{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 391,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0025575447570332483,
      "grad_norm": 42.05885932037307,
      "learning_rate": 2.5e-09,
      "logits/chosen": -4.623842239379883,
      "logits/rejected": -4.85917854309082,
      "logps/chosen": -239.31422424316406,
      "logps/rejected": -207.56365966796875,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.02557544757033248,
      "grad_norm": 39.560773735648084,
      "learning_rate": 2.5e-08,
      "logits/chosen": -4.334544658660889,
      "logits/rejected": -4.644796848297119,
      "logps/chosen": -265.15618896484375,
      "logps/rejected": -215.6714630126953,
      "loss": 0.693,
      "rewards/accuracies": 0.4166666567325592,
      "rewards/chosen": -0.0004928099224343896,
      "rewards/margins": -0.0008595392573624849,
      "rewards/rejected": 0.00036672933492809534,
      "step": 10
    },
    {
      "epoch": 0.05115089514066496,
      "grad_norm": 40.83271143256618,
      "learning_rate": 5e-08,
      "logits/chosen": -4.509532928466797,
      "logits/rejected": -4.744012832641602,
      "logps/chosen": -267.80267333984375,
      "logps/rejected": -216.80471801757812,
      "loss": 0.6919,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.0013727399054914713,
      "rewards/margins": 0.0033264080993831158,
      "rewards/rejected": -0.0019536681938916445,
      "step": 20
    },
    {
      "epoch": 0.07672634271099744,
      "grad_norm": 43.48154475134036,
      "learning_rate": 7.5e-08,
      "logits/chosen": -4.5965423583984375,
      "logits/rejected": -4.777901649475098,
      "logps/chosen": -257.59088134765625,
      "logps/rejected": -215.49658203125,
      "loss": 0.6845,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.00941941887140274,
      "rewards/margins": 0.019057607278227806,
      "rewards/rejected": -0.00963818933814764,
      "step": 30
    },
    {
      "epoch": 0.10230179028132992,
      "grad_norm": 43.11247032025707,
      "learning_rate": 1e-07,
      "logits/chosen": -4.648722171783447,
      "logits/rejected": -4.745718002319336,
      "logps/chosen": -250.10897827148438,
      "logps/rejected": -223.86532592773438,
      "loss": 0.6588,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 0.03490210697054863,
      "rewards/margins": 0.07684428989887238,
      "rewards/rejected": -0.041942186653614044,
      "step": 40
    },
    {
      "epoch": 0.1278772378516624,
      "grad_norm": 47.11742069616159,
      "learning_rate": 9.979985922607475e-08,
      "logits/chosen": -4.593738555908203,
      "logits/rejected": -4.8337082862854,
      "logps/chosen": -267.30694580078125,
      "logps/rejected": -239.9588623046875,
      "loss": 0.6057,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.04662395641207695,
      "rewards/margins": 0.1874973475933075,
      "rewards/rejected": -0.14087337255477905,
      "step": 50
    },
    {
      "epoch": 0.1534526854219949,
      "grad_norm": 39.18274034042972,
      "learning_rate": 9.92010391574745e-08,
      "logits/chosen": -4.788964748382568,
      "logits/rejected": -4.883444786071777,
      "logps/chosen": -237.8981475830078,
      "logps/rejected": -257.84942626953125,
      "loss": 0.5174,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.03224308043718338,
      "rewards/margins": 0.46052321791648865,
      "rewards/rejected": -0.4282800555229187,
      "step": 60
    },
    {
      "epoch": 0.17902813299232737,
      "grad_norm": 34.832880831116846,
      "learning_rate": 9.820833372667812e-08,
      "logits/chosen": -4.657534599304199,
      "logits/rejected": -4.817151069641113,
      "logps/chosen": -249.996337890625,
      "logps/rejected": -280.097412109375,
      "loss": 0.4614,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.031456105411052704,
      "rewards/margins": 0.6548057198524475,
      "rewards/rejected": -0.6233495473861694,
      "step": 70
    },
    {
      "epoch": 0.20460358056265984,
      "grad_norm": 36.47722570862778,
      "learning_rate": 9.682969016701356e-08,
      "logits/chosen": -4.626967430114746,
      "logits/rejected": -4.778214454650879,
      "logps/chosen": -250.9975128173828,
      "logps/rejected": -311.1219177246094,
      "loss": 0.3904,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.10222460329532623,
      "rewards/margins": 0.9102567434310913,
      "rewards/rejected": -0.8080320358276367,
      "step": 80
    },
    {
      "epoch": 0.23017902813299232,
      "grad_norm": 30.998854450156045,
      "learning_rate": 9.507614539004081e-08,
      "logits/chosen": -4.739785194396973,
      "logits/rejected": -4.909841060638428,
      "logps/chosen": -237.671875,
      "logps/rejected": -307.8204040527344,
      "loss": 0.3509,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.1354086697101593,
      "rewards/margins": 1.1853126287460327,
      "rewards/rejected": -1.0499038696289062,
      "step": 90
    },
    {
      "epoch": 0.2557544757033248,
      "grad_norm": 42.52785579314538,
      "learning_rate": 9.296173762811083e-08,
      "logits/chosen": -4.647661209106445,
      "logits/rejected": -4.924945831298828,
      "logps/chosen": -244.45303344726562,
      "logps/rejected": -355.6828918457031,
      "loss": 0.3413,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.12265179306268692,
      "rewards/margins": 1.4083904027938843,
      "rewards/rejected": -1.2857385873794556,
      "step": 100
    },
    {
      "epoch": 0.2557544757033248,
      "eval_logits/chosen": -4.821703910827637,
      "eval_logits/rejected": -5.045117378234863,
      "eval_logps/chosen": -444.5645751953125,
      "eval_logps/rejected": -575.9554443359375,
      "eval_loss": 0.7230384349822998,
      "eval_rewards/accuracies": 0.515625,
      "eval_rewards/chosen": -0.5409007668495178,
      "eval_rewards/margins": 0.03477693349123001,
      "eval_rewards/rejected": -0.575677752494812,
      "eval_runtime": 98.6304,
      "eval_samples_per_second": 20.278,
      "eval_steps_per_second": 0.324,
      "step": 100
    },
    {
      "epoch": 0.2813299232736573,
      "grad_norm": 42.40417010662429,
      "learning_rate": 9.050339404945832e-08,
      "logits/chosen": -4.8084492683410645,
      "logits/rejected": -5.027788162231445,
      "logps/chosen": -247.86376953125,
      "logps/rejected": -369.55267333984375,
      "loss": 0.3143,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.01321962010115385,
      "rewards/margins": 1.5418872833251953,
      "rewards/rejected": -1.528667688369751,
      "step": 110
    },
    {
      "epoch": 0.3069053708439898,
      "grad_norm": 31.11429497548564,
      "learning_rate": 8.77207952455395e-08,
      "logits/chosen": -4.781357765197754,
      "logits/rejected": -5.055319786071777,
      "logps/chosen": -271.8451843261719,
      "logps/rejected": -396.73046875,
      "loss": 0.3042,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.041443757712841034,
      "rewards/margins": 1.7226619720458984,
      "rewards/rejected": -1.6812183856964111,
      "step": 120
    },
    {
      "epoch": 0.33248081841432225,
      "grad_norm": 32.05773581279916,
      "learning_rate": 8.463621767547997e-08,
      "logits/chosen": -4.876931190490723,
      "logits/rejected": -5.202266693115234,
      "logps/chosen": -264.2982482910156,
      "logps/rejected": -409.0570983886719,
      "loss": 0.2914,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.058543670922517776,
      "rewards/margins": 1.8787403106689453,
      "rewards/rejected": -1.937284231185913,
      "step": 130
    },
    {
      "epoch": 0.35805626598465473,
      "grad_norm": 31.96087329942538,
      "learning_rate": 8.127435532896387e-08,
      "logits/chosen": -4.971903324127197,
      "logits/rejected": -5.277985095977783,
      "logps/chosen": -305.4132385253906,
      "logps/rejected": -457.46343994140625,
      "loss": 0.274,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": -0.2121816873550415,
      "rewards/margins": 2.040717601776123,
      "rewards/rejected": -2.252899408340454,
      "step": 140
    },
    {
      "epoch": 0.3836317135549872,
      "grad_norm": 40.46461234858551,
      "learning_rate": 7.766212203526569e-08,
      "logits/chosen": -5.087113857269287,
      "logits/rejected": -5.368134498596191,
      "logps/chosen": -274.01080322265625,
      "logps/rejected": -457.38330078125,
      "loss": 0.2819,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.2159254252910614,
      "rewards/margins": 2.166714906692505,
      "rewards/rejected": -2.3826401233673096,
      "step": 150
    },
    {
      "epoch": 0.4092071611253197,
      "grad_norm": 32.057320142788335,
      "learning_rate": 7.382843600106538e-08,
      "logits/chosen": -5.177260398864746,
      "logits/rejected": -5.416450023651123,
      "logps/chosen": -284.1901550292969,
      "logps/rejected": -474.3257751464844,
      "loss": 0.2436,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.3273366093635559,
      "rewards/margins": 2.2598299980163574,
      "rewards/rejected": -2.5871663093566895,
      "step": 160
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 33.151157821087715,
      "learning_rate": 6.980398830195784e-08,
      "logits/chosen": -5.109088897705078,
      "logits/rejected": -5.438628196716309,
      "logps/chosen": -296.1925964355469,
      "logps/rejected": -516.4288940429688,
      "loss": 0.2364,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.3511837124824524,
      "rewards/margins": 2.6150753498077393,
      "rewards/rejected": -2.966259002685547,
      "step": 170
    },
    {
      "epoch": 0.46035805626598464,
      "grad_norm": 34.18806970089564,
      "learning_rate": 6.562099718102787e-08,
      "logits/chosen": -5.2773332595825195,
      "logits/rejected": -5.568037509918213,
      "logps/chosen": -284.951904296875,
      "logps/rejected": -486.5365295410156,
      "loss": 0.2628,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.480882465839386,
      "rewards/margins": 2.4242804050445557,
      "rewards/rejected": -2.9051625728607178,
      "step": 180
    },
    {
      "epoch": 0.4859335038363171,
      "grad_norm": 33.03269272782741,
      "learning_rate": 6.131295012148612e-08,
      "logits/chosen": -5.19248104095459,
      "logits/rejected": -5.355208396911621,
      "logps/chosen": -311.060791015625,
      "logps/rejected": -542.6156005859375,
      "loss": 0.2517,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": -0.5016793012619019,
      "rewards/margins": 2.5728163719177246,
      "rewards/rejected": -3.074495792388916,
      "step": 190
    },
    {
      "epoch": 0.5115089514066496,
      "grad_norm": 40.925552268276135,
      "learning_rate": 5.691433575823665e-08,
      "logits/chosen": -5.236765384674072,
      "logits/rejected": -5.465119361877441,
      "logps/chosen": -302.1981201171875,
      "logps/rejected": -515.2794799804688,
      "loss": 0.2653,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.5005888342857361,
      "rewards/margins": 2.4987406730651855,
      "rewards/rejected": -2.9993293285369873,
      "step": 200
    },
    {
      "epoch": 0.5115089514066496,
      "eval_logits/chosen": -5.226232528686523,
      "eval_logits/rejected": -5.50424337387085,
      "eval_logps/chosen": -540.43896484375,
      "eval_logps/rejected": -679.8809814453125,
      "eval_loss": 0.7765124440193176,
      "eval_rewards/accuracies": 0.54296875,
      "eval_rewards/chosen": -1.4996453523635864,
      "eval_rewards/margins": 0.11528739333152771,
      "eval_rewards/rejected": -1.6149327754974365,
      "eval_runtime": 98.5941,
      "eval_samples_per_second": 20.285,
      "eval_steps_per_second": 0.325,
      "step": 200
    },
    {
      "epoch": 0.5370843989769821,
      "grad_norm": 26.659672604447973,
      "learning_rate": 5.2460367774593905e-08,
      "logits/chosen": -5.310137748718262,
      "logits/rejected": -5.583542823791504,
      "logps/chosen": -303.67047119140625,
      "logps/rejected": -573.1016845703125,
      "loss": 0.2296,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.37785404920578003,
      "rewards/margins": 3.0667028427124023,
      "rewards/rejected": -3.444556713104248,
      "step": 210
    },
    {
      "epoch": 0.5626598465473146,
      "grad_norm": 35.355172011912686,
      "learning_rate": 4.798670299452925e-08,
      "logits/chosen": -5.1389665603637695,
      "logits/rejected": -5.567061424255371,
      "logps/chosen": -304.0540466308594,
      "logps/rejected": -569.4851684570312,
      "loss": 0.245,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -0.41242700815200806,
      "rewards/margins": 3.0356929302215576,
      "rewards/rejected": -3.4481201171875,
      "step": 220
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 37.35765448344736,
      "learning_rate": 4.3529155927297226e-08,
      "logits/chosen": -5.210625648498535,
      "logits/rejected": -5.601117134094238,
      "logps/chosen": -323.33135986328125,
      "logps/rejected": -591.130126953125,
      "loss": 0.2477,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -0.623622715473175,
      "rewards/margins": 3.050567150115967,
      "rewards/rejected": -3.674190044403076,
      "step": 230
    },
    {
      "epoch": 0.6138107416879796,
      "grad_norm": 27.168387739658527,
      "learning_rate": 3.9123412049691636e-08,
      "logits/chosen": -5.26107120513916,
      "logits/rejected": -5.582613945007324,
      "logps/chosen": -341.65289306640625,
      "logps/rejected": -593.1688232421875,
      "loss": 0.2349,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.68829745054245,
      "rewards/margins": 3.034055233001709,
      "rewards/rejected": -3.7223525047302246,
      "step": 240
    },
    {
      "epoch": 0.639386189258312,
      "grad_norm": 34.59601076495169,
      "learning_rate": 3.480474212128766e-08,
      "logits/chosen": -5.441601753234863,
      "logits/rejected": -5.72822380065918,
      "logps/chosen": -329.5417175292969,
      "logps/rejected": -537.7394409179688,
      "loss": 0.2339,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -0.8084025382995605,
      "rewards/margins": 2.4929001331329346,
      "rewards/rejected": -3.301302433013916,
      "step": 250
    },
    {
      "epoch": 0.6649616368286445,
      "grad_norm": 44.5395657806438,
      "learning_rate": 3.060771981975726e-08,
      "logits/chosen": -5.302738666534424,
      "logits/rejected": -5.622676372528076,
      "logps/chosen": -326.24041748046875,
      "logps/rejected": -637.6575927734375,
      "loss": 0.2325,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -0.7219182848930359,
      "rewards/margins": 3.4571731090545654,
      "rewards/rejected": -4.179091453552246,
      "step": 260
    },
    {
      "epoch": 0.690537084398977,
      "grad_norm": 33.64914034772639,
      "learning_rate": 2.6565944956764818e-08,
      "logits/chosen": -5.4421281814575195,
      "logits/rejected": -5.695931911468506,
      "logps/chosen": -332.70892333984375,
      "logps/rejected": -598.5055541992188,
      "loss": 0.2433,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -0.7167563438415527,
      "rewards/margins": 3.118049144744873,
      "rewards/rejected": -3.834805727005005,
      "step": 270
    },
    {
      "epoch": 0.7161125319693095,
      "grad_norm": 38.28164920230575,
      "learning_rate": 2.2711774490274766e-08,
      "logits/chosen": -5.344332695007324,
      "logits/rejected": -5.591184616088867,
      "logps/chosen": -331.06939697265625,
      "logps/rejected": -640.959228515625,
      "loss": 0.2174,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.665625810623169,
      "rewards/margins": 3.3145720958709717,
      "rewards/rejected": -3.9801979064941406,
      "step": 280
    },
    {
      "epoch": 0.7416879795396419,
      "grad_norm": 42.555865291815444,
      "learning_rate": 1.9076063486687256e-08,
      "logits/chosen": -5.223475933074951,
      "logits/rejected": -5.618660926818848,
      "logps/chosen": -328.63055419921875,
      "logps/rejected": -579.0905151367188,
      "loss": 0.2228,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -0.5602216124534607,
      "rewards/margins": 3.074389696121216,
      "rewards/rejected": -3.634611129760742,
      "step": 290
    },
    {
      "epoch": 0.7672634271099744,
      "grad_norm": 40.820437800178965,
      "learning_rate": 1.5687918106563324e-08,
      "logits/chosen": -5.369271755218506,
      "logits/rejected": -5.632781028747559,
      "logps/chosen": -320.268798828125,
      "logps/rejected": -608.9943237304688,
      "loss": 0.2424,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.6686061024665833,
      "rewards/margins": 3.2616829872131348,
      "rewards/rejected": -3.9302895069122314,
      "step": 300
    },
    {
      "epoch": 0.7672634271099744,
      "eval_logits/chosen": -5.394677639007568,
      "eval_logits/rejected": -5.655616283416748,
      "eval_logps/chosen": -579.441650390625,
      "eval_logps/rejected": -718.423828125,
      "eval_loss": 0.8002758622169495,
      "eval_rewards/accuracies": 0.52734375,
      "eval_rewards/chosen": -1.889671802520752,
      "eval_rewards/margins": 0.11068924516439438,
      "eval_rewards/rejected": -2.000361442565918,
      "eval_runtime": 98.5861,
      "eval_samples_per_second": 20.287,
      "eval_steps_per_second": 0.325,
      "step": 300
    },
    {
      "epoch": 0.7928388746803069,
      "grad_norm": 33.64379879568246,
      "learning_rate": 1.257446259144494e-08,
      "logits/chosen": -5.246872425079346,
      "logits/rejected": -5.653367042541504,
      "logps/chosen": -315.7105407714844,
      "logps/rejected": -625.9619140625,
      "loss": 0.2323,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": -0.5457090735435486,
      "rewards/margins": 3.491931200027466,
      "rewards/rejected": -4.03764009475708,
      "step": 310
    },
    {
      "epoch": 0.8184143222506394,
      "grad_norm": 35.35694379401523,
      "learning_rate": 9.760622117187234e-09,
      "logits/chosen": -5.381436824798584,
      "logits/rejected": -5.7473673820495605,
      "logps/chosen": -314.6531677246094,
      "logps/rejected": -594.841552734375,
      "loss": 0.2466,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.6948888897895813,
      "rewards/margins": 3.1890125274658203,
      "rewards/rejected": -3.8839008808135986,
      "step": 320
    },
    {
      "epoch": 0.8439897698209718,
      "grad_norm": 34.515465680243125,
      "learning_rate": 7.2689232521989885e-09,
      "logits/chosen": -5.308783531188965,
      "logits/rejected": -5.656357765197754,
      "logps/chosen": -347.4857482910156,
      "logps/rejected": -629.4615478515625,
      "loss": 0.2233,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.7933691143989563,
      "rewards/margins": 3.1742498874664307,
      "rewards/rejected": -3.967618942260742,
      "step": 330
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 28.542655038843865,
      "learning_rate": 5.119313618049309e-09,
      "logits/chosen": -5.346091270446777,
      "logits/rejected": -5.756931781768799,
      "logps/chosen": -338.65509033203125,
      "logps/rejected": -575.4979858398438,
      "loss": 0.2174,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.6450907588005066,
      "rewards/margins": 3.0420687198638916,
      "rewards/rejected": -3.687159299850464,
      "step": 340
    },
    {
      "epoch": 0.8951406649616368,
      "grad_norm": 33.42105425863571,
      "learning_rate": 3.3290021961708158e-09,
      "logits/chosen": -5.374421119689941,
      "logits/rejected": -5.536851406097412,
      "logps/chosen": -333.8661193847656,
      "logps/rejected": -595.2741088867188,
      "loss": 0.2467,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.7839492559432983,
      "rewards/margins": 2.891091823577881,
      "rewards/rejected": -3.6750411987304688,
      "step": 350
    },
    {
      "epoch": 0.9207161125319693,
      "grad_norm": 33.175441995042306,
      "learning_rate": 1.9123215591052013e-09,
      "logits/chosen": -5.3232526779174805,
      "logits/rejected": -5.559803485870361,
      "logps/chosen": -337.17694091796875,
      "logps/rejected": -596.7660522460938,
      "loss": 0.2397,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.7554206252098083,
      "rewards/margins": 2.937615156173706,
      "rewards/rejected": -3.693035840988159,
      "step": 360
    },
    {
      "epoch": 0.9462915601023018,
      "grad_norm": 33.50889046296721,
      "learning_rate": 8.806131292167618e-10,
      "logits/chosen": -5.363125801086426,
      "logits/rejected": -5.561426162719727,
      "logps/chosen": -327.09295654296875,
      "logps/rejected": -608.7786865234375,
      "loss": 0.238,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.7027177214622498,
      "rewards/margins": 3.01659893989563,
      "rewards/rejected": -3.719316005706787,
      "step": 370
    },
    {
      "epoch": 0.9718670076726342,
      "grad_norm": 42.30140132740828,
      "learning_rate": 2.4213638345040867e-10,
      "logits/chosen": -5.489308834075928,
      "logits/rejected": -5.787456512451172,
      "logps/chosen": -332.35858154296875,
      "logps/rejected": -607.3480224609375,
      "loss": 0.2341,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -0.7027586698532104,
      "rewards/margins": 3.165475368499756,
      "rewards/rejected": -3.8682339191436768,
      "step": 380
    },
    {
      "epoch": 0.9974424552429667,
      "grad_norm": 37.88179259111206,
      "learning_rate": 2.0027310073833516e-12,
      "logits/chosen": -5.485334873199463,
      "logits/rejected": -5.764852046966553,
      "logps/chosen": -331.56610107421875,
      "logps/rejected": -614.2138061523438,
      "loss": 0.2223,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -0.7189357876777649,
      "rewards/margins": 3.2180511951446533,
      "rewards/rejected": -3.9369864463806152,
      "step": 390
    },
    {
      "epoch": 1.0,
      "step": 391,
      "total_flos": 0.0,
      "train_loss": 0.3220548828315857,
      "train_runtime": 6253.066,
      "train_samples_per_second": 7.996,
      "train_steps_per_second": 0.063
    }
  ],
  "logging_steps": 10,
  "max_steps": 391,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}