": -60.566200256347656,
      "loss": 0.1551,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.691102147102356,
      "rewards/margins": 5.743535041809082,
      "rewards/rejected": -4.052433013916016,
      "step": 116
    },
    {
      "epoch": 1.4523076923076923,
      "grad_norm": 4.32184296596162,
      "learning_rate": 1.0525330254703788e-07,
      "logits/chosen": -0.4085264205932617,
      "logits/rejected": -0.4047048091888428,
      "logps/chosen": -40.241764068603516,
      "logps/rejected": -48.12974166870117,
      "loss": 0.1681,
      "rewards/accuracies": 0.796875,
      "rewards/chosen": 1.730395793914795,
      "rewards/margins": 4.292797088623047,
      "rewards/rejected": -2.5624003410339355,
      "step": 118
    },
    {
      "epoch": 1.476923076923077,
      "grad_norm": 6.544542585214649,
      "learning_rate": 9.655933126436563e-08,
      "logits/chosen": -0.4505607783794403,
      "logits/rejected": -0.44797009229660034,
      "logps/chosen": -42.54741668701172,
      "logps/rejected": -34.47932052612305,
      "loss": 0.1668,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.5434067249298096,
      "rewards/margins": 3.265448570251465,
      "rewards/rejected": -1.7220419645309448,
      "step": 120
    },
    {
      "epoch": 1.476923076923077,
      "eval_logits/chosen": -0.4385281801223755,
      "eval_logits/rejected": -0.43417835235595703,
      "eval_logps/chosen": -41.83414077758789,
      "eval_logps/rejected": -48.51993179321289,
      "eval_loss": 0.19258952140808105,
      "eval_rewards/accuracies": 0.8527397513389587,
      "eval_rewards/chosen": 1.6856427192687988,
      "eval_rewards/margins": 4.5064897537231445,
      "eval_rewards/rejected": -2.8208470344543457,
      "eval_runtime": 398.4093,
      "eval_samples_per_second": 4.352,
      "eval_steps_per_second": 0.183,
      "step": 120
    },
    {
      "epoch": 1.5015384615384615,
      "grad_norm": 3.8817405874572923,
      "learning_rate": 8.81534288045431e-08,
      "logits/chosen": -0.4345233738422394,
      "logits/rejected": -0.432235985994339,
      "logps/chosen": -38.9439697265625,
      "logps/rejected": -40.60162353515625,
      "loss": 0.155,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 1.4539457559585571,
      "rewards/margins": 3.9496073722839355,
      "rewards/rejected": -2.4956612586975098,
      "step": 122
    },
    {
      "epoch": 1.5261538461538462,
      "grad_norm": 4.9773505616027425,
      "learning_rate": 8.005137637112302e-08,
      "logits/chosen": -0.42935946583747864,
      "logits/rejected": -0.4225008487701416,
      "logps/chosen": -40.33964157104492,
      "logps/rejected": -37.91059875488281,
      "loss": 0.1685,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.3645392656326294,
      "rewards/margins": 3.7918734550476074,
      "rewards/rejected": -2.4273343086242676,
      "step": 124
    },
    {
      "epoch": 1.5384615384615383,
      "eval_logits/chosen": -0.4404720067977905,
      "eval_logits/rejected": -0.4361659288406372,
      "eval_logps/chosen": -41.75288391113281,
      "eval_logps/rejected": -48.5331916809082,
      "eval_loss": 0.19096876680850983,
      "eval_rewards/accuracies": 0.8561643958091736,
      "eval_rewards/chosen": 1.7262725830078125,
      "eval_rewards/margins": 4.553750991821289,
      "eval_rewards/rejected": -2.8274781703948975,
      "eval_runtime": 397.8296,
      "eval_samples_per_second": 4.359,
      "eval_steps_per_second": 0.183,
      "step": 125
    },
    {
      "epoch": 1.5507692307692307,
      "grad_norm": 3.816285562997914,
      "learning_rate": 7.226838472098237e-08,
      "logits/chosen": -0.4305260181427002,
      "logits/rejected": -0.427184522151947,
      "logps/chosen": -38.50707244873047,
      "logps/rejected": -31.813642501831055,
      "loss": 0.1785,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.653688669204712,
      "rewards/margins": 3.0635831356048584,
      "rewards/rejected": -1.4098942279815674,
      "step": 126
    },
    {
      "epoch": 1.5753846153846154,
      "grad_norm": 5.070564484576485,
      "learning_rate": 6.481906560771524e-08,
      "logits/chosen": -0.45029404759407043,
      "logits/rejected": -0.44603830575942993,
      "logps/chosen": -40.4008674621582,
      "logps/rejected": -34.62392044067383,
      "loss": 0.162,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 1.847829818725586,
      "rewards/margins": 3.5869626998901367,
      "rewards/rejected": -1.7391327619552612,
      "step": 128
    },
    {
      "epoch": 1.6,
      "grad_norm": 4.291727231427144,
      "learning_rate": 5.771740434959277e-08,
      "logits/chosen": -0.43359375,
      "logits/rejected": -0.4286433160305023,
      "logps/chosen": -38.0141716003418,
      "logps/rejected": -41.48249053955078,
      "loss": 0.1637,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 1.5690267086029053,
      "rewards/margins": 3.8614985942840576,
      "rewards/rejected": -2.2924716472625732,
      "step": 130
    },
    {
      "epoch": 1.6,
      "eval_logits/chosen": -0.43712207674980164,
      "eval_logits/rejected": -0.43287885189056396,
      "eval_logps/chosen": -41.706233978271484,
      "eval_logps/rejected": -48.57079315185547,
      "eval_loss": 0.18982651829719543,
      "eval_rewards/accuracies": 0.8595890402793884,
      "eval_rewards/chosen": 1.7495962381362915,
      "eval_rewards/margins": 4.5958757400512695,
      "eval_rewards/rejected": -2.8462791442871094,
      "eval_runtime": 399.3255,
      "eval_samples_per_second": 4.342,
      "eval_steps_per_second": 0.183,
      "step": 130
    }
  ],
  "logging_steps": 2,
  "max_steps": 162,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}