{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 50,
  "global_step": 436,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022935779816513763,
      "grad_norm": 1.6432102724209243,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.619717597961426,
      "logits/rejected": -2.552694320678711,
      "logps/chosen": -265.4180908203125,
      "logps/rejected": -236.1053924560547,
      "loss": 0.0154,
      "rewards/accuracies": 0.3812499940395355,
      "rewards/chosen": 0.00027717588818632066,
      "rewards/margins": 3.1426665373146534e-05,
      "rewards/rejected": 0.0002457492519170046,
      "step": 10
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 1.5125635433247804,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.6578421592712402,
      "logits/rejected": -2.5760762691497803,
      "logps/chosen": -298.83837890625,
      "logps/rejected": -274.30596923828125,
      "loss": 0.0155,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": -0.0005307936808094382,
      "rewards/margins": 0.0014294281136244535,
      "rewards/rejected": -0.0019602221436798573,
      "step": 20
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 1.3722478894206875,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6761069297790527,
      "logits/rejected": -2.602224826812744,
      "logps/chosen": -290.43157958984375,
      "logps/rejected": -234.3854522705078,
      "loss": 0.0146,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.00901214499026537,
      "rewards/margins": 0.013047738000750542,
      "rewards/rejected": -0.004035593010485172,
      "step": 30
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 1.2692399849923888,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.6601767539978027,
      "logits/rejected": -2.610305070877075,
      "logps/chosen": -280.9813537597656,
      "logps/rejected": -267.60333251953125,
      "loss": 0.013,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.04187846928834915,
      "rewards/margins": 0.0399814248085022,
      "rewards/rejected": 0.0018970475066453218,
      "step": 40
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 1.4063277329047579,
      "learning_rate": 4.997110275491701e-07,
      "logits/chosen": -2.6256864070892334,
      "logits/rejected": -2.6204254627227783,
      "logps/chosen": -289.7392578125,
      "logps/rejected": -299.10064697265625,
      "loss": 0.012,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.06511950492858887,
      "rewards/margins": 0.06428412348031998,
      "rewards/rejected": 0.0008353829616680741,
      "step": 50
    },
    {
      "epoch": 0.11467889908256881,
      "eval_logits/chosen": -2.578613519668579,
      "eval_logits/rejected": -2.502546787261963,
      "eval_logps/chosen": -277.33428955078125,
      "eval_logps/rejected": -247.25254821777344,
      "eval_loss": 0.01130068488419056,
      "eval_rewards/accuracies": 0.6982758641242981,
      "eval_rewards/chosen": 0.0775580108165741,
      "eval_rewards/margins": 0.0848592072725296,
      "eval_rewards/rejected": -0.007301207631826401,
      "eval_runtime": 93.3854,
      "eval_samples_per_second": 19.468,
      "eval_steps_per_second": 0.311,
      "step": 50
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 1.422544468484653,
      "learning_rate": 4.979475034558115e-07,
      "logits/chosen": -2.580706834793091,
      "logits/rejected": -2.525763750076294,
      "logps/chosen": -284.323486328125,
      "logps/rejected": -259.6044006347656,
      "loss": 0.0112,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.05161212012171745,
      "rewards/margins": 0.07209066301584244,
      "rewards/rejected": -0.020478537306189537,
      "step": 60
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 1.4554279086187263,
      "learning_rate": 4.945923025551788e-07,
      "logits/chosen": -2.5086066722869873,
      "logits/rejected": -2.471884250640869,
      "logps/chosen": -319.5674743652344,
      "logps/rejected": -265.2959899902344,
      "loss": 0.0115,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.08003517985343933,
      "rewards/margins": 0.11891887336969376,
      "rewards/rejected": -0.038883693516254425,
      "step": 70
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 1.3828352586845378,
      "learning_rate": 4.896669632591651e-07,
      "logits/chosen": -2.480229616165161,
      "logits/rejected": -2.3888580799102783,
      "logps/chosen": -279.42529296875,
      "logps/rejected": -256.35528564453125,
      "loss": 0.0109,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.06424586474895477,
      "rewards/margins": 0.10123773664236069,
      "rewards/rejected": -0.036991871893405914,
      "step": 80
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 1.341734639019618,
      "learning_rate": 4.832031033425662e-07,
      "logits/chosen": -2.430302381515503,
      "logits/rejected": -2.3838858604431152,
      "logps/chosen": -280.6355895996094,
      "logps/rejected": -252.35107421875,
      "loss": 0.0103,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.06957819312810898,
      "rewards/margins": 0.12507636845111847,
      "rewards/rejected": -0.0554981604218483,
      "step": 90
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 1.503281183515388,
      "learning_rate": 4.752422169756047e-07,
      "logits/chosen": -2.392822265625,
      "logits/rejected": -2.337982654571533,
      "logps/chosen": -269.06829833984375,
      "logps/rejected": -280.2345886230469,
      "loss": 0.0111,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.033564209938049316,
      "rewards/margins": 0.11662591993808746,
      "rewards/rejected": -0.08306171000003815,
      "step": 100
    },
    {
      "epoch": 0.22935779816513763,
      "eval_logits/chosen": -2.445643424987793,
      "eval_logits/rejected": -2.345205545425415,
      "eval_logps/chosen": -281.08892822265625,
      "eval_logps/rejected": -254.68817138671875,
      "eval_loss": 0.01000975538045168,
      "eval_rewards/accuracies": 0.7112069129943848,
      "eval_rewards/chosen": 0.04001150280237198,
      "eval_rewards/margins": 0.1216687560081482,
      "eval_rewards/rejected": -0.08165725320577621,
      "eval_runtime": 92.3022,
      "eval_samples_per_second": 19.696,
      "eval_steps_per_second": 0.314,
      "step": 100
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 2.297190949384611,
      "learning_rate": 4.658354083558188e-07,
      "logits/chosen": -2.466749668121338,
      "logits/rejected": -2.361680030822754,
      "logps/chosen": -263.682373046875,
      "logps/rejected": -241.5611572265625,
      "loss": 0.0113,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.015057370066642761,
      "rewards/margins": 0.09333699941635132,
      "rewards/rejected": -0.07827961444854736,
      "step": 110
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 1.422856756971901,
      "learning_rate": 4.550430636492389e-07,
      "logits/chosen": -2.3867597579956055,
      "logits/rejected": -2.3550314903259277,
      "logps/chosen": -275.4966125488281,
      "logps/rejected": -260.20843505859375,
      "loss": 0.0103,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0032271749805659056,
      "rewards/margins": 0.09067237377166748,
      "rewards/rejected": -0.08744519203901291,
      "step": 120
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 2.508161764096464,
      "learning_rate": 4.429344633468004e-07,
      "logits/chosen": -2.4056849479675293,
      "logits/rejected": -2.361074686050415,
      "logps/chosen": -254.7312469482422,
      "logps/rejected": -252.63455200195312,
      "loss": 0.0105,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.014488674700260162,
      "rewards/margins": 0.11470804363489151,
      "rewards/rejected": -0.10021935403347015,
      "step": 130
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 3.762306745768528,
      "learning_rate": 4.2958733752443187e-07,
      "logits/chosen": -2.412415027618408,
      "logits/rejected": -2.3416905403137207,
      "logps/chosen": -273.1282043457031,
      "logps/rejected": -231.5812225341797,
      "loss": 0.0122,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.03056950867176056,
      "rewards/margins": 0.10341660678386688,
      "rewards/rejected": -0.07284711301326752,
      "step": 140
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 1.323990928834917,
      "learning_rate": 4.150873668617898e-07,
      "logits/chosen": -2.454942226409912,
      "logits/rejected": -2.3723654747009277,
      "logps/chosen": -268.99053955078125,
      "logps/rejected": -247.30032348632812,
      "loss": 0.0104,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.02446950413286686,
      "rewards/margins": 0.11665806919336319,
      "rewards/rejected": -0.1411275714635849,
      "step": 150
    },
    {
      "epoch": 0.3440366972477064,
      "eval_logits/chosen": -2.4970576763153076,
      "eval_logits/rejected": -2.400646924972534,
      "eval_logps/chosen": -286.011474609375,
      "eval_logps/rejected": -260.7337646484375,
      "eval_loss": 0.009798681363463402,
      "eval_rewards/accuracies": 0.7284482717514038,
      "eval_rewards/chosen": -0.009214152581989765,
      "eval_rewards/margins": 0.13289883732795715,
      "eval_rewards/rejected": -0.14211300015449524,
      "eval_runtime": 93.2661,
      "eval_samples_per_second": 19.493,
      "eval_steps_per_second": 0.311,
      "step": 150
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 2.7039803270775495,
      "learning_rate": 3.9952763262280397e-07,
      "logits/chosen": -2.460155963897705,
      "logits/rejected": -2.39618182182312,
      "logps/chosen": -293.5860595703125,
      "logps/rejected": -295.9964294433594,
      "loss": 0.01,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.02965056337416172,
      "rewards/margins": 0.1299286037683487,
      "rewards/rejected": -0.10027804225683212,
      "step": 160
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 1.3210709780236793,
      "learning_rate": 3.8300801912883414e-07,
      "logits/chosen": -2.503695249557495,
      "logits/rejected": -2.4564812183380127,
      "logps/chosen": -267.4649353027344,
      "logps/rejected": -279.4906921386719,
      "loss": 0.0091,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.041359588503837585,
      "rewards/margins": 0.13557755947113037,
      "rewards/rejected": -0.09421796351671219,
      "step": 170
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 1.4526636355174283,
      "learning_rate": 3.6563457256020884e-07,
      "logits/chosen": -2.5071005821228027,
      "logits/rejected": -2.4085752964019775,
      "logps/chosen": -306.22930908203125,
      "logps/rejected": -253.84475708007812,
      "loss": 0.0096,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.028713583946228027,
      "rewards/margins": 0.12478481233119965,
      "rewards/rejected": -0.15349839627742767,
      "step": 180
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 4.8242221106390115,
      "learning_rate": 3.475188202022617e-07,
      "logits/chosen": -2.414685010910034,
      "logits/rejected": -2.3840391635894775,
      "logps/chosen": -256.085205078125,
      "logps/rejected": -276.09039306640625,
      "loss": 0.0106,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.020293032750487328,
      "rewards/margins": 0.13775238394737244,
      "rewards/rejected": -0.11745934188365936,
      "step": 190
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 1.2637397948323732,
      "learning_rate": 3.287770545059052e-07,
      "logits/chosen": -2.5293562412261963,
      "logits/rejected": -2.431912899017334,
      "logps/chosen": -277.83734130859375,
      "logps/rejected": -256.05206298828125,
      "loss": 0.0096,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.004443052224814892,
      "rewards/margins": 0.12640419602394104,
      "rewards/rejected": -0.12196113169193268,
      "step": 200
    },
    {
      "epoch": 0.45871559633027525,
      "eval_logits/chosen": -2.5115060806274414,
      "eval_logits/rejected": -2.4205849170684814,
      "eval_logps/chosen": -282.7939147949219,
      "eval_logps/rejected": -258.3851013183594,
      "eval_loss": 0.00931267999112606,
      "eval_rewards/accuracies": 0.7887930870056152,
      "eval_rewards/chosen": 0.022961637005209923,
      "eval_rewards/margins": 0.14158831536769867,
      "eval_rewards/rejected": -0.1186266764998436,
      "eval_runtime": 92.3817,
      "eval_samples_per_second": 19.679,
      "eval_steps_per_second": 0.314,
      "step": 200
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 1.093362739669567,
      "learning_rate": 3.0952958655864954e-07,
      "logits/chosen": -2.4653537273406982,
      "logits/rejected": -2.448239803314209,
      "logps/chosen": -271.6222229003906,
      "logps/rejected": -265.7300109863281,
      "loss": 0.0094,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.01682201772928238,
      "rewards/margins": 0.13070014119148254,
      "rewards/rejected": -0.11387811601161957,
      "step": 210
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 1.5040749987763142,
      "learning_rate": 2.898999737583448e-07,
      "logits/chosen": -2.50272798538208,
      "logits/rejected": -2.4043869972229004,
      "logps/chosen": -322.4222717285156,
      "logps/rejected": -300.2145690917969,
      "loss": 0.0091,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.0025803411845117807,
      "rewards/margins": 0.15502406656742096,
      "rewards/rejected": -0.15244372189044952,
      "step": 220
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 1.2805773201998596,
      "learning_rate": 2.7001422664752333e-07,
      "logits/chosen": -2.396461009979248,
      "logits/rejected": -2.344364881515503,
      "logps/chosen": -268.88946533203125,
      "logps/rejected": -278.57122802734375,
      "loss": 0.0093,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.009459306485950947,
      "rewards/margins": 0.12005350738763809,
      "rewards/rejected": -0.12951281666755676,
      "step": 230
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 1.346452258119606,
      "learning_rate": 2.5e-07,
      "logits/chosen": -2.471860885620117,
      "logits/rejected": -2.379460573196411,
      "logps/chosen": -288.55877685546875,
      "logps/rejected": -274.3955078125,
      "loss": 0.0096,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.010269422084093094,
      "rewards/margins": 0.12018589675426483,
      "rewards/rejected": -0.13045531511306763,
      "step": 240
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 1.1636311800865655,
      "learning_rate": 2.2998577335247667e-07,
      "logits/chosen": -2.5101351737976074,
      "logits/rejected": -2.4231343269348145,
      "logps/chosen": -303.5893249511719,
      "logps/rejected": -274.129638671875,
      "loss": 0.0093,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 0.0020445596892386675,
      "rewards/margins": 0.13890263438224792,
      "rewards/rejected": -0.13685807585716248,
      "step": 250
    },
    {
      "epoch": 0.573394495412844,
      "eval_logits/chosen": -2.459127187728882,
      "eval_logits/rejected": -2.365269899368286,
      "eval_logps/chosen": -286.2547912597656,
      "eval_logps/rejected": -263.3385925292969,
      "eval_loss": 0.008932252414524555,
      "eval_rewards/accuracies": 0.7844827771186829,
      "eval_rewards/chosen": -0.011646694503724575,
      "eval_rewards/margins": 0.1565149575471878,
      "eval_rewards/rejected": -0.16816167533397675,
      "eval_runtime": 93.0437,
      "eval_samples_per_second": 19.539,
      "eval_steps_per_second": 0.312,
      "step": 250
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 1.1997721634582499,
      "learning_rate": 2.1010002624165524e-07,
      "logits/chosen": -2.4458436965942383,
      "logits/rejected": -2.4075264930725098,
      "logps/chosen": -273.95843505859375,
      "logps/rejected": -292.99114990234375,
      "loss": 0.0097,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.032438840717077255,
      "rewards/margins": 0.1535351574420929,
      "rewards/rejected": -0.18597400188446045,
      "step": 260
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 1.2414114105136553,
      "learning_rate": 1.9047041344135043e-07,
      "logits/chosen": -2.437638759613037,
      "logits/rejected": -2.409536600112915,
      "logps/chosen": -270.0658264160156,
      "logps/rejected": -270.24798583984375,
      "loss": 0.0092,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.04003521427512169,
      "rewards/margins": 0.14397411048412323,
      "rewards/rejected": -0.1840093433856964,
      "step": 270
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 1.288255010907884,
      "learning_rate": 1.7122294549409482e-07,
      "logits/chosen": -2.5124118328094482,
      "logits/rejected": -2.4503026008605957,
      "logps/chosen": -273.12841796875,
      "logps/rejected": -280.5234375,
      "loss": 0.0098,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.0012958184815943241,
      "rewards/margins": 0.165980726480484,
      "rewards/rejected": -0.167276531457901,
      "step": 280
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 1.3451525886655082,
      "learning_rate": 1.524811797977383e-07,
      "logits/chosen": -2.494507312774658,
      "logits/rejected": -2.4191977977752686,
      "logps/chosen": -287.3388366699219,
      "logps/rejected": -267.7834167480469,
      "loss": 0.009,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 0.003920386545360088,
      "rewards/margins": 0.1425868272781372,
      "rewards/rejected": -0.13866642117500305,
      "step": 290
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 1.3558837079563066,
      "learning_rate": 1.3436542743979125e-07,
      "logits/chosen": -2.498835563659668,
      "logits/rejected": -2.4762415885925293,
      "logps/chosen": -306.97705078125,
      "logps/rejected": -270.83367919921875,
      "loss": 0.0096,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.010277172550559044,
      "rewards/margins": 0.10946089029312134,
      "rewards/rejected": -0.11973806470632553,
      "step": 300
    },
    {
      "epoch": 0.6880733944954128,
      "eval_logits/chosen": -2.4814062118530273,
      "eval_logits/rejected": -2.38907790184021,
      "eval_logps/chosen": -285.9173278808594,
      "eval_logps/rejected": -262.4115295410156,
      "eval_loss": 0.008779831230640411,
      "eval_rewards/accuracies": 0.7844827771186829,
      "eval_rewards/chosen": -0.008272184059023857,
      "eval_rewards/margins": 0.15061868727207184,
      "eval_rewards/rejected": -0.15889087319374084,
      "eval_runtime": 92.1897,
      "eval_samples_per_second": 19.72,
      "eval_steps_per_second": 0.315,
      "step": 300
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 1.3381129170655774,
      "learning_rate": 1.1699198087116588e-07,
      "logits/chosen": -2.515507221221924,
      "logits/rejected": -2.4294919967651367,
      "logps/chosen": -280.1512756347656,
      "logps/rejected": -280.0994873046875,
      "loss": 0.0094,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.02186095342040062,
      "rewards/margins": 0.11463706195354462,
      "rewards/rejected": -0.13649800419807434,
      "step": 310
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 1.338991121862378,
      "learning_rate": 1.00472367377196e-07,
      "logits/chosen": -2.451615571975708,
      "logits/rejected": -2.404048442840576,
      "logps/chosen": -275.2835388183594,
      "logps/rejected": -251.0447998046875,
      "loss": 0.0093,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.010515814647078514,
      "rewards/margins": 0.15502998232841492,
      "rewards/rejected": -0.16554580628871918,
      "step": 320
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 1.3896883263831044,
      "learning_rate": 8.49126331382102e-08,
      "logits/chosen": -2.4412286281585693,
      "logits/rejected": -2.392331600189209,
      "logps/chosen": -275.5787353515625,
      "logps/rejected": -260.4464111328125,
      "loss": 0.0095,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.010565539821982384,
      "rewards/margins": 0.12289756536483765,
      "rewards/rejected": -0.13346309959888458,
      "step": 330
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 1.360283909503132,
      "learning_rate": 7.041266247556812e-08,
      "logits/chosen": -2.525031566619873,
      "logits/rejected": -2.481128215789795,
      "logps/chosen": -289.6778564453125,
      "logps/rejected": -272.6752014160156,
      "loss": 0.009,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.016852576285600662,
      "rewards/margins": 0.11005936563014984,
      "rewards/rejected": -0.1269119530916214,
      "step": 340
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 1.3078891508366295,
      "learning_rate": 5.706553665319955e-08,
      "logits/chosen": -2.501155138015747,
      "logits/rejected": -2.399773597717285,
      "logps/chosen": -277.02081298828125,
      "logps/rejected": -253.2749786376953,
      "loss": 0.0096,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.025297433137893677,
      "rewards/margins": 0.13083800673484802,
      "rewards/rejected": -0.1561354398727417,
      "step": 350
    },
    {
      "epoch": 0.8027522935779816,
      "eval_logits/chosen": -2.4845948219299316,
      "eval_logits/rejected": -2.390641927719116,
      "eval_logps/chosen": -285.5014343261719,
      "eval_logps/rejected": -262.4816589355469,
      "eval_loss": 0.008705741725862026,
      "eval_rewards/accuracies": 0.7801724076271057,
      "eval_rewards/chosen": -0.004113705363124609,
      "eval_rewards/margins": 0.15547847747802734,
      "eval_rewards/rejected": -0.15959219634532928,
      "eval_runtime": 93.024,
      "eval_samples_per_second": 19.543,
      "eval_steps_per_second": 0.312,
      "step": 350
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 1.1167320443463302,
      "learning_rate": 4.4956936350761005e-08,
      "logits/chosen": -2.4814512729644775,
      "logits/rejected": -2.4325201511383057,
      "logps/chosen": -252.41159057617188,
      "logps/rejected": -271.5174865722656,
      "loss": 0.0088,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.0009152599377557635,
      "rewards/margins": 0.128888800740242,
      "rewards/rejected": -0.12980404496192932,
      "step": 360
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 1.2531440295822198,
      "learning_rate": 3.416459164418123e-08,
      "logits/chosen": -2.5166125297546387,
      "logits/rejected": -2.4559273719787598,
      "logps/chosen": -299.89935302734375,
      "logps/rejected": -276.8605041503906,
      "loss": 0.0087,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.003148370888084173,
      "rewards/margins": 0.13971056044101715,
      "rewards/rejected": -0.14285892248153687,
      "step": 370
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 1.494922388543688,
      "learning_rate": 2.475778302439524e-08,
      "logits/chosen": -2.4861674308776855,
      "logits/rejected": -2.4356298446655273,
      "logps/chosen": -296.1441650390625,
      "logps/rejected": -272.86956787109375,
      "loss": 0.009,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.0023386895190924406,
      "rewards/margins": 0.15447205305099487,
      "rewards/rejected": -0.15213337540626526,
      "step": 380
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 1.505735003806321,
      "learning_rate": 1.6796896657433805e-08,
      "logits/chosen": -2.4833850860595703,
      "logits/rejected": -2.381284236907959,
      "logps/chosen": -252.93844604492188,
      "logps/rejected": -242.3744659423828,
      "loss": 0.0095,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.009809909388422966,
      "rewards/margins": 0.1208723783493042,
      "rewards/rejected": -0.13068227469921112,
      "step": 390
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 1.3636138422869082,
      "learning_rate": 1.0333036740834855e-08,
      "logits/chosen": -2.41188383102417,
      "logits/rejected": -2.378366231918335,
      "logps/chosen": -223.8693389892578,
      "logps/rejected": -241.79000854492188,
      "loss": 0.0093,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.006902736611664295,
      "rewards/margins": 0.13264694809913635,
      "rewards/rejected": -0.12574420869350433,
      "step": 400
    },
    {
      "epoch": 0.9174311926605505,
      "eval_logits/chosen": -2.4834423065185547,
      "eval_logits/rejected": -2.389298439025879,
      "eval_logps/chosen": -285.1916809082031,
      "eval_logps/rejected": -261.8979797363281,
      "eval_loss": 0.008685999549925327,
      "eval_rewards/accuracies": 0.7758620977401733,
      "eval_rewards/chosen": -0.0010154928313568234,
      "eval_rewards/margins": 0.15273970365524292,
      "eval_rewards/rejected": -0.15375518798828125,
      "eval_runtime": 92.569,
      "eval_samples_per_second": 19.639,
      "eval_steps_per_second": 0.313,
      "step": 400
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 1.3075058243660094,
      "learning_rate": 5.4076974448211685e-09,
      "logits/chosen": -2.419281005859375,
      "logits/rejected": -2.368946075439453,
      "logps/chosen": -271.1397705078125,
      "logps/rejected": -253.4750213623047,
      "loss": 0.009,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.0018791112815961242,
      "rewards/margins": 0.15240374207496643,
      "rewards/rejected": -0.1542828381061554,
      "step": 410
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 2.8175803183046364,
      "learning_rate": 2.052496544188487e-09,
      "logits/chosen": -2.462890148162842,
      "logits/rejected": -2.377373218536377,
      "logps/chosen": -260.7967834472656,
      "logps/rejected": -261.8900451660156,
      "loss": 0.0093,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.013578072190284729,
      "rewards/margins": 0.14654859900474548,
      "rewards/rejected": -0.16012665629386902,
      "step": 420
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 1.3745801458045255,
      "learning_rate": 2.889724508297886e-10,
      "logits/chosen": -2.4770209789276123,
      "logits/rejected": -2.3631789684295654,
      "logps/chosen": -303.4706115722656,
      "logps/rejected": -256.81878662109375,
      "loss": 0.0094,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.0029516436625272036,
      "rewards/margins": 0.12487111985683441,
      "rewards/rejected": -0.12782277166843414,
      "step": 430
    },
    {
      "epoch": 1.0,
      "step": 436,
      "total_flos": 0.0,
      "train_loss": 0.010263856175705927,
      "train_runtime": 11602.0724,
      "train_samples_per_second": 4.806,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 10,
  "max_steps": 436,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}