{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6837606837606838,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017094017094017096,
      "grad_norm": 35.00202204303521,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7455849647521973,
      "logits/rejected": -2.7442612648010254,
      "logps/chosen": -164.2725830078125,
      "logps/rejected": -170.57113647460938,
      "loss": 0.6934,
      "rewards/accuracies": 0.23749999701976776,
      "rewards/chosen": 0.0026612328365445137,
      "rewards/margins": -0.001539617427624762,
      "rewards/rejected": 0.004200850613415241,
      "step": 5
    },
    {
      "epoch": 0.03418803418803419,
      "grad_norm": 36.29266486314593,
      "learning_rate": 1e-06,
      "logits/chosen": -2.709902763366699,
      "logits/rejected": -2.7155404090881348,
      "logps/chosen": -171.80032348632812,
      "logps/rejected": -165.20169067382812,
      "loss": 0.6879,
      "rewards/accuracies": 0.5062500238418579,
      "rewards/chosen": 0.012009668163955212,
      "rewards/margins": 0.0021203968208283186,
      "rewards/rejected": 0.009889272041618824,
      "step": 10
    },
    {
      "epoch": 0.05128205128205128,
      "grad_norm": 33.83921269470837,
      "learning_rate": 9.999177507263144e-07,
      "logits/chosen": -2.6502068042755127,
      "logits/rejected": -2.628007411956787,
      "logps/chosen": -174.082275390625,
      "logps/rejected": -174.13429260253906,
      "loss": 0.6698,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": 0.23495244979858398,
      "rewards/margins": 0.1125468835234642,
      "rewards/rejected": 0.12240554392337799,
      "step": 15
    },
    {
      "epoch": 0.06837606837606838,
      "grad_norm": 34.14427373918799,
      "learning_rate": 9.996710299650301e-07,
      "logits/chosen": -2.473665714263916,
      "logits/rejected": -2.4469008445739746,
      "logps/chosen": -158.2163848876953,
      "logps/rejected": -158.0710906982422,
      "loss": 0.661,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": 0.4233472943305969,
      "rewards/margins": 0.1434161365032196,
      "rewards/rejected": 0.2799311578273773,
      "step": 20
    },
    {
      "epoch": 0.08547008547008547,
      "grad_norm": 33.2696083475879,
      "learning_rate": 9.992599188865604e-07,
      "logits/chosen": -2.314507007598877,
      "logits/rejected": -2.3168132305145264,
      "logps/chosen": -150.67019653320312,
      "logps/rejected": -156.8417510986328,
      "loss": 0.6501,
      "rewards/accuracies": 0.543749988079071,
      "rewards/chosen": 0.4975205063819885,
      "rewards/margins": 0.15743504464626312,
      "rewards/rejected": 0.3400854766368866,
      "step": 25
    },
    {
      "epoch": 0.10256410256410256,
      "grad_norm": 34.42253361988952,
      "learning_rate": 9.98684552745256e-07,
      "logits/chosen": -2.243194103240967,
      "logits/rejected": -2.251340866088867,
      "logps/chosen": -161.2266845703125,
      "logps/rejected": -161.32298278808594,
      "loss": 0.6289,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": 0.4243805408477783,
      "rewards/margins": 0.2635195851325989,
      "rewards/rejected": 0.16086098551750183,
      "step": 30
    },
    {
      "epoch": 0.11965811965811966,
      "grad_norm": 31.414296706456245,
      "learning_rate": 9.979451208349055e-07,
      "logits/chosen": -2.30315899848938,
      "logits/rejected": -2.289762496948242,
      "logps/chosen": -171.71713256835938,
      "logps/rejected": -174.50900268554688,
      "loss": 0.6296,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.019384615123271942,
      "rewards/margins": 0.318477988243103,
      "rewards/rejected": -0.3378625512123108,
      "step": 35
    },
    {
      "epoch": 0.13675213675213677,
      "grad_norm": 32.071830655862556,
      "learning_rate": 9.970418664264595e-07,
      "logits/chosen": -2.3935599327087402,
      "logits/rejected": -2.3812546730041504,
      "logps/chosen": -171.0698699951172,
      "logps/rejected": -176.58578491210938,
      "loss": 0.5991,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.26089853048324585,
      "rewards/margins": 0.5235068202018738,
      "rewards/rejected": -0.7844053506851196,
      "step": 40
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 36.19466541168301,
      "learning_rate": 9.95975086687994e-07,
      "logits/chosen": -2.4914021492004395,
      "logits/rejected": -2.4973323345184326,
      "logps/chosen": -163.68099975585938,
      "logps/rejected": -167.174072265625,
      "loss": 0.6141,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.2962096929550171,
      "rewards/margins": 0.4588828682899475,
      "rewards/rejected": -0.7550925016403198,
      "step": 45
    },
    {
      "epoch": 0.17094017094017094,
      "grad_norm": 31.16276115760231,
      "learning_rate": 9.947451325869439e-07,
      "logits/chosen": -2.5575203895568848,
      "logits/rejected": -2.557717800140381,
      "logps/chosen": -172.04318237304688,
      "logps/rejected": -177.67672729492188,
      "loss": 0.5777,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.2069791853427887,
      "rewards/margins": 0.6018465757369995,
      "rewards/rejected": -0.808825671672821,
      "step": 50
    },
    {
      "epoch": 0.17094017094017094,
      "eval_logits/chosen": -2.5221025943756104,
      "eval_logits/rejected": -2.5152711868286133,
      "eval_logps/chosen": -163.01820373535156,
      "eval_logps/rejected": -169.54832458496094,
      "eval_loss": 0.5812540650367737,
      "eval_rewards/accuracies": 0.6682692170143127,
      "eval_rewards/chosen": -0.45408713817596436,
      "eval_rewards/margins": 0.6127156615257263,
      "eval_rewards/rejected": -1.0668028593063354,
      "eval_runtime": 510.3361,
      "eval_samples_per_second": 16.291,
      "eval_steps_per_second": 0.255,
      "step": 50
    },
    {
      "epoch": 0.18803418803418803,
      "grad_norm": 31.575578721339145,
      "learning_rate": 9.933524087746347e-07,
      "logits/chosen": -2.490377426147461,
      "logits/rejected": -2.4825081825256348,
      "logps/chosen": -168.06161499023438,
      "logps/rejected": -175.0494384765625,
      "loss": 0.5706,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.5060762763023376,
      "rewards/margins": 0.7589826583862305,
      "rewards/rejected": -1.2650587558746338,
      "step": 55
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 30.171745273288415,
      "learning_rate": 9.917973734531549e-07,
      "logits/chosen": -2.48228120803833,
      "logits/rejected": -2.4833157062530518,
      "logps/chosen": -159.47142028808594,
      "logps/rejected": -170.63671875,
      "loss": 0.5753,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.35752761363983154,
      "rewards/margins": 0.5991309881210327,
      "rewards/rejected": -0.9566585421562195,
      "step": 60
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 32.13878319029882,
      "learning_rate": 9.90080538224607e-07,
      "logits/chosen": -2.585407018661499,
      "logits/rejected": -2.5767769813537598,
      "logps/chosen": -157.43936157226562,
      "logps/rejected": -166.13589477539062,
      "loss": 0.566,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.02057185396552086,
      "rewards/margins": 0.47568243741989136,
      "rewards/rejected": -0.4962543547153473,
      "step": 65
    },
    {
      "epoch": 0.23931623931623933,
      "grad_norm": 29.494674721856043,
      "learning_rate": 9.882024679227938e-07,
      "logits/chosen": -2.6504979133605957,
      "logits/rejected": -2.6398470401763916,
      "logps/chosen": -178.0801239013672,
      "logps/rejected": -179.46328735351562,
      "loss": 0.5444,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.43436694145202637,
      "rewards/margins": 0.8427752256393433,
      "rewards/rejected": -1.27714204788208,
      "step": 70
    },
    {
      "epoch": 0.2564102564102564,
      "grad_norm": 28.856733948308104,
      "learning_rate": 9.861637804273881e-07,
      "logits/chosen": -2.660489082336426,
      "logits/rejected": -2.655539035797119,
      "logps/chosen": -162.1233673095703,
      "logps/rejected": -170.16131591796875,
      "loss": 0.5568,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.4032784402370453,
      "rewards/margins": 0.6959114074707031,
      "rewards/rejected": -1.0991899967193604,
      "step": 75
    },
    {
      "epoch": 0.27350427350427353,
      "grad_norm": 26.646061534818323,
      "learning_rate": 9.83965146460653e-07,
      "logits/chosen": -2.6391615867614746,
      "logits/rejected": -2.628577709197998,
      "logps/chosen": -168.58099365234375,
      "logps/rejected": -179.22805786132812,
      "loss": 0.5448,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.6665827035903931,
      "rewards/margins": 0.8240470886230469,
      "rewards/rejected": -1.4906299114227295,
      "step": 80
    },
    {
      "epoch": 0.2905982905982906,
      "grad_norm": 36.04159750418885,
      "learning_rate": 9.816072893667758e-07,
      "logits/chosen": -2.6322970390319824,
      "logits/rejected": -2.6053385734558105,
      "logps/chosen": -174.82640075683594,
      "logps/rejected": -186.0735626220703,
      "loss": 0.5579,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -1.0639268159866333,
      "rewards/margins": 1.0258175134658813,
      "rewards/rejected": -2.0897443294525146,
      "step": 85
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 26.922939193632168,
      "learning_rate": 9.790909848738904e-07,
      "logits/chosen": -2.60801362991333,
      "logits/rejected": -2.6101624965667725,
      "logps/chosen": -176.20538330078125,
      "logps/rejected": -184.7812957763672,
      "loss": 0.5215,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.9929834604263306,
      "rewards/margins": 0.8646324276924133,
      "rewards/rejected": -1.8576160669326782,
      "step": 90
    },
    {
      "epoch": 0.3247863247863248,
      "grad_norm": 30.3564450245371,
      "learning_rate": 9.764170608388647e-07,
      "logits/chosen": -2.6054036617279053,
      "logits/rejected": -2.5733799934387207,
      "logps/chosen": -168.037109375,
      "logps/rejected": -174.51144409179688,
      "loss": 0.5197,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.6652337312698364,
      "rewards/margins": 1.060430884361267,
      "rewards/rejected": -1.725664734840393,
      "step": 95
    },
    {
      "epoch": 0.3418803418803419,
      "grad_norm": 28.936164680674203,
      "learning_rate": 9.735863969749371e-07,
      "logits/chosen": -2.5255179405212402,
      "logits/rejected": -2.4874520301818848,
      "logps/chosen": -177.73861694335938,
      "logps/rejected": -189.82369995117188,
      "loss": 0.4982,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.8511013984680176,
      "rewards/margins": 1.1354777812957764,
      "rewards/rejected": -1.986579179763794,
      "step": 100
    },
    {
      "epoch": 0.3418803418803419,
      "eval_logits/chosen": -2.484687328338623,
      "eval_logits/rejected": -2.460559368133545,
      "eval_logps/chosen": -168.28323364257812,
      "eval_logps/rejected": -180.8539276123047,
      "eval_loss": 0.5161151885986328,
      "eval_rewards/accuracies": 0.7211538553237915,
      "eval_rewards/chosen": -0.9805887937545776,
      "eval_rewards/margins": 1.2167747020721436,
      "eval_rewards/rejected": -2.1973636150360107,
      "eval_runtime": 510.3447,
      "eval_samples_per_second": 16.291,
      "eval_steps_per_second": 0.255,
      "step": 100
    },
    {
      "epoch": 0.358974358974359,
      "grad_norm": 31.089971589067016,
      "learning_rate": 9.705999245622956e-07,
      "logits/chosen": -2.4702706336975098,
      "logits/rejected": -2.4523651599884033,
      "logps/chosen": -170.59246826171875,
      "logps/rejected": -182.99813842773438,
      "loss": 0.4991,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.8999192118644714,
      "rewards/margins": 0.8702341318130493,
      "rewards/rejected": -1.770153284072876,
      "step": 105
    },
    {
      "epoch": 0.37606837606837606,
      "grad_norm": 27.339023914835686,
      "learning_rate": 9.674586261416873e-07,
      "logits/chosen": -2.4866347312927246,
      "logits/rejected": -2.4518179893493652,
      "logps/chosen": -179.46290588378906,
      "logps/rejected": -188.7920379638672,
      "loss": 0.5213,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.6938365697860718,
      "rewards/margins": 1.0765600204467773,
      "rewards/rejected": -1.7703965902328491,
      "step": 110
    },
    {
      "epoch": 0.39316239316239315,
      "grad_norm": 31.312902469600562,
      "learning_rate": 9.641635351911664e-07,
      "logits/chosen": -2.4456398487091064,
      "logits/rejected": -2.426159381866455,
      "logps/chosen": -170.3855438232422,
      "logps/rejected": -181.9676513671875,
      "loss": 0.4823,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.8490931391716003,
      "rewards/margins": 1.2224478721618652,
      "rewards/rejected": -2.0715408325195312,
      "step": 115
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 25.966469642807997,
      "learning_rate": 9.607157357860821e-07,
      "logits/chosen": -2.4072113037109375,
      "logits/rejected": -2.3874144554138184,
      "logps/chosen": -187.41197204589844,
      "logps/rejected": -201.69454956054688,
      "loss": 0.5037,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.0798847675323486,
      "rewards/margins": 1.3365159034729004,
      "rewards/rejected": -2.41640043258667,
      "step": 120
    },
    {
      "epoch": 0.42735042735042733,
      "grad_norm": 32.18242375190423,
      "learning_rate": 9.571163622424225e-07,
      "logits/chosen": -2.2766659259796143,
      "logits/rejected": -2.252072811126709,
      "logps/chosen": -174.78514099121094,
      "logps/rejected": -187.40646362304688,
      "loss": 0.5019,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -1.5247443914413452,
      "rewards/margins": 1.2177503108978271,
      "rewards/rejected": -2.742494821548462,
      "step": 125
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 30.579550576640443,
      "learning_rate": 9.533665987436261e-07,
      "logits/chosen": -2.182610034942627,
      "logits/rejected": -2.128113269805908,
      "logps/chosen": -178.1033477783203,
      "logps/rejected": -197.4458465576172,
      "loss": 0.4957,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.5623412132263184,
      "rewards/margins": 1.2608497142791748,
      "rewards/rejected": -2.823190689086914,
      "step": 130
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 29.287644373971865,
      "learning_rate": 9.494676789509899e-07,
      "logits/chosen": -2.1067867279052734,
      "logits/rejected": -2.0683400630950928,
      "logps/chosen": -176.67918395996094,
      "logps/rejected": -193.65371704101562,
      "loss": 0.4939,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -1.1053364276885986,
      "rewards/margins": 1.3829355239868164,
      "rewards/rejected": -2.488272190093994,
      "step": 135
    },
    {
      "epoch": 0.47863247863247865,
      "grad_norm": 27.563555703636343,
      "learning_rate": 9.454208855977985e-07,
      "logits/chosen": -2.0855822563171387,
      "logits/rejected": -2.013296127319336,
      "logps/chosen": -178.40390014648438,
      "logps/rejected": -196.03305053710938,
      "loss": 0.4715,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -1.443866491317749,
      "rewards/margins": 1.600778341293335,
      "rewards/rejected": -3.044644832611084,
      "step": 140
    },
    {
      "epoch": 0.49572649572649574,
      "grad_norm": 30.234814125811326,
      "learning_rate": 9.41227550067308e-07,
      "logits/chosen": -2.0734238624572754,
      "logits/rejected": -2.0634400844573975,
      "logps/chosen": -179.51080322265625,
      "logps/rejected": -191.87046813964844,
      "loss": 0.4798,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.6011661291122437,
      "rewards/margins": 1.445229172706604,
      "rewards/rejected": -3.0463955402374268,
      "step": 145
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 31.371346339775513,
      "learning_rate": 9.36889051954725e-07,
      "logits/chosen": -2.127821683883667,
      "logits/rejected": -2.080082416534424,
      "logps/chosen": -180.66383361816406,
      "logps/rejected": -196.1031494140625,
      "loss": 0.4954,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.6335647106170654,
      "rewards/margins": 1.5422546863555908,
      "rewards/rejected": -3.1758196353912354,
      "step": 150
    },
    {
      "epoch": 0.5128205128205128,
      "eval_logits/chosen": -2.1472573280334473,
      "eval_logits/rejected": -2.0990829467773438,
      "eval_logps/chosen": -173.8290557861328,
      "eval_logps/rejected": -191.683349609375,
      "eval_loss": 0.47699737548828125,
      "eval_rewards/accuracies": 0.754807710647583,
      "eval_rewards/chosen": -1.5351712703704834,
      "eval_rewards/margins": 1.7451337575912476,
      "eval_rewards/rejected": -3.2803049087524414,
      "eval_runtime": 510.7048,
      "eval_samples_per_second": 16.279,
      "eval_steps_per_second": 0.255,
      "step": 150
    },
    {
      "epoch": 0.5299145299145299,
      "grad_norm": 27.790777356361556,
      "learning_rate": 9.324068186133245e-07,
      "logits/chosen": -2.1372084617614746,
      "logits/rejected": -2.124948024749756,
      "logps/chosen": -172.3369598388672,
      "logps/rejected": -186.5850372314453,
      "loss": 0.4644,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -1.3583369255065918,
      "rewards/margins": 1.7439367771148682,
      "rewards/rejected": -3.102273464202881,
      "step": 155
    },
    {
      "epoch": 0.5470085470085471,
      "grad_norm": 27.612583401785376,
      "learning_rate": 9.277823246848536e-07,
      "logits/chosen": -2.2635793685913086,
      "logits/rejected": -2.2123026847839355,
      "logps/chosen": -186.25137329101562,
      "logps/rejected": -196.69564819335938,
      "loss": 0.4547,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.2192834615707397,
      "rewards/margins": 1.371063470840454,
      "rewards/rejected": -2.5903468132019043,
      "step": 160
    },
    {
      "epoch": 0.5641025641025641,
      "grad_norm": 28.168886287584876,
      "learning_rate": 9.230170916143793e-07,
      "logits/chosen": -2.3309006690979004,
      "logits/rejected": -2.2978808879852295,
      "logps/chosen": -174.7559814453125,
      "logps/rejected": -195.68280029296875,
      "loss": 0.4967,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.0855658054351807,
      "rewards/margins": 1.6249233484268188,
      "rewards/rejected": -2.710489273071289,
      "step": 165
    },
    {
      "epoch": 0.5811965811965812,
      "grad_norm": 28.93959851544435,
      "learning_rate": 9.181126871497378e-07,
      "logits/chosen": -2.376833915710449,
      "logits/rejected": -2.340681552886963,
      "logps/chosen": -175.3675537109375,
      "logps/rejected": -194.9619903564453,
      "loss": 0.4651,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.9624043703079224,
      "rewards/margins": 1.7745708227157593,
      "rewards/rejected": -2.7369751930236816,
      "step": 170
    },
    {
      "epoch": 0.5982905982905983,
      "grad_norm": 30.43477724579486,
      "learning_rate": 9.130707248257491e-07,
      "logits/chosen": -2.458378553390503,
      "logits/rejected": -2.4171223640441895,
      "logps/chosen": -168.79849243164062,
      "logps/rejected": -178.6556396484375,
      "loss": 0.4728,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.9254748225212097,
      "rewards/margins": 1.5751961469650269,
      "rewards/rejected": -2.500671148300171,
      "step": 175
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 25.220318056395065,
      "learning_rate": 9.078928634333698e-07,
      "logits/chosen": -2.4454641342163086,
      "logits/rejected": -2.4170265197753906,
      "logps/chosen": -181.41317749023438,
      "logps/rejected": -199.88668823242188,
      "loss": 0.4526,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": -0.84205561876297,
      "rewards/margins": 1.7803510427474976,
      "rewards/rejected": -2.622406482696533,
      "step": 180
    },
    {
      "epoch": 0.6324786324786325,
      "grad_norm": 29.414031929374275,
      "learning_rate": 9.025808064739549e-07,
      "logits/chosen": -2.4103400707244873,
      "logits/rejected": -2.370731830596924,
      "logps/chosen": -178.70916748046875,
      "logps/rejected": -193.0004119873047,
      "loss": 0.4891,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -1.174070119857788,
      "rewards/margins": 1.535239338874817,
      "rewards/rejected": -2.7093093395233154,
      "step": 185
    },
    {
      "epoch": 0.6495726495726496,
      "grad_norm": 26.0372223221703,
      "learning_rate": 8.971363015988113e-07,
      "logits/chosen": -2.3428735733032227,
      "logits/rejected": -2.2986531257629395,
      "logps/chosen": -173.8651580810547,
      "logps/rejected": -194.9317626953125,
      "loss": 0.4643,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -1.1425771713256836,
      "rewards/margins": 1.611919641494751,
      "rewards/rejected": -2.7544968128204346,
      "step": 190
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 33.87434178682573,
      "learning_rate": 8.91561140034225e-07,
      "logits/chosen": -2.2664923667907715,
      "logits/rejected": -2.2088184356689453,
      "logps/chosen": -172.7240753173828,
      "logps/rejected": -193.1275177001953,
      "loss": 0.5029,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.3134868144989014,
      "rewards/margins": 1.5050963163375854,
      "rewards/rejected": -2.8185834884643555,
      "step": 195
    },
    {
      "epoch": 0.6837606837606838,
      "grad_norm": 25.21313391058931,
      "learning_rate": 8.858571559921537e-07,
      "logits/chosen": -2.191737174987793,
      "logits/rejected": -2.1188113689422607,
      "logps/chosen": -174.46722412109375,
      "logps/rejected": -188.4197540283203,
      "loss": 0.4567,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.026064157485962,
      "rewards/margins": 1.5865710973739624,
      "rewards/rejected": -2.612635374069214,
      "step": 200
    },
    {
      "epoch": 0.6837606837606838,
      "eval_logits/chosen": -2.1586899757385254,
      "eval_logits/rejected": -2.1090493202209473,
      "eval_logps/chosen": -170.4287872314453,
      "eval_logps/rejected": -187.2865447998047,
      "eval_loss": 0.45979756116867065,
      "eval_rewards/accuracies": 0.7596153616905212,
      "eval_rewards/chosen": -1.1951465606689453,
      "eval_rewards/margins": 1.64547860622406,
      "eval_rewards/rejected": -2.840625047683716,
      "eval_runtime": 510.4854,
      "eval_samples_per_second": 16.286,
      "eval_steps_per_second": 0.255,
      "step": 200
    }
  ],
  "logging_steps": 5,
  "max_steps": 876,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2358113407598592.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}