{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 329,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "dpo_losses": 0.6931471824645996,
      "epoch": 0.0,
      "grad_norm": 2.0956511335509704,
      "learning_rate": 1.5151515151515152e-08,
      "logits/chosen": -2.6820077896118164,
      "logits/rejected": -2.6930205821990967,
      "logps/chosen": -281.2528381347656,
      "logps/rejected": -258.0622253417969,
      "loss": 0.6931,
      "positive_losses": 0.0,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/margins_max": 0.0,
      "rewards/margins_min": 0.0,
      "rewards/margins_std": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "dpo_losses": 0.6928904056549072,
      "epoch": 0.03,
      "grad_norm": 25.14467901589718,
      "learning_rate": 1.5151515151515152e-07,
      "logits/chosen": -2.769256353378296,
      "logits/rejected": -2.7548859119415283,
      "logps/chosen": -284.5684814453125,
      "logps/rejected": -249.87168884277344,
      "loss": 0.6979,
      "positive_losses": 0.05147833377122879,
      "rewards/accuracies": 0.4583333432674408,
      "rewards/chosen": 0.0003567033272702247,
      "rewards/margins": 0.000515056774020195,
      "rewards/margins_max": 0.0027906966861337423,
      "rewards/margins_min": -0.0017034021439030766,
      "rewards/margins_std": 0.0019198498921468854,
      "rewards/rejected": -0.00015835335943847895,
      "step": 10
    },
    {
      "dpo_losses": 0.6927427053451538,
      "epoch": 0.06,
      "grad_norm": 15.081930583547756,
      "learning_rate": 3.0303030303030305e-07,
      "logits/chosen": -2.8346035480499268,
      "logits/rejected": -2.7817800045013428,
      "logps/chosen": -291.3860168457031,
      "logps/rejected": -270.3315124511719,
      "loss": 0.6968,
      "positive_losses": 0.02039351500570774,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.001457269536331296,
      "rewards/margins": 0.0008110570488497615,
      "rewards/margins_max": 0.0036527395714074373,
      "rewards/margins_min": -0.0018750065937638283,
      "rewards/margins_std": 0.0025213216431438923,
      "rewards/rejected": 0.0006462126038968563,
      "step": 20
    },
    {
      "dpo_losses": 0.6933337450027466,
      "epoch": 0.09,
      "grad_norm": 9.990963095062185,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.8629353046417236,
      "logits/rejected": -2.8151936531066895,
      "logps/chosen": -258.76837158203125,
      "logps/rejected": -226.8423614501953,
      "loss": 0.6951,
      "positive_losses": 0.014158916659653187,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 0.004769365303218365,
      "rewards/margins": -0.00037118454929441214,
      "rewards/margins_max": 0.002330085728317499,
      "rewards/margins_min": -0.0032659140415489674,
      "rewards/margins_std": 0.0024877875111997128,
      "rewards/rejected": 0.005140549503266811,
      "step": 30
    },
    {
      "dpo_losses": 0.6926981806755066,
      "epoch": 0.12,
      "grad_norm": 1.604193261987099,
      "learning_rate": 4.993103596812268e-07,
      "logits/chosen": -2.829625129699707,
      "logits/rejected": -2.764366626739502,
      "logps/chosen": -316.5284729003906,
      "logps/rejected": -223.76123046875,
      "loss": 0.693,
      "positive_losses": 0.007878494448959827,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.010137422941625118,
      "rewards/margins": 0.0009021027944982052,
      "rewards/margins_max": 0.005337115842849016,
      "rewards/margins_min": -0.0031714539509266615,
      "rewards/margins_std": 0.0037328810431063175,
      "rewards/rejected": 0.0092353206127882,
      "step": 40
    },
    {
      "dpo_losses": 0.6929700374603271,
      "epoch": 0.15,
      "grad_norm": 7.523440542250164,
      "learning_rate": 4.959416858332709e-07,
      "logits/chosen": -2.790869951248169,
      "logits/rejected": -2.8046393394470215,
      "logps/chosen": -241.8709259033203,
      "logps/rejected": -278.8560791015625,
      "loss": 0.6937,
      "positive_losses": 0.0050521851517260075,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.010639366693794727,
      "rewards/margins": 0.0003602326032705605,
      "rewards/margins_max": 0.005497736390680075,
      "rewards/margins_min": -0.004975494462996721,
      "rewards/margins_std": 0.0046127657406032085,
      "rewards/rejected": 0.010279135778546333,
      "step": 50
    },
    {
      "dpo_losses": 0.6924153566360474,
      "epoch": 0.18,
      "grad_norm": 8.317758374112413,
      "learning_rate": 4.898051734555674e-07,
      "logits/chosen": -2.834188938140869,
      "logits/rejected": -2.8444936275482178,
      "logps/chosen": -320.539794921875,
      "logps/rejected": -281.9419860839844,
      "loss": 0.6928,
      "positive_losses": 0.0005203246837481856,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.013813069090247154,
      "rewards/margins": 0.0014724148204550147,
      "rewards/margins_max": 0.008079244755208492,
      "rewards/margins_min": -0.003974889405071735,
      "rewards/margins_std": 0.005347874015569687,
      "rewards/rejected": 0.012340654619038105,
      "step": 60
    },
    {
      "dpo_losses": 0.693148672580719,
      "epoch": 0.21,
      "grad_norm": 7.343753116688168,
      "learning_rate": 4.809698831278217e-07,
      "logits/chosen": -2.7492079734802246,
      "logits/rejected": -2.735715389251709,
      "logps/chosen": -265.208251953125,
      "logps/rejected": -245.1059112548828,
      "loss": 0.6937,
      "positive_losses": 0.01350479107350111,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.012691942974925041,
      "rewards/margins": 6.041547749191523e-06,
      "rewards/margins_max": 0.0061768521554768085,
      "rewards/margins_min": -0.0060454607009887695,
      "rewards/margins_std": 0.00548047199845314,
      "rewards/rejected": 0.012685902416706085,
      "step": 70
    },
    {
      "dpo_losses": 0.6920666694641113,
      "epoch": 0.24,
      "grad_norm": 2.0944283368710837,
      "learning_rate": 4.6953524759527053e-07,
      "logits/chosen": -2.8422322273254395,
      "logits/rejected": -2.815117597579956,
      "logps/chosen": -280.7608947753906,
      "logps/rejected": -273.6212158203125,
      "loss": 0.6925,
      "positive_losses": 0.00601539621129632,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.01512569934129715,
      "rewards/margins": 0.0021733222529292107,
      "rewards/margins_max": 0.008934634737670422,
      "rewards/margins_min": -0.004246932454407215,
      "rewards/margins_std": 0.006057220045477152,
      "rewards/rejected": 0.012952374294400215,
      "step": 80
    },
    {
      "dpo_losses": 0.6931439638137817,
      "epoch": 0.27,
      "grad_norm": 7.341115665723385,
      "learning_rate": 4.5562995274820283e-07,
      "logits/chosen": -2.799257278442383,
      "logits/rejected": -2.7461183071136475,
      "logps/chosen": -294.0331726074219,
      "logps/rejected": -290.0491638183594,
      "loss": 0.6927,
      "positive_losses": 0.0,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.0151880057528615,
      "rewards/margins": 1.8423608707962558e-05,
      "rewards/margins_max": 0.006872598081827164,
      "rewards/margins_min": -0.007915332913398743,
      "rewards/margins_std": 0.006555147469043732,
      "rewards/rejected": 0.015169581398367882,
      "step": 90
    },
    {
      "dpo_losses": 0.6915194988250732,
      "epoch": 0.3,
      "grad_norm": 1.7021064862560185,
      "learning_rate": 4.394104893853007e-07,
      "logits/chosen": -2.8969879150390625,
      "logits/rejected": -2.8577048778533936,
      "logps/chosen": -271.9718017578125,
      "logps/rejected": -255.94015502929688,
      "loss": 0.6925,
      "positive_losses": 0.0063987732864916325,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.01490433793514967,
      "rewards/margins": 0.0032706786878407,
      "rewards/margins_max": 0.01071108691394329,
      "rewards/margins_min": -0.004061442334204912,
      "rewards/margins_std": 0.006633488927036524,
      "rewards/rejected": 0.011633659712970257,
      "step": 100
    },
    {
      "epoch": 0.3,
      "eval_dpo_losses": 0.6918572783470154,
      "eval_logits/chosen": -2.8058130741119385,
      "eval_logits/rejected": -2.7668540477752686,
      "eval_logps/chosen": -282.8757629394531,
      "eval_logps/rejected": -257.1208801269531,
      "eval_loss": 0.6926050782203674,
      "eval_positive_losses": 0.002994521986693144,
      "eval_rewards/accuracies": 0.6330000162124634,
      "eval_rewards/chosen": 0.017176620662212372,
      "eval_rewards/margins": 0.0025962977670133114,
      "eval_rewards/margins_max": 0.014017846435308456,
      "eval_rewards/margins_min": -0.007906544953584671,
      "eval_rewards/margins_std": 0.00730311032384634,
      "eval_rewards/rejected": 0.014580323360860348,
      "eval_runtime": 429.0754,
      "eval_samples_per_second": 4.661,
      "eval_steps_per_second": 0.291,
      "step": 100
    },
    {
      "dpo_losses": 0.6914526224136353,
      "epoch": 0.33,
      "grad_norm": 2.1403573418786492,
      "learning_rate": 4.2105939205932005e-07,
      "logits/chosen": -2.764038324356079,
      "logits/rejected": -2.7474615573883057,
      "logps/chosen": -310.06982421875,
      "logps/rejected": -234.0756072998047,
      "loss": 0.6917,
      "positive_losses": 0.0032207488548010588,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.016864223405718803,
      "rewards/margins": 0.0034028415102511644,
      "rewards/margins_max": 0.010497726500034332,
      "rewards/margins_min": -0.0030488395132124424,
      "rewards/margins_std": 0.006098336540162563,
      "rewards/rejected": 0.013461383990943432,
      "step": 110
    },
    {
      "dpo_losses": 0.691475510597229,
      "epoch": 0.36,
      "grad_norm": 2.00482561166422,
      "learning_rate": 4.0078318482522114e-07,
      "logits/chosen": -2.752159833908081,
      "logits/rejected": -2.750898838043213,
      "logps/chosen": -321.77813720703125,
      "logps/rejected": -272.95037841796875,
      "loss": 0.6922,
      "positive_losses": 0.011810302734375,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.017863672226667404,
      "rewards/margins": 0.003372038947418332,
      "rewards/margins_max": 0.014931520447134972,
      "rewards/margins_min": -0.006266787648200989,
      "rewards/margins_std": 0.009545858018100262,
      "rewards/rejected": 0.01449163444340229,
      "step": 120
    },
    {
      "dpo_losses": 0.692450225353241,
      "epoch": 0.4,
      "grad_norm": 1.6365937615424984,
      "learning_rate": 3.7881005700938627e-07,
      "logits/chosen": -2.8213768005371094,
      "logits/rejected": -2.831343650817871,
      "logps/chosen": -264.4930114746094,
      "logps/rejected": -232.359130859375,
      "loss": 0.6926,
      "positive_losses": 0.008326339535415173,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.01900574564933777,
      "rewards/margins": 0.001420500222593546,
      "rewards/margins_max": 0.011069941334426403,
      "rewards/margins_min": -0.009186895564198494,
      "rewards/margins_std": 0.008960539475083351,
      "rewards/rejected": 0.01758524402976036,
      "step": 130
    },
    {
      "dpo_losses": 0.690600574016571,
      "epoch": 0.43,
      "grad_norm": 1.9064212159356546,
      "learning_rate": 3.5538729515692354e-07,
      "logits/chosen": -2.7800724506378174,
      "logits/rejected": -2.763850450515747,
      "logps/chosen": -292.2286682128906,
      "logps/rejected": -268.72857666015625,
      "loss": 0.6912,
      "positive_losses": 0.013485336676239967,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.02164594829082489,
      "rewards/margins": 0.005139264743775129,
      "rewards/margins_max": 0.017013436183333397,
      "rewards/margins_min": -0.0060675484128296375,
      "rewards/margins_std": 0.010294691659510136,
      "rewards/rejected": 0.016506681218743324,
      "step": 140
    },
    {
      "dpo_losses": 0.6907236576080322,
      "epoch": 0.46,
      "grad_norm": 1.4378221399671829,
      "learning_rate": 3.3077850005803125e-07,
      "logits/chosen": -2.842320203781128,
      "logits/rejected": -2.8206865787506104,
      "logps/chosen": -268.4656066894531,
      "logps/rejected": -243.43276977539062,
      "loss": 0.6925,
      "positive_losses": 0.0013946533435955644,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.021758494898676872,
      "rewards/margins": 0.004882398992776871,
      "rewards/margins_max": 0.017139647156000137,
      "rewards/margins_min": -0.0063543543219566345,
      "rewards/margins_std": 0.010589024983346462,
      "rewards/rejected": 0.01687609776854515,
      "step": 150
    },
    {
      "dpo_losses": 0.6922906041145325,
      "epoch": 0.49,
      "grad_norm": 2.1178244445563013,
      "learning_rate": 3.0526062017313247e-07,
      "logits/chosen": -2.7998666763305664,
      "logits/rejected": -2.7823493480682373,
      "logps/chosen": -253.3588104248047,
      "logps/rejected": -238.626708984375,
      "loss": 0.6933,
      "positive_losses": 0.0011047363514080644,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.021453257650136948,
      "rewards/margins": 0.001750586787238717,
      "rewards/margins_max": 0.014365032315254211,
      "rewards/margins_min": -0.009759698994457722,
      "rewards/margins_std": 0.010657011531293392,
      "rewards/rejected": 0.019702669233083725,
      "step": 160
    },
    {
      "dpo_losses": 0.6903497576713562,
      "epoch": 0.52,
      "grad_norm": 1.5792742029110545,
      "learning_rate": 2.791208348427426e-07,
      "logits/chosen": -2.817678451538086,
      "logits/rejected": -2.7354776859283447,
      "logps/chosen": -301.37139892578125,
      "logps/rejected": -271.1383972167969,
      "loss": 0.6907,
      "positive_losses": 0.0042434693314135075,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.02347148209810257,
      "rewards/margins": 0.0056489938870072365,
      "rewards/margins_max": 0.017104174941778183,
      "rewards/margins_min": -0.0068167769350111485,
      "rewards/margins_std": 0.010564757511019707,
      "rewards/rejected": 0.01782248727977276,
      "step": 170
    },
    {
      "dpo_losses": 0.6911171674728394,
      "epoch": 0.55,
      "grad_norm": 11.74545654760677,
      "learning_rate": 2.526533223585641e-07,
      "logits/chosen": -2.8419086933135986,
      "logits/rejected": -2.7769436836242676,
      "logps/chosen": -253.8654022216797,
      "logps/rejected": -227.02719116210938,
      "loss": 0.6917,
      "positive_losses": 0.0033275603782385588,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.02268325164914131,
      "rewards/margins": 0.0040952288545668125,
      "rewards/margins_max": 0.016258299350738525,
      "rewards/margins_min": -0.004967542830854654,
      "rewards/margins_std": 0.00928699504584074,
      "rewards/rejected": 0.018588021397590637,
      "step": 180
    },
    {
      "dpo_losses": 0.6901928186416626,
      "epoch": 0.58,
      "grad_norm": 9.4811716718965,
      "learning_rate": 2.261559492680755e-07,
      "logits/chosen": -2.784456729888916,
      "logits/rejected": -2.766946792602539,
      "logps/chosen": -298.0020446777344,
      "logps/rejected": -268.62164306640625,
      "loss": 0.6923,
      "positive_losses": 0.0034267425071448088,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.025788750499486923,
      "rewards/margins": 0.005970729514956474,
      "rewards/margins_max": 0.02267562597990036,
      "rewards/margins_min": -0.008256749249994755,
      "rewards/margins_std": 0.013814790174365044,
      "rewards/rejected": 0.01981802098453045,
      "step": 190
    },
    {
      "dpo_losses": 0.6899710893630981,
      "epoch": 0.61,
      "grad_norm": 2.1557416454862444,
      "learning_rate": 1.9992691817133024e-07,
      "logits/chosen": -2.787891387939453,
      "logits/rejected": -2.756988048553467,
      "logps/chosen": -279.0212097167969,
      "logps/rejected": -286.3407287597656,
      "loss": 0.6928,
      "positive_losses": 0.02213897742331028,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.025793900713324547,
      "rewards/margins": 0.006411595735698938,
      "rewards/margins_max": 0.022542249411344528,
      "rewards/margins_min": -0.00852292776107788,
      "rewards/margins_std": 0.013604698702692986,
      "rewards/rejected": 0.019382307305932045,
      "step": 200
    },
    {
      "epoch": 0.61,
      "eval_dpo_losses": 0.6903740763664246,
      "eval_logits/chosen": -2.8064374923706055,
      "eval_logits/rejected": -2.7678143978118896,
      "eval_logps/chosen": -282.0021667480469,
      "eval_logps/rejected": -256.5478515625,
      "eval_loss": 0.6918678879737854,
      "eval_positive_losses": 0.008644149638712406,
      "eval_rewards/accuracies": 0.6470000147819519,
      "eval_rewards/chosen": 0.025912432000041008,
      "eval_rewards/margins": 0.005601577460765839,
      "eval_rewards/margins_max": 0.02654200792312622,
      "eval_rewards/margins_min": -0.01304223295301199,
      "eval_rewards/margins_std": 0.013095145113766193,
      "eval_rewards/rejected": 0.02031085453927517,
      "eval_runtime": 428.4364,
      "eval_samples_per_second": 4.668,
      "eval_steps_per_second": 0.292,
      "step": 200
    },
    {
      "dpo_losses": 0.6895982623100281,
      "epoch": 0.64,
      "grad_norm": 9.2776467878139,
      "learning_rate": 1.742614117358029e-07,
      "logits/chosen": -2.8029887676239014,
      "logits/rejected": -2.7589237689971924,
      "logps/chosen": -302.7414855957031,
      "logps/rejected": -286.43438720703125,
      "loss": 0.6912,
      "positive_losses": 0.013084793463349342,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.02575690671801567,
      "rewards/margins": 0.007164796348661184,
      "rewards/margins_max": 0.023669250309467316,
      "rewards/margins_min": -0.007106063421815634,
      "rewards/margins_std": 0.014227842912077904,
      "rewards/rejected": 0.0185921099036932,
      "step": 210
    },
    {
      "dpo_losses": 0.6903053522109985,
      "epoch": 0.67,
      "grad_norm": 1.955123489265158,
      "learning_rate": 1.4944827069769122e-07,
      "logits/chosen": -2.8536009788513184,
      "logits/rejected": -2.8277313709259033,
      "logps/chosen": -310.047607421875,
      "logps/rejected": -264.05352783203125,
      "loss": 0.6902,
      "positive_losses": 0.002166366670280695,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.02925548516213894,
      "rewards/margins": 0.005735241807997227,
      "rewards/margins_max": 0.02009030058979988,
      "rewards/margins_min": -0.007059932686388493,
      "rewards/margins_std": 0.01263638585805893,
      "rewards/rejected": 0.023520244285464287,
      "step": 220
    },
    {
      "dpo_losses": 0.6920837163925171,
      "epoch": 0.7,
      "grad_norm": 8.688484410559708,
      "learning_rate": 1.2576674323558928e-07,
      "logits/chosen": -2.822300910949707,
      "logits/rejected": -2.8432376384735107,
      "logps/chosen": -286.09466552734375,
      "logps/rejected": -260.4286804199219,
      "loss": 0.6922,
      "positive_losses": 0.0013641357654705644,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.025508206337690353,
      "rewards/margins": 0.002203943906351924,
      "rewards/margins_max": 0.020830942317843437,
      "rewards/margins_min": -0.015516755171120167,
      "rewards/margins_std": 0.01637136936187744,
      "rewards/rejected": 0.02330426312983036,
      "step": 230
    },
    {
      "dpo_losses": 0.690080463886261,
      "epoch": 0.73,
      "grad_norm": 2.0416140906382805,
      "learning_rate": 1.0348334229922676e-07,
      "logits/chosen": -2.8787901401519775,
      "logits/rejected": -2.831031322479248,
      "logps/chosen": -288.3709411621094,
      "logps/rejected": -273.1520080566406,
      "loss": 0.6918,
      "positive_losses": 0.01643219031393528,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.02646791562438011,
      "rewards/margins": 0.00618840241804719,
      "rewards/margins_max": 0.022397857159376144,
      "rewards/margins_min": -0.006807761732488871,
      "rewards/margins_std": 0.013137358240783215,
      "rewards/rejected": 0.02027951553463936,
      "step": 240
    },
    {
      "dpo_losses": 0.6899579763412476,
      "epoch": 0.76,
      "grad_norm": 1.9842983441302535,
      "learning_rate": 8.284884626103164e-08,
      "logits/chosen": -2.8189456462860107,
      "logits/rejected": -2.787537097930908,
      "logps/chosen": -298.2633056640625,
      "logps/rejected": -302.3707580566406,
      "loss": 0.6937,
      "positive_losses": 0.0013935088645666838,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.02826567366719246,
      "rewards/margins": 0.006455945782363415,
      "rewards/margins_max": 0.022190529853105545,
      "rewards/margins_min": -0.010183097794651985,
      "rewards/margins_std": 0.014948679134249687,
      "rewards/rejected": 0.02180972695350647,
      "step": 250
    },
    {
      "dpo_losses": 0.6888389587402344,
      "epoch": 0.79,
      "grad_norm": 2.1130940747529685,
      "learning_rate": 6.409547664531733e-08,
      "logits/chosen": -2.8461780548095703,
      "logits/rejected": -2.8126139640808105,
      "logps/chosen": -330.7644958496094,
      "logps/rejected": -310.1560974121094,
      "loss": 0.6907,
      "positive_losses": 0.0,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.03217402845621109,
      "rewards/margins": 0.008677511475980282,
      "rewards/margins_max": 0.021433433517813683,
      "rewards/margins_min": -0.0022177654318511486,
      "rewards/margins_std": 0.010701724328100681,
      "rewards/rejected": 0.023496516048908234,
      "step": 260
    },
    {
      "dpo_losses": 0.6895651817321777,
      "epoch": 0.82,
      "grad_norm": 1.9914333935784974,
      "learning_rate": 4.743428469705335e-08,
      "logits/chosen": -2.7977194786071777,
      "logits/rejected": -2.792222499847412,
      "logps/chosen": -301.0746154785156,
      "logps/rejected": -305.9708557128906,
      "loss": 0.6919,
      "positive_losses": 0.020720671862363815,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.027332518249750137,
      "rewards/margins": 0.007225664798170328,
      "rewards/margins_max": 0.021714068949222565,
      "rewards/margins_min": -0.0065118856728076935,
      "rewards/margins_std": 0.012981800362467766,
      "rewards/rejected": 0.020106855779886246,
      "step": 270
    },
    {
      "dpo_losses": 0.6888093948364258,
      "epoch": 0.85,
      "grad_norm": 2.1081832344794846,
      "learning_rate": 3.305277620188826e-08,
      "logits/chosen": -2.8479342460632324,
      "logits/rejected": -2.829345226287842,
      "logps/chosen": -322.39654541015625,
      "logps/rejected": -267.4809875488281,
      "loss": 0.6907,
      "positive_losses": 0.00835342425853014,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.03169160336256027,
      "rewards/margins": 0.008761906065046787,
      "rewards/margins_max": 0.025315579026937485,
      "rewards/margins_min": -0.00724324444308877,
      "rewards/margins_std": 0.014560697600245476,
      "rewards/rejected": 0.02292969450354576,
      "step": 280
    },
    {
      "dpo_losses": 0.690091609954834,
      "epoch": 0.88,
      "grad_norm": 1.7109015474069318,
      "learning_rate": 2.1112801287806375e-08,
      "logits/chosen": -2.7887511253356934,
      "logits/rejected": -2.753249168395996,
      "logps/chosen": -271.52618408203125,
      "logps/rejected": -243.4734344482422,
      "loss": 0.691,
      "positive_losses": 0.0007474899175576866,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.02663729526102543,
      "rewards/margins": 0.0061707692220807076,
      "rewards/margins_max": 0.020485369488596916,
      "rewards/margins_min": -0.006647027097642422,
      "rewards/margins_std": 0.012255651876330376,
      "rewards/rejected": 0.020466525107622147,
      "step": 290
    },
    {
      "dpo_losses": 0.6898671984672546,
      "epoch": 0.91,
      "grad_norm": 6.914969485002492,
      "learning_rate": 1.1748732956682023e-08,
      "logits/chosen": -2.8814024925231934,
      "logits/rejected": -2.8128414154052734,
      "logps/chosen": -321.03570556640625,
      "logps/rejected": -283.5750732421875,
      "loss": 0.6916,
      "positive_losses": 0.020281601697206497,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.02677486464381218,
      "rewards/margins": 0.006634272634983063,
      "rewards/margins_max": 0.02378956601023674,
      "rewards/margins_min": -0.009901603683829308,
      "rewards/margins_std": 0.015127221122384071,
      "rewards/rejected": 0.020140592008829117,
      "step": 300
    },
    {
      "epoch": 0.91,
      "eval_dpo_losses": 0.6899807453155518,
      "eval_logits/chosen": -2.8024394512176514,
      "eval_logits/rejected": -2.763456344604492,
      "eval_logps/chosen": -281.7664489746094,
      "eval_logps/rejected": -256.39239501953125,
      "eval_loss": 0.6914138197898865,
      "eval_positive_losses": 0.006967452820390463,
      "eval_rewards/accuracies": 0.6629999876022339,
      "eval_rewards/chosen": 0.028269650414586067,
      "eval_rewards/margins": 0.0064042662270367146,
      "eval_rewards/margins_max": 0.030468182638287544,
      "eval_rewards/margins_min": -0.014824692159891129,
      "eval_rewards/margins_std": 0.014939413405954838,
      "eval_rewards/rejected": 0.02186538279056549,
      "eval_runtime": 428.1048,
      "eval_samples_per_second": 4.672,
      "eval_steps_per_second": 0.292,
      "step": 300
    },
    {
      "dpo_losses": 0.690555214881897,
      "epoch": 0.94,
      "grad_norm": 1.5197312877950733,
      "learning_rate": 5.065954844616721e-09,
      "logits/chosen": -2.827882766723633,
      "logits/rejected": -2.781325340270996,
      "logps/chosen": -274.24688720703125,
      "logps/rejected": -278.71307373046875,
      "loss": 0.6911,
      "positive_losses": 0.0,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.028748828917741776,
      "rewards/margins": 0.005246045999228954,
      "rewards/margins_max": 0.022478515282273293,
      "rewards/margins_min": -0.010068441741168499,
      "rewards/margins_std": 0.014408007264137268,
      "rewards/rejected": 0.023502781987190247,
      "step": 310
    },
    {
      "dpo_losses": 0.6897249817848206,
      "epoch": 0.97,
      "grad_norm": 1.9024323554426086,
      "learning_rate": 1.1396752298723499e-09,
      "logits/chosen": -2.869424819946289,
      "logits/rejected": -2.8175909519195557,
      "logps/chosen": -246.2608642578125,
      "logps/rejected": -255.5459442138672,
      "loss": 0.6898,
      "positive_losses": 0.0,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.026843314990401268,
      "rewards/margins": 0.006898718420416117,
      "rewards/margins_max": 0.020682070404291153,
      "rewards/margins_min": -0.0074739838019013405,
      "rewards/margins_std": 0.012545767240226269,
      "rewards/rejected": 0.01994459703564644,
      "step": 320
    },
    {
      "epoch": 1.0,
      "step": 329,
      "total_flos": 0.0,
      "train_loss": 0.6924228537771114,
      "train_runtime": 3896.0637,
      "train_samples_per_second": 1.351,
      "train_steps_per_second": 0.084
    }
  ],
  "logging_steps": 10,
  "max_steps": 329,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}