{
  "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": [
    {
      "epoch": 0.0,
      "grad_norm": 2.0906090021590473,
      "learning_rate": 1.5151515151515152e-08,
      "logits/chosen": -2.6820077896118164,
      "logits/rejected": -2.6930205821990967,
      "logps/chosen": -281.2528381347656,
      "logps/rejected": -258.0622253417969,
      "loss": 0.6931,
      "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
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.1561337868153565,
      "learning_rate": 1.5151515151515152e-07,
      "logits/chosen": -2.7683067321777344,
      "logits/rejected": -2.7538461685180664,
      "logps/chosen": -284.59912109375,
      "logps/rejected": -249.83580017089844,
      "loss": 0.6931,
      "rewards/accuracies": 0.3888888955116272,
      "rewards/chosen": 5.0317983550485224e-05,
      "rewards/margins": -0.00015015894314274192,
      "rewards/margins_max": 0.0020335663575679064,
      "rewards/margins_min": -0.0025187418796122074,
      "rewards/margins_std": 0.0020784963853657246,
      "rewards/rejected": 0.00020047693396918476,
      "step": 10
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9751920510122447,
      "learning_rate": 3.0303030303030305e-07,
      "logits/chosen": -2.8347439765930176,
      "logits/rejected": -2.7819018363952637,
      "logps/chosen": -291.50921630859375,
      "logps/rejected": -270.4449768066406,
      "loss": 0.693,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.00022530628484673798,
      "rewards/margins": 0.0007139825029298663,
      "rewards/margins_max": 0.004345210734754801,
      "rewards/margins_min": -0.002943573985248804,
      "rewards/margins_std": 0.0032876902259886265,
      "rewards/rejected": -0.0004886762471869588,
      "step": 20
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.6654288543237405,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.8627753257751465,
      "logits/rejected": -2.8151745796203613,
      "logps/chosen": -259.2825927734375,
      "logps/rejected": -227.37350463867188,
      "loss": 0.6932,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.00037297833478078246,
      "rewards/margins": -0.00020202626183163375,
      "rewards/margins_max": 0.0030375297646969557,
      "rewards/margins_min": -0.0033374775666743517,
      "rewards/margins_std": 0.0028792533557862043,
      "rewards/rejected": -0.00017095205839723349,
      "step": 30
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.6057052994928989,
      "learning_rate": 4.993103596812268e-07,
      "logits/chosen": -2.8291430473327637,
      "logits/rejected": -2.7638001441955566,
      "logps/chosen": -317.513916015625,
      "logps/rejected": -224.7698211669922,
      "loss": 0.6927,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.000282805209280923,
      "rewards/margins": 0.0011336280731484294,
      "rewards/margins_max": 0.005270844791084528,
      "rewards/margins_min": -0.002321633044630289,
      "rewards/margins_std": 0.003387246746569872,
      "rewards/rejected": -0.000850822776556015,
      "step": 40
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.75475734119915,
      "learning_rate": 4.959416858332709e-07,
      "logits/chosen": -2.79063081741333,
      "logits/rejected": -2.804368495941162,
      "logps/chosen": -242.9667510986328,
      "logps/rejected": -280.0011901855469,
      "loss": 0.6926,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.00031891773687675595,
      "rewards/margins": 0.0008532041683793068,
      "rewards/margins_max": 0.004698004573583603,
      "rewards/margins_min": -0.002724443329498172,
      "rewards/margins_std": 0.00329922279343009,
      "rewards/rejected": -0.0011721218470484018,
      "step": 50
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.91854731579599,
      "learning_rate": 4.898051734555674e-07,
      "logits/chosen": -2.8335373401641846,
      "logits/rejected": -2.8440303802490234,
      "logps/chosen": -321.90625,
      "logps/rejected": -283.37994384765625,
      "loss": 0.6921,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.00014833270688541234,
      "rewards/margins": 0.0021868678741157055,
      "rewards/margins_max": 0.008168894797563553,
      "rewards/margins_min": -0.0031033740378916264,
      "rewards/margins_std": 0.005018442869186401,
      "rewards/rejected": -0.0020385351963341236,
      "step": 60
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.5964843630528198,
      "learning_rate": 4.809698831278217e-07,
      "logits/chosen": -2.748741865158081,
      "logits/rejected": -2.735199213027954,
      "logps/chosen": -266.52606201171875,
      "logps/rejected": -246.6175079345703,
      "loss": 0.6922,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.0004858696775045246,
      "rewards/margins": 0.0019440820906311274,
      "rewards/margins_max": 0.008044283837080002,
      "rewards/margins_min": -0.003816543845459819,
      "rewards/margins_std": 0.005177702754735947,
      "rewards/rejected": -0.002429951447993517,
      "step": 70
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.1014040068240663,
      "learning_rate": 4.6953524759527053e-07,
      "logits/chosen": -2.8426356315612793,
      "logits/rejected": -2.8158562183380127,
      "logps/chosen": -282.353515625,
      "logps/rejected": -275.220458984375,
      "loss": 0.6918,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.000800526118837297,
      "rewards/margins": 0.0022392510436475277,
      "rewards/margins_max": 0.00997895933687687,
      "rewards/margins_min": -0.005192113574594259,
      "rewards/margins_std": 0.0067059798166155815,
      "rewards/rejected": -0.0030397772789001465,
      "step": 80
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.9914287244112958,
      "learning_rate": 4.5562995274820283e-07,
      "logits/chosen": -2.7992029190063477,
      "logits/rejected": -2.746138095855713,
      "logps/chosen": -295.78399658203125,
      "logps/rejected": -291.9333190917969,
      "loss": 0.6919,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.002320217899978161,
      "rewards/margins": 0.001351921702735126,
      "rewards/margins_max": 0.010480575263500214,
      "rewards/margins_min": -0.009322223253548145,
      "rewards/margins_std": 0.008863108232617378,
      "rewards/rejected": -0.0036721397191286087,
      "step": 90
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6705180570693485,
      "learning_rate": 4.394104893853007e-07,
      "logits/chosen": -2.896794557571411,
      "logits/rejected": -2.85756254196167,
      "logps/chosen": -273.5906982421875,
      "logps/rejected": -257.73284912109375,
      "loss": 0.6914,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.0012845676392316818,
      "rewards/margins": 0.005008908919990063,
      "rewards/margins_max": 0.013928805477917194,
      "rewards/margins_min": -0.003106380347162485,
      "rewards/margins_std": 0.007625125348567963,
      "rewards/rejected": -0.006293477024883032,
      "step": 100
    },
    {
      "epoch": 0.3,
      "eval_logits/chosen": -2.806475877761841,
      "eval_logits/rejected": -2.767702102661133,
      "eval_logps/chosen": -284.6319274902344,
      "eval_logps/rejected": -258.9901123046875,
      "eval_loss": 0.691525399684906,
      "eval_rewards/accuracies": 0.6460000276565552,
      "eval_rewards/chosen": -0.00038527295691892505,
      "eval_rewards/margins": 0.003726556431502104,
      "eval_rewards/margins_max": 0.01873905211687088,
      "eval_rewards/margins_min": -0.009459242224693298,
      "eval_rewards/margins_std": 0.00931489747017622,
      "eval_rewards/rejected": -0.004111829213798046,
      "eval_runtime": 428.4684,
      "eval_samples_per_second": 4.668,
      "eval_steps_per_second": 0.292,
      "step": 100
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.1453641236519685,
      "learning_rate": 4.2105939205932005e-07,
      "logits/chosen": -2.7631096839904785,
      "logits/rejected": -2.746663808822632,
      "logps/chosen": -311.8393249511719,
      "logps/rejected": -235.84280395507812,
      "loss": 0.6911,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.0008311712299473584,
      "rewards/margins": 0.0033794320188462734,
      "rewards/margins_max": 0.013278109021484852,
      "rewards/margins_min": -0.00541637372225523,
      "rewards/margins_std": 0.008299448527395725,
      "rewards/rejected": -0.0042106034234166145,
      "step": 110
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.024896986425123,
      "learning_rate": 4.0078318482522114e-07,
      "logits/chosen": -2.7521708011627197,
      "logits/rejected": -2.750868082046509,
      "logps/chosen": -323.51666259765625,
      "logps/rejected": -274.75970458984375,
      "loss": 0.6909,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.0004785500350408256,
      "rewards/margins": 0.004080395679920912,
      "rewards/margins_max": 0.015328818932175636,
      "rewards/margins_min": -0.0073760440573096275,
      "rewards/margins_std": 0.00990099273622036,
      "rewards/rejected": -0.0036018460523337126,
      "step": 120
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6346525930252072,
      "learning_rate": 3.7881005700938627e-07,
      "logits/chosen": -2.8206729888916016,
      "logits/rejected": -2.8308663368225098,
      "logps/chosen": -266.37469482421875,
      "logps/rejected": -234.52035522460938,
      "loss": 0.6906,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.00018907712365034968,
      "rewards/margins": 0.00421659741550684,
      "rewards/margins_max": 0.015676384791731834,
      "rewards/margins_min": -0.007556927390396595,
      "rewards/margins_std": 0.010246575810015202,
      "rewards/rejected": -0.004027520306408405,
      "step": 130
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.9044203185149946,
      "learning_rate": 3.5538729515692354e-07,
      "logits/chosen": -2.780360460281372,
      "logits/rejected": -2.7639012336730957,
      "logps/chosen": -294.11309814453125,
      "logps/rejected": -270.84710693359375,
      "loss": 0.6896,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.0028018890880048275,
      "rewards/margins": 0.007480897009372711,
      "rewards/margins_max": 0.021374408155679703,
      "rewards/margins_min": -0.0061719887889921665,
      "rewards/margins_std": 0.01222699973732233,
      "rewards/rejected": -0.004679008387029171,
      "step": 140
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4256559970133287,
      "learning_rate": 3.3077850005803125e-07,
      "logits/chosen": -2.8410263061523438,
      "logits/rejected": -2.8195314407348633,
      "logps/chosen": -270.49615478515625,
      "logps/rejected": -245.65200805664062,
      "loss": 0.6903,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0014524383004754782,
      "rewards/margins": 0.006768654100596905,
      "rewards/margins_max": 0.025039460510015488,
      "rewards/margins_min": -0.01076546311378479,
      "rewards/margins_std": 0.015858832746744156,
      "rewards/rejected": -0.0053162164986133575,
      "step": 150
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.1265057109843077,
      "learning_rate": 3.0526062017313247e-07,
      "logits/chosen": -2.79884672164917,
      "logits/rejected": -2.7815585136413574,
      "logps/chosen": -255.3964080810547,
      "logps/rejected": -241.00271606445312,
      "loss": 0.6909,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.0010771710658445954,
      "rewards/margins": 0.005134746432304382,
      "rewards/margins_max": 0.022996146231889725,
      "rewards/margins_min": -0.009730304591357708,
      "rewards/margins_std": 0.014861812815070152,
      "rewards/rejected": -0.004057575948536396,
      "step": 160
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.59020242230242,
      "learning_rate": 2.791208348427426e-07,
      "logits/chosen": -2.814671039581299,
      "logits/rejected": -2.732504367828369,
      "logps/chosen": -303.4354553222656,
      "logps/rejected": -273.4683837890625,
      "loss": 0.6887,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.002831272780895233,
      "rewards/margins": 0.008308259770274162,
      "rewards/margins_max": 0.02398056350648403,
      "rewards/margins_min": -0.007372391410171986,
      "rewards/margins_std": 0.014078010804951191,
      "rewards/rejected": -0.005476987920701504,
      "step": 170
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.791424187804565,
      "learning_rate": 2.526533223585641e-07,
      "logits/chosen": -2.8398988246917725,
      "logits/rejected": -2.775310754776001,
      "logps/chosen": -256.0347595214844,
      "logps/rejected": -229.332763671875,
      "loss": 0.6897,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.0009899451397359371,
      "rewards/margins": 0.005457731895148754,
      "rewards/margins_max": 0.021505217999219894,
      "rewards/margins_min": -0.008436702191829681,
      "rewards/margins_std": 0.013385000638663769,
      "rewards/rejected": -0.00446778628975153,
      "step": 180
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.7305988753672774,
      "learning_rate": 2.261559492680755e-07,
      "logits/chosen": -2.781790256500244,
      "logits/rejected": -2.7643322944641113,
      "logps/chosen": -300.09393310546875,
      "logps/rejected": -271.13116455078125,
      "loss": 0.6891,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.004870180506259203,
      "rewards/margins": 0.0101470947265625,
      "rewards/margins_max": 0.03561341017484665,
      "rewards/margins_min": -0.00919102318584919,
      "rewards/margins_std": 0.019990354776382446,
      "rewards/rejected": -0.005276912357658148,
      "step": 190
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.169958133205736,
      "learning_rate": 1.9992691817133024e-07,
      "logits/chosen": -2.7859396934509277,
      "logits/rejected": -2.755178213119507,
      "logps/chosen": -281.18170166015625,
      "logps/rejected": -288.84930419921875,
      "loss": 0.6884,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0041890377178788185,
      "rewards/margins": 0.009892629459500313,
      "rewards/margins_max": 0.03310906141996384,
      "rewards/margins_min": -0.012689237482845783,
      "rewards/margins_std": 0.02011021040380001,
      "rewards/rejected": -0.005703592207282782,
      "step": 200
    },
    {
      "epoch": 0.61,
      "eval_logits/chosen": -2.804927349090576,
      "eval_logits/rejected": -2.766470432281494,
      "eval_logps/chosen": -284.3610534667969,
      "eval_logps/rejected": -259.1879577636719,
      "eval_loss": 0.6895014643669128,
      "eval_rewards/accuracies": 0.6850000023841858,
      "eval_rewards/chosen": 0.0023234861437231302,
      "eval_rewards/margins": 0.008414038456976414,
      "eval_rewards/margins_max": 0.038945525884628296,
      "eval_rewards/margins_min": -0.018883490934967995,
      "eval_rewards/margins_std": 0.018971558660268784,
      "eval_rewards/rejected": -0.006090551149100065,
      "eval_runtime": 427.7798,
      "eval_samples_per_second": 4.675,
      "eval_steps_per_second": 0.292,
      "step": 200
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.9906194761669704,
      "learning_rate": 1.742614117358029e-07,
      "logits/chosen": -2.80131196975708,
      "logits/rejected": -2.7576537132263184,
      "logps/chosen": -304.849853515625,
      "logps/rejected": -289.08197021484375,
      "loss": 0.6877,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0046735843643546104,
      "rewards/margins": 0.012557747773826122,
      "rewards/margins_max": 0.03481978923082352,
      "rewards/margins_min": -0.00802917592227459,
      "rewards/margins_std": 0.019201457500457764,
      "rewards/rejected": -0.007884165272116661,
      "step": 210
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.9658311065665528,
      "learning_rate": 1.4944827069769122e-07,
      "logits/chosen": -2.851292133331299,
      "logits/rejected": -2.8257217407226562,
      "logps/chosen": -312.4863586425781,
      "logps/rejected": -266.73626708984375,
      "loss": 0.6891,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.004867873154580593,
      "rewards/margins": 0.008174732327461243,
      "rewards/margins_max": 0.028449540957808495,
      "rewards/margins_min": -0.011079356074333191,
      "rewards/margins_std": 0.01735488697886467,
      "rewards/rejected": -0.003306858241558075,
      "step": 220
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8987994692738805,
      "learning_rate": 1.2576674323558928e-07,
      "logits/chosen": -2.821254014968872,
      "logits/rejected": -2.8421223163604736,
      "logps/chosen": -288.6875,
      "logps/rejected": -263.0277099609375,
      "loss": 0.6906,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.00041991579928435385,
      "rewards/margins": 0.0022654212079942226,
      "rewards/margins_max": 0.024668732658028603,
      "rewards/margins_min": -0.022191215306520462,
      "rewards/margins_std": 0.020731808617711067,
      "rewards/rejected": -0.002685337094590068,
      "step": 230
    },
    {
      "epoch": 0.73,
      "grad_norm": 2.049682090113544,
      "learning_rate": 1.0348334229922676e-07,
      "logits/chosen": -2.877260684967041,
      "logits/rejected": -2.8300554752349854,
      "logps/chosen": -290.80633544921875,
      "logps/rejected": -275.846435546875,
      "loss": 0.6893,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.0021143355406820774,
      "rewards/margins": 0.00877899769693613,
      "rewards/margins_max": 0.03138250857591629,
      "rewards/margins_min": -0.01147426012903452,
      "rewards/margins_std": 0.019380424171686172,
      "rewards/rejected": -0.006664662156254053,
      "step": 240
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.008481756505904,
      "learning_rate": 8.284884626103164e-08,
      "logits/chosen": -2.817871570587158,
      "logits/rejected": -2.786424398422241,
      "logps/chosen": -300.6135559082031,
      "logps/rejected": -305.0606994628906,
      "loss": 0.6882,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.0047633713111281395,
      "rewards/margins": 0.009853017516434193,
      "rewards/margins_max": 0.034555986523628235,
      "rewards/margins_min": -0.011890431866049767,
      "rewards/margins_std": 0.020635981112718582,
      "rewards/rejected": -0.0050896452739834785,
      "step": 250
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.120277542804363,
      "learning_rate": 6.409547664531733e-08,
      "logits/chosen": -2.844655752182007,
      "logits/rejected": -2.811575412750244,
      "logps/chosen": -333.072265625,
      "logps/rejected": -312.94317626953125,
      "loss": 0.6874,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.009096643887460232,
      "rewards/margins": 0.013471168465912342,
      "rewards/margins_max": 0.0355917289853096,
      "rewards/margins_min": -0.005764602217823267,
      "rewards/margins_std": 0.018313560634851456,
      "rewards/rejected": -0.00437452457845211,
      "step": 260
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.015529778175616,
      "learning_rate": 4.743428469705335e-08,
      "logits/chosen": -2.7949509620666504,
      "logits/rejected": -2.7894396781921387,
      "logps/chosen": -303.4598693847656,
      "logps/rejected": -308.66522216796875,
      "loss": 0.6889,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.003480118466541171,
      "rewards/margins": 0.010316994972527027,
      "rewards/margins_max": 0.033331625163555145,
      "rewards/margins_min": -0.010707234963774681,
      "rewards/margins_std": 0.01953895017504692,
      "rewards/rejected": -0.006836875341832638,
      "step": 270
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.1162209644793024,
      "learning_rate": 3.305277620188826e-08,
      "logits/chosen": -2.844252347946167,
      "logits/rejected": -2.8254075050354004,
      "logps/chosen": -324.8486633300781,
      "logps/rejected": -270.613037109375,
      "loss": 0.6865,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.0071704513393342495,
      "rewards/margins": 0.015561411157250404,
      "rewards/margins_max": 0.041363365948200226,
      "rewards/margins_min": -0.010495706461369991,
      "rewards/margins_std": 0.0231755543500185,
      "rewards/rejected": -0.008390960283577442,
      "step": 280
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7280929479679055,
      "learning_rate": 2.1112801287806375e-08,
      "logits/chosen": -2.783881187438965,
      "logits/rejected": -2.747999668121338,
      "logps/chosen": -273.90185546875,
      "logps/rejected": -246.3704833984375,
      "loss": 0.6881,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0028805662877857685,
      "rewards/margins": 0.011384439654648304,
      "rewards/margins_max": 0.036729536950588226,
      "rewards/margins_min": -0.009171558544039726,
      "rewards/margins_std": 0.021134525537490845,
      "rewards/rejected": -0.008503873832523823,
      "step": 290
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.8137366581326853,
      "learning_rate": 1.1748732956682023e-08,
      "logits/chosen": -2.878770351409912,
      "logits/rejected": -2.8104898929595947,
      "logps/chosen": -323.51312255859375,
      "logps/rejected": -286.44964599609375,
      "loss": 0.6873,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0020009407307952642,
      "rewards/margins": 0.010605795308947563,
      "rewards/margins_max": 0.03404298424720764,
      "rewards/margins_min": -0.010880110785365105,
      "rewards/margins_std": 0.020114842802286148,
      "rewards/rejected": -0.008604854345321655,
      "step": 300
    },
    {
      "epoch": 0.91,
      "eval_logits/chosen": -2.802642822265625,
      "eval_logits/rejected": -2.7640159130096436,
      "eval_logps/chosen": -284.2815246582031,
      "eval_logps/rejected": -259.24383544921875,
      "eval_loss": 0.6889453530311584,
      "eval_rewards/accuracies": 0.6759999990463257,
      "eval_rewards/chosen": 0.0031190679874271154,
      "eval_rewards/margins": 0.009768038988113403,
      "eval_rewards/margins_max": 0.044922519475221634,
      "eval_rewards/margins_min": -0.021590130403637886,
      "eval_rewards/margins_std": 0.021896740421652794,
      "eval_rewards/rejected": -0.006648970767855644,
      "eval_runtime": 427.9336,
      "eval_samples_per_second": 4.674,
      "eval_steps_per_second": 0.292,
      "step": 300
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.5476563917272619,
      "learning_rate": 5.065954844616721e-09,
      "logits/chosen": -2.8241655826568604,
      "logits/rejected": -2.7778286933898926,
      "logps/chosen": -276.5940856933594,
      "logps/rejected": -281.5748596191406,
      "loss": 0.6885,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.005276383366435766,
      "rewards/margins": 0.010391583666205406,
      "rewards/margins_max": 0.036186523735523224,
      "rewards/margins_min": -0.010774780064821243,
      "rewards/margins_std": 0.02108721435070038,
      "rewards/rejected": -0.005115201231092215,
      "step": 310
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.9217088208809332,
      "learning_rate": 1.1396752298723499e-09,
      "logits/chosen": -2.8640575408935547,
      "logits/rejected": -2.8119149208068848,
      "logps/chosen": -249.0362548828125,
      "logps/rejected": -258.521484375,
      "loss": 0.6879,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.0009104462224058807,
      "rewards/margins": 0.008900880813598633,
      "rewards/margins_max": 0.02946281060576439,
      "rewards/margins_min": -0.010788346640765667,
      "rewards/margins_std": 0.017393799498677254,
      "rewards/rejected": -0.009811325930058956,
      "step": 320
    },
    {
      "epoch": 1.0,
      "step": 329,
      "total_flos": 0.0,
      "train_loss": 0.6900796745323483,
      "train_runtime": 3893.3874,
      "train_samples_per_second": 1.352,
      "train_steps_per_second": 0.085
    }
  ],
  "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
}