{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.39952912638675847,
  "eval_steps": 200,
  "global_step": 1400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005707558948382264,
      "grad_norm": 7.28125,
      "learning_rate": 4.7499999999999995e-08,
      "log_odds_chosen": 0.023618780076503754,
      "log_odds_ratio": -0.6926277279853821,
      "logits/chosen": 1.2535090446472168,
      "logits/rejected": 1.2576510906219482,
      "logps/chosen": -0.19766047596931458,
      "logps/rejected": -0.20112021267414093,
      "loss": 0.2323,
      "nll_loss": 0.19766047596931458,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": -0.009883022867143154,
      "rewards/margins": 0.00017298731836490333,
      "rewards/rejected": -0.010056011378765106,
      "step": 20
    },
    {
      "epoch": 0.011415117896764528,
      "grad_norm": 7.40625,
      "learning_rate": 9.749999999999999e-08,
      "log_odds_chosen": 0.07222548872232437,
      "log_odds_ratio": -0.669521152973175,
      "logits/chosen": 1.2201169729232788,
      "logits/rejected": 1.2268095016479492,
      "logps/chosen": -0.1903044879436493,
      "logps/rejected": -0.205613374710083,
      "loss": 0.2238,
      "nll_loss": 0.1903044879436493,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.009515224024653435,
      "rewards/margins": 0.0007654451765120029,
      "rewards/rejected": -0.010280669666826725,
      "step": 40
    },
    {
      "epoch": 0.017122676845146793,
      "grad_norm": 8.0625,
      "learning_rate": 1.475e-07,
      "log_odds_chosen": 0.03168485313653946,
      "log_odds_ratio": -0.6886086463928223,
      "logits/chosen": 1.2384475469589233,
      "logits/rejected": 1.2428100109100342,
      "logps/chosen": -0.2004697322845459,
      "logps/rejected": -0.20637699961662292,
      "loss": 0.2349,
      "nll_loss": 0.2004697322845459,
      "rewards/accuracies": 0.515625,
      "rewards/chosen": -0.010023487731814384,
      "rewards/margins": 0.0002953629591502249,
      "rewards/rejected": -0.010318851098418236,
      "step": 60
    },
    {
      "epoch": 0.022830235793529056,
      "grad_norm": 7.125,
      "learning_rate": 1.975e-07,
      "log_odds_chosen": 0.051982104778289795,
      "log_odds_ratio": -0.6787282228469849,
      "logits/chosen": 1.2293108701705933,
      "logits/rejected": 1.2379376888275146,
      "logps/chosen": -0.2004239857196808,
      "logps/rejected": -0.21183133125305176,
      "loss": 0.2344,
      "nll_loss": 0.2004239857196808,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": -0.010021200403571129,
      "rewards/margins": 0.0005703666829504073,
      "rewards/rejected": -0.010591565631330013,
      "step": 80
    },
    {
      "epoch": 0.02853779474191132,
      "grad_norm": 7.25,
      "learning_rate": 2.475e-07,
      "log_odds_chosen": 0.032506946474313736,
      "log_odds_ratio": -0.6894897222518921,
      "logits/chosen": 1.2407578229904175,
      "logits/rejected": 1.2470386028289795,
      "logps/chosen": -0.19914038479328156,
      "logps/rejected": -0.20606501400470734,
      "loss": 0.2336,
      "nll_loss": 0.19914038479328156,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.009957019239664078,
      "rewards/margins": 0.000346232729498297,
      "rewards/rejected": -0.010303251445293427,
      "step": 100
    },
    {
      "epoch": 0.034245353690293585,
      "grad_norm": 8.1875,
      "learning_rate": 2.9749999999999996e-07,
      "log_odds_chosen": 0.04109399393200874,
      "log_odds_ratio": -0.6831611394882202,
      "logits/chosen": 1.2408243417739868,
      "logits/rejected": 1.2486960887908936,
      "logps/chosen": -0.19329828023910522,
      "logps/rejected": -0.2012052834033966,
      "loss": 0.2275,
      "nll_loss": 0.19329828023910522,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.009664914570748806,
      "rewards/margins": 0.00039534951793029904,
      "rewards/rejected": -0.010060264728963375,
      "step": 120
    },
    {
      "epoch": 0.03995291263867585,
      "grad_norm": 7.4375,
      "learning_rate": 3.4749999999999996e-07,
      "log_odds_chosen": 0.050192832946777344,
      "log_odds_ratio": -0.680444598197937,
      "logits/chosen": 1.2217073440551758,
      "logits/rejected": 1.22607421875,
      "logps/chosen": -0.19278880953788757,
      "logps/rejected": -0.2016700953245163,
      "loss": 0.2268,
      "nll_loss": 0.19278880953788757,
      "rewards/accuracies": 0.5718749761581421,
      "rewards/chosen": -0.009639441035687923,
      "rewards/margins": 0.00044406444067135453,
      "rewards/rejected": -0.010083505883812904,
      "step": 140
    },
    {
      "epoch": 0.04566047158705811,
      "grad_norm": 7.5625,
      "learning_rate": 3.975e-07,
      "log_odds_chosen": 0.05574320629239082,
      "log_odds_ratio": -0.679256796836853,
      "logits/chosen": 1.2088888883590698,
      "logits/rejected": 1.2116034030914307,
      "logps/chosen": -0.19427359104156494,
      "logps/rejected": -0.2055772989988327,
      "loss": 0.2282,
      "nll_loss": 0.19427359104156494,
      "rewards/accuracies": 0.5718749761581421,
      "rewards/chosen": -0.009713678620755672,
      "rewards/margins": 0.0005651867832057178,
      "rewards/rejected": -0.010278865694999695,
      "step": 160
    },
    {
      "epoch": 0.051368030535440375,
      "grad_norm": 7.65625,
      "learning_rate": 4.475e-07,
      "log_odds_chosen": 0.06304959952831268,
      "log_odds_ratio": -0.67559415102005,
      "logits/chosen": 1.2389880418777466,
      "logits/rejected": 1.247880458831787,
      "logps/chosen": -0.19238482415676117,
      "logps/rejected": -0.20644333958625793,
      "loss": 0.2262,
      "nll_loss": 0.19238482415676117,
      "rewards/accuracies": 0.5687500238418579,
      "rewards/chosen": -0.009619241580367088,
      "rewards/margins": 0.0007029252592474222,
      "rewards/rejected": -0.010322166606783867,
      "step": 180
    },
    {
      "epoch": 0.05707558948382264,
      "grad_norm": 7.1875,
      "learning_rate": 4.975e-07,
      "log_odds_chosen": -0.0036049254704266787,
      "log_odds_ratio": -0.7065373659133911,
      "logits/chosen": 1.2162582874298096,
      "logits/rejected": 1.2224491834640503,
      "logps/chosen": -0.1787000596523285,
      "logps/rejected": -0.179734006524086,
      "loss": 0.214,
      "nll_loss": 0.1787000596523285,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.00893500354140997,
      "rewards/margins": 5.1697123126359656e-05,
      "rewards/rejected": -0.008986701257526875,
      "step": 200
    },
    {
      "epoch": 0.06278314843220491,
      "grad_norm": 7.25,
      "learning_rate": 4.778184825674965e-07,
      "log_odds_chosen": 0.06815891712903976,
      "log_odds_ratio": -0.6725326776504517,
      "logits/chosen": 1.2341724634170532,
      "logits/rejected": 1.238364815711975,
      "logps/chosen": -0.17411001026630402,
      "logps/rejected": -0.1845506727695465,
      "loss": 0.2077,
      "nll_loss": 0.17410999536514282,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.008705498650670052,
      "rewards/margins": 0.000522034359164536,
      "rewards/rejected": -0.00922753382474184,
      "step": 220
    },
    {
      "epoch": 0.06849070738058717,
      "grad_norm": 6.78125,
      "learning_rate": 4.573893537463482e-07,
      "log_odds_chosen": 0.050056636333465576,
      "log_odds_ratio": -0.6814659237861633,
      "logits/chosen": 1.2188810110092163,
      "logits/rejected": 1.2238280773162842,
      "logps/chosen": -0.16769398748874664,
      "logps/rejected": -0.17724591493606567,
      "loss": 0.2018,
      "nll_loss": 0.16769398748874664,
      "rewards/accuracies": 0.596875011920929,
      "rewards/chosen": -0.008384699001908302,
      "rewards/margins": 0.0004775961278937757,
      "rewards/rejected": -0.008862295188009739,
      "step": 240
    },
    {
      "epoch": 0.07419826632896943,
      "grad_norm": 7.625,
      "learning_rate": 4.393747751637468e-07,
      "log_odds_chosen": 0.05101919174194336,
      "log_odds_ratio": -0.6815733313560486,
      "logits/chosen": 1.217728853225708,
      "logits/rejected": 1.2232797145843506,
      "logps/chosen": -0.16077013313770294,
      "logps/rejected": -0.1699983775615692,
      "loss": 0.1948,
      "nll_loss": 0.16077013313770294,
      "rewards/accuracies": 0.5687500238418579,
      "rewards/chosen": -0.008038505911827087,
      "rewards/margins": 0.00046141305938363075,
      "rewards/rejected": -0.00849991850554943,
      "step": 260
    },
    {
      "epoch": 0.0799058252773517,
      "grad_norm": 7.625,
      "learning_rate": 4.233337566673016e-07,
      "log_odds_chosen": 0.05960080772638321,
      "log_odds_ratio": -0.6763849258422852,
      "logits/chosen": 1.1978760957717896,
      "logits/rejected": 1.2018585205078125,
      "logps/chosen": -0.16003164649009705,
      "logps/rejected": -0.17044484615325928,
      "loss": 0.1939,
      "nll_loss": 0.16003163158893585,
      "rewards/accuracies": 0.5843750238418579,
      "rewards/chosen": -0.008001581765711308,
      "rewards/margins": 0.0005206596106290817,
      "rewards/rejected": -0.008522240445017815,
      "step": 280
    },
    {
      "epoch": 0.08561338422573396,
      "grad_norm": 7.6875,
      "learning_rate": 4.089304100547653e-07,
      "log_odds_chosen": 0.05554483085870743,
      "log_odds_ratio": -0.6802533864974976,
      "logits/chosen": 1.2009286880493164,
      "logits/rejected": 1.2058742046356201,
      "logps/chosen": -0.15855595469474792,
      "logps/rejected": -0.16927847266197205,
      "loss": 0.1926,
      "nll_loss": 0.15855596959590912,
      "rewards/accuracies": 0.5406249761581421,
      "rewards/chosen": -0.007927799597382545,
      "rewards/margins": 0.0005361254443414509,
      "rewards/rejected": -0.008463923819363117,
      "step": 300
    },
    {
      "epoch": 0.09132094317411622,
      "grad_norm": 7.15625,
      "learning_rate": 3.959037912324479e-07,
      "log_odds_chosen": 0.06532987207174301,
      "log_odds_ratio": -0.6775387525558472,
      "logits/chosen": 1.1810499429702759,
      "logits/rejected": 1.183556318283081,
      "logps/chosen": -0.1590016484260559,
      "logps/rejected": -0.16949968039989471,
      "loss": 0.1929,
      "nll_loss": 0.1590016484260559,
      "rewards/accuracies": 0.565625011920929,
      "rewards/chosen": -0.007950082421302795,
      "rewards/margins": 0.0005249014939181507,
      "rewards/rejected": -0.00847498420625925,
      "step": 320
    },
    {
      "epoch": 0.09702850212249849,
      "grad_norm": 7.78125,
      "learning_rate": 3.8404768632128423e-07,
      "log_odds_chosen": 0.09168948233127594,
      "log_odds_ratio": -0.6634136438369751,
      "logits/chosen": 1.1914002895355225,
      "logits/rejected": 1.200673222541809,
      "logps/chosen": -0.15473826229572296,
      "logps/rejected": -0.16917651891708374,
      "loss": 0.1879,
      "nll_loss": 0.15473827719688416,
      "rewards/accuracies": 0.621874988079071,
      "rewards/chosen": -0.007736912462860346,
      "rewards/margins": 0.0007219129474833608,
      "rewards/rejected": -0.008458825759589672,
      "step": 340
    },
    {
      "epoch": 0.10273606107088075,
      "grad_norm": 8.0,
      "learning_rate": 3.7319668543103794e-07,
      "log_odds_chosen": 0.061370015144348145,
      "log_odds_ratio": -0.6798322200775146,
      "logits/chosen": 1.1793339252471924,
      "logits/rejected": 1.1871938705444336,
      "logps/chosen": -0.15096887946128845,
      "logps/rejected": -0.16278162598609924,
      "loss": 0.185,
      "nll_loss": 0.15096887946128845,
      "rewards/accuracies": 0.543749988079071,
      "rewards/chosen": -0.007548443973064423,
      "rewards/margins": 0.000590637675486505,
      "rewards/rejected": -0.008139082230627537,
      "step": 360
    },
    {
      "epoch": 0.10844362001926301,
      "grad_norm": 7.8125,
      "learning_rate": 3.6321635614607396e-07,
      "log_odds_chosen": 0.05262409523129463,
      "log_odds_ratio": -0.6827314496040344,
      "logits/chosen": 1.1874498128890991,
      "logits/rejected": 1.1972339153289795,
      "logps/chosen": -0.15195965766906738,
      "logps/rejected": -0.16172973811626434,
      "loss": 0.1861,
      "nll_loss": 0.15195965766906738,
      "rewards/accuracies": 0.5218750238418579,
      "rewards/chosen": -0.007597983814775944,
      "rewards/margins": 0.0004885048256255686,
      "rewards/rejected": -0.008086487650871277,
      "step": 380
    },
    {
      "epoch": 0.11415117896764528,
      "grad_norm": 6.8125,
      "learning_rate": 3.539961627023943e-07,
      "log_odds_chosen": 0.10117591917514801,
      "log_odds_ratio": -0.6626842021942139,
      "logits/chosen": 1.2180874347686768,
      "logits/rejected": 1.224051833152771,
      "logps/chosen": -0.14667002856731415,
      "logps/rejected": -0.16229018568992615,
      "loss": 0.1798,
      "nll_loss": 0.14667002856731415,
      "rewards/accuracies": 0.596875011920929,
      "rewards/chosen": -0.007333502173423767,
      "rewards/margins": 0.0007810078095644712,
      "rewards/rejected": -0.008114509284496307,
      "step": 400
    },
    {
      "epoch": 0.11985873791602754,
      "grad_norm": 7.625,
      "learning_rate": 3.4544426792673337e-07,
      "log_odds_chosen": 0.07183325290679932,
      "log_odds_ratio": -0.676596999168396,
      "logits/chosen": 1.2141412496566772,
      "logits/rejected": 1.215499997138977,
      "logps/chosen": -0.14280927181243896,
      "logps/rejected": -0.1540253758430481,
      "loss": 0.1766,
      "nll_loss": 0.14280927181243896,
      "rewards/accuracies": 0.590624988079071,
      "rewards/chosen": -0.007140463683754206,
      "rewards/margins": 0.0005608043866232038,
      "rewards/rejected": -0.007701267953962088,
      "step": 420
    },
    {
      "epoch": 0.12556629686440982,
      "grad_norm": 7.03125,
      "learning_rate": 3.374836535314135e-07,
      "log_odds_chosen": 0.0553693063557148,
      "log_odds_ratio": -0.6874627470970154,
      "logits/chosen": 1.1983997821807861,
      "logits/rejected": 1.1994131803512573,
      "logps/chosen": -0.1475822627544403,
      "logps/rejected": -0.15458670258522034,
      "loss": 0.182,
      "nll_loss": 0.1475822627544403,
      "rewards/accuracies": 0.565625011920929,
      "rewards/chosen": -0.00737911369651556,
      "rewards/margins": 0.00035022044903598726,
      "rewards/rejected": -0.0077293342910707,
      "step": 440
    },
    {
      "epoch": 0.13127385581279208,
      "grad_norm": 7.59375,
      "learning_rate": 3.3004918099222473e-07,
      "log_odds_chosen": 0.09319862723350525,
      "log_odds_ratio": -0.6688340902328491,
      "logits/chosen": 1.1878206729888916,
      "logits/rejected": 1.1925140619277954,
      "logps/chosen": -0.14523671567440033,
      "logps/rejected": -0.15989655256271362,
      "loss": 0.1787,
      "nll_loss": 0.14523671567440033,
      "rewards/accuracies": 0.5843750238418579,
      "rewards/chosen": -0.007261836435645819,
      "rewards/margins": 0.0007329917862080038,
      "rewards/rejected": -0.007994827814400196,
      "step": 460
    },
    {
      "epoch": 0.13698141476117434,
      "grad_norm": 6.5,
      "learning_rate": 3.2308533487561923e-07,
      "log_odds_chosen": 0.041309647262096405,
      "log_odds_ratio": -0.6908696293830872,
      "logits/chosen": 1.1932710409164429,
      "logits/rejected": 1.1951032876968384,
      "logps/chosen": -0.14346230030059814,
      "logps/rejected": -0.1491231620311737,
      "loss": 0.178,
      "nll_loss": 0.14346230030059814,
      "rewards/accuracies": 0.528124988079071,
      "rewards/chosen": -0.00717311492189765,
      "rewards/margins": 0.0002830428129527718,
      "rewards/rejected": -0.0074561582878232,
      "step": 480
    },
    {
      "epoch": 0.1426889737095566,
      "grad_norm": 6.6875,
      "learning_rate": 3.1654446891645916e-07,
      "log_odds_chosen": 0.08073492348194122,
      "log_odds_ratio": -0.6769333481788635,
      "logits/chosen": 1.204514741897583,
      "logits/rejected": 1.2099024057388306,
      "logps/chosen": -0.14232878386974335,
      "logps/rejected": -0.15676438808441162,
      "loss": 0.1762,
      "nll_loss": 0.14232878386974335,
      "rewards/accuracies": 0.543749988079071,
      "rewards/chosen": -0.007116439286619425,
      "rewards/margins": 0.0007217799429781735,
      "rewards/rejected": -0.007838220335543156,
      "step": 500
    },
    {
      "epoch": 0.14839653265793887,
      "grad_norm": 6.40625,
      "learning_rate": 3.103854276221004e-07,
      "log_odds_chosen": 0.10610455274581909,
      "log_odds_ratio": -0.6604468822479248,
      "logits/chosen": 1.1745558977127075,
      "logits/rejected": 1.182629942893982,
      "logps/chosen": -0.1395079791545868,
      "logps/rejected": -0.15413886308670044,
      "loss": 0.1725,
      "nll_loss": 0.1395079791545868,
      "rewards/accuracies": 0.596875011920929,
      "rewards/chosen": -0.006975398398935795,
      "rewards/margins": 0.0007315444527193904,
      "rewards/rejected": -0.007706942968070507,
      "step": 520
    },
    {
      "epoch": 0.15410409160632113,
      "grad_norm": 7.53125,
      "learning_rate": 3.0457245193658633e-07,
      "log_odds_chosen": 0.0635085254907608,
      "log_odds_ratio": -0.6808339953422546,
      "logits/chosen": 1.1945526599884033,
      "logits/rejected": 1.2023719549179077,
      "logps/chosen": -0.14418356120586395,
      "logps/rejected": -0.15572601556777954,
      "loss": 0.1782,
      "nll_loss": 0.14418357610702515,
      "rewards/accuracies": 0.5562499761581421,
      "rewards/chosen": -0.007209178060293198,
      "rewards/margins": 0.0005771225551143289,
      "rewards/rejected": -0.007786300964653492,
      "step": 540
    },
    {
      "epoch": 0.1598116505547034,
      "grad_norm": 6.4375,
      "learning_rate": 2.9907430227728956e-07,
      "log_odds_chosen": 0.10175992548465729,
      "log_odds_ratio": -0.6643875241279602,
      "logits/chosen": 1.1899529695510864,
      "logits/rejected": 1.196588158607483,
      "logps/chosen": -0.12970320880413055,
      "logps/rejected": -0.14286372065544128,
      "loss": 0.1629,
      "nll_loss": 0.12970320880413055,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": -0.006485159508883953,
      "rewards/margins": 0.0006580256158486009,
      "rewards/rejected": -0.007143185939639807,
      "step": 560
    },
    {
      "epoch": 0.16551920950308566,
      "grad_norm": 6.59375,
      "learning_rate": 2.938635496921166e-07,
      "log_odds_chosen": 0.1222359910607338,
      "log_odds_ratio": -0.6543433666229248,
      "logits/chosen": 1.165566086769104,
      "logits/rejected": 1.176872968673706,
      "logps/chosen": -0.1328657865524292,
      "logps/rejected": -0.15196380019187927,
      "loss": 0.1656,
      "nll_loss": 0.1328657865524292,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.0066432892344892025,
      "rewards/margins": 0.0009549002861604095,
      "rewards/rejected": -0.0075981891714036465,
      "step": 580
    },
    {
      "epoch": 0.17122676845146792,
      "grad_norm": 6.40625,
      "learning_rate": 2.8891599832849384e-07,
      "log_odds_chosen": 0.07435475289821625,
      "log_odds_ratio": -0.6786702275276184,
      "logits/chosen": 1.1653727293014526,
      "logits/rejected": 1.1729456186294556,
      "logps/chosen": -0.13466762006282806,
      "logps/rejected": -0.14524760842323303,
      "loss": 0.1686,
      "nll_loss": 0.13466762006282806,
      "rewards/accuracies": 0.543749988079071,
      "rewards/chosen": -0.006733381655067205,
      "rewards/margins": 0.0005289987893775105,
      "rewards/rejected": -0.007262379862368107,
      "step": 600
    },
    {
      "epoch": 0.17693432739985018,
      "grad_norm": 6.46875,
      "learning_rate": 2.842102114016023e-07,
      "log_odds_chosen": 0.06968313455581665,
      "log_odds_ratio": -0.6842463612556458,
      "logits/chosen": 1.1924859285354614,
      "logits/rejected": 1.1970815658569336,
      "logps/chosen": -0.1375359296798706,
      "logps/rejected": -0.14428077638149261,
      "loss": 0.1717,
      "nll_loss": 0.1375359296798706,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.006876795552670956,
      "rewards/margins": 0.0003372425271663815,
      "rewards/rejected": -0.0072140381671488285,
      "step": 620
    },
    {
      "epoch": 0.18264188634823245,
      "grad_norm": 6.3125,
      "learning_rate": 2.7972711943222965e-07,
      "log_odds_chosen": 0.050189681351184845,
      "log_odds_ratio": -0.6886711120605469,
      "logits/chosen": 1.18123459815979,
      "logits/rejected": 1.1853077411651611,
      "logps/chosen": -0.13307775557041168,
      "logps/rejected": -0.14148366451263428,
      "loss": 0.1675,
      "nll_loss": 0.13307775557041168,
      "rewards/accuracies": 0.543749988079071,
      "rewards/chosen": -0.006653887685388327,
      "rewards/margins": 0.0004202956333756447,
      "rewards/rejected": -0.007074182387441397,
      "step": 640
    },
    {
      "epoch": 0.1883494452966147,
      "grad_norm": 6.75,
      "learning_rate": 2.754496943956198e-07,
      "log_odds_chosen": 0.10923095792531967,
      "log_odds_ratio": -0.6653646230697632,
      "logits/chosen": 1.1724183559417725,
      "logits/rejected": 1.1764357089996338,
      "logps/chosen": -0.13218267261981964,
      "logps/rejected": -0.14843502640724182,
      "loss": 0.1655,
      "nll_loss": 0.13218264281749725,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": -0.00660913297906518,
      "rewards/margins": 0.0008126182365231216,
      "rewards/rejected": -0.007421751506626606,
      "step": 660
    },
    {
      "epoch": 0.19405700424499697,
      "grad_norm": 7.5625,
      "learning_rate": 2.7136267706462855e-07,
      "log_odds_chosen": 0.11492220312356949,
      "log_odds_ratio": -0.6578065752983093,
      "logits/chosen": 1.1791291236877441,
      "logits/rejected": 1.1879496574401855,
      "logps/chosen": -0.12462460994720459,
      "logps/rejected": -0.14196214079856873,
      "loss": 0.1575,
      "nll_loss": 0.1246245950460434,
      "rewards/accuracies": 0.609375,
      "rewards/chosen": -0.00623122975230217,
      "rewards/margins": 0.0008668772061355412,
      "rewards/rejected": -0.007098107133060694,
      "step": 680
    },
    {
      "epoch": 0.19976456319337924,
      "grad_norm": 6.28125,
      "learning_rate": 2.674523475799702e-07,
      "log_odds_chosen": 0.08160345256328583,
      "log_odds_ratio": -0.6759176850318909,
      "logits/chosen": 1.1840394735336304,
      "logits/rejected": 1.1869642734527588,
      "logps/chosen": -0.1288476288318634,
      "logps/rejected": -0.1388583481311798,
      "loss": 0.1626,
      "nll_loss": 0.1288476288318634,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.006442382000386715,
      "rewards/margins": 0.000500535941682756,
      "rewards/rejected": -0.006942917592823505,
      "step": 700
    },
    {
      "epoch": 0.2054721221417615,
      "grad_norm": 6.5625,
      "learning_rate": 2.637063313749449e-07,
      "log_odds_chosen": 0.09184299409389496,
      "log_odds_ratio": -0.6702265739440918,
      "logits/chosen": 1.1718956232070923,
      "logits/rejected": 1.175795316696167,
      "logps/chosen": -0.12900219857692719,
      "logps/rejected": -0.14239074289798737,
      "loss": 0.1625,
      "nll_loss": 0.12900219857692719,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.0064501105807721615,
      "rewards/margins": 0.0006694265175610781,
      "rewards/rejected": -0.007119536399841309,
      "step": 720
    },
    {
      "epoch": 0.21117968109014376,
      "grad_norm": 6.375,
      "learning_rate": 2.6011343419135436e-07,
      "log_odds_chosen": 0.09168614447116852,
      "log_odds_ratio": -0.6726126670837402,
      "logits/chosen": 1.1679967641830444,
      "logits/rejected": 1.1757028102874756,
      "logps/chosen": -0.12767893075942993,
      "logps/rejected": -0.1408797949552536,
      "loss": 0.1613,
      "nll_loss": 0.12767891585826874,
      "rewards/accuracies": 0.5531250238418579,
      "rewards/chosen": -0.006383946631103754,
      "rewards/margins": 0.000660042860545218,
      "rewards/rejected": -0.0070439898408949375,
      "step": 740
    },
    {
      "epoch": 0.21688724003852602,
      "grad_norm": 7.21875,
      "learning_rate": 2.566635011696727e-07,
      "log_odds_chosen": 0.06486343592405319,
      "log_odds_ratio": -0.6854541301727295,
      "logits/chosen": 1.1893186569213867,
      "logits/rejected": 1.1990292072296143,
      "logps/chosen": -0.13231287896633148,
      "logps/rejected": -0.14083275198936462,
      "loss": 0.1666,
      "nll_loss": 0.13231287896633148,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.006615643855184317,
      "rewards/margins": 0.0004259937850292772,
      "rewards/rejected": -0.007041637785732746,
      "step": 760
    },
    {
      "epoch": 0.2225947989869083,
      "grad_norm": 6.5,
      "learning_rate": 2.5334729596906997e-07,
      "log_odds_chosen": 0.10854773223400116,
      "log_odds_ratio": -0.6595977544784546,
      "logits/chosen": 1.133418083190918,
      "logits/rejected": 1.1365370750427246,
      "logps/chosen": -0.1260526478290558,
      "logps/rejected": -0.13903895020484924,
      "loss": 0.159,
      "nll_loss": 0.1260526478290558,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.006302633322775364,
      "rewards/margins": 0.000649315130431205,
      "rewards/rejected": -0.006951947696506977,
      "step": 780
    },
    {
      "epoch": 0.22830235793529055,
      "grad_norm": 6.8125,
      "learning_rate": 2.5015639663712994e-07,
      "log_odds_chosen": 0.11962361633777618,
      "log_odds_ratio": -0.6557859182357788,
      "logits/chosen": 1.193050742149353,
      "logits/rejected": 1.1949045658111572,
      "logps/chosen": -0.12155069410800934,
      "logps/rejected": -0.1389068216085434,
      "loss": 0.1543,
      "nll_loss": 0.12155069410800934,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": -0.006077534519135952,
      "rewards/margins": 0.0008678067242726684,
      "rewards/rejected": -0.0069453418254852295,
      "step": 800
    },
    {
      "epoch": 0.2340099168836728,
      "grad_norm": 6.96875,
      "learning_rate": 2.4708310555370036e-07,
      "log_odds_chosen": 0.11173542588949203,
      "log_odds_ratio": -0.6594327092170715,
      "logits/chosen": 1.1577401161193848,
      "logits/rejected": 1.168442964553833,
      "logps/chosen": -0.12523651123046875,
      "logps/rejected": -0.13914619386196136,
      "loss": 0.1582,
      "nll_loss": 0.12523649632930756,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": -0.0062618255615234375,
      "rewards/margins": 0.0006954848067834973,
      "rewards/rejected": -0.006957309786230326,
      "step": 820
    },
    {
      "epoch": 0.23971747583205508,
      "grad_norm": 6.375,
      "learning_rate": 2.44120371254637e-07,
      "log_odds_chosen": 0.08974882960319519,
      "log_odds_ratio": -0.6731756329536438,
      "logits/chosen": 1.1761608123779297,
      "logits/rejected": 1.1835229396820068,
      "logps/chosen": -0.12705373764038086,
      "logps/rejected": -0.13870368897914886,
      "loss": 0.1607,
      "nll_loss": 0.12705373764038086,
      "rewards/accuracies": 0.578125,
      "rewards/chosen": -0.0063526867888867855,
      "rewards/margins": 0.0005824980325996876,
      "rewards/rejected": -0.00693518528714776,
      "step": 840
    },
    {
      "epoch": 0.24542503478043734,
      "grad_norm": 6.46875,
      "learning_rate": 2.412617203266904e-07,
      "log_odds_chosen": 0.07684284448623657,
      "log_odds_ratio": -0.6797510385513306,
      "logits/chosen": 1.1856390237808228,
      "logits/rejected": 1.1910351514816284,
      "logps/chosen": -0.13384665548801422,
      "logps/rejected": -0.14304396510124207,
      "loss": 0.1678,
      "nll_loss": 0.13384665548801422,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.006692333612591028,
      "rewards/margins": 0.0004598658415488899,
      "rewards/rejected": -0.007152198348194361,
      "step": 860
    },
    {
      "epoch": 0.25113259372881963,
      "grad_norm": 6.59375,
      "learning_rate": 2.385011978752745e-07,
      "log_odds_chosen": 0.09443817287683487,
      "log_odds_ratio": -0.6688668727874756,
      "logits/chosen": 1.1798049211502075,
      "logits/rejected": 1.1851075887680054,
      "logps/chosen": -0.13026410341262817,
      "logps/rejected": -0.14304585754871368,
      "loss": 0.1637,
      "nll_loss": 0.13026410341262817,
      "rewards/accuracies": 0.5843750238418579,
      "rewards/chosen": -0.006513205356895924,
      "rewards/margins": 0.0006390881026163697,
      "rewards/rejected": -0.007152292877435684,
      "step": 880
    },
    {
      "epoch": 0.2568401526772019,
      "grad_norm": 6.34375,
      "learning_rate": 2.3583331531828906e-07,
      "log_odds_chosen": 0.0780445784330368,
      "log_odds_ratio": -0.6762874722480774,
      "logits/chosen": 1.1711924076080322,
      "logits/rejected": 1.174644112586975,
      "logps/chosen": -0.1311054527759552,
      "logps/rejected": -0.13908042013645172,
      "loss": 0.1649,
      "nll_loss": 0.1311054676771164,
      "rewards/accuracies": 0.5843750238418579,
      "rewards/chosen": -0.0065552727319300175,
      "rewards/margins": 0.0003987480013165623,
      "rewards/rejected": -0.006954020820558071,
      "step": 900
    },
    {
      "epoch": 0.26254771162558416,
      "grad_norm": 6.28125,
      "learning_rate": 2.3325300446383042e-07,
      "log_odds_chosen": 0.09321997314691544,
      "log_odds_ratio": -0.6692286729812622,
      "logits/chosen": 1.162477731704712,
      "logits/rejected": 1.167890191078186,
      "logps/chosen": -0.1252012997865677,
      "logps/rejected": -0.13648080825805664,
      "loss": 0.1587,
      "nll_loss": 0.1252012997865677,
      "rewards/accuracies": 0.5843750238418579,
      "rewards/chosen": -0.006260064896196127,
      "rewards/margins": 0.0005639748414978385,
      "rewards/rejected": -0.006824040319770575,
      "step": 920
    },
    {
      "epoch": 0.2682552705739664,
      "grad_norm": 6.125,
      "learning_rate": 2.307555769970229e-07,
      "log_odds_chosen": 0.07238370180130005,
      "log_odds_ratio": -0.6811813116073608,
      "logits/chosen": 1.14841628074646,
      "logits/rejected": 1.1509320735931396,
      "logps/chosen": -0.13035395741462708,
      "logps/rejected": -0.13957537710666656,
      "loss": 0.1644,
      "nll_loss": 0.13035395741462708,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.006517698056995869,
      "rewards/margins": 0.0004610706528183073,
      "rewards/rejected": -0.006978767924010754,
      "step": 940
    },
    {
      "epoch": 0.2739628295223487,
      "grad_norm": 7.0,
      "learning_rate": 2.283366886387715e-07,
      "log_odds_chosen": 0.12806378304958344,
      "log_odds_ratio": -0.6543666124343872,
      "logits/chosen": 1.1922695636749268,
      "logits/rejected": 1.2018707990646362,
      "logps/chosen": -0.12585455179214478,
      "logps/rejected": -0.1432829350233078,
      "loss": 0.1586,
      "nll_loss": 0.12585455179214478,
      "rewards/accuracies": 0.590624988079071,
      "rewards/chosen": -0.006292728241533041,
      "rewards/margins": 0.0008714188006706536,
      "rewards/rejected": -0.00716414675116539,
      "step": 960
    },
    {
      "epoch": 0.27967038847073095,
      "grad_norm": 7.0,
      "learning_rate": 2.2599230735278433e-07,
      "log_odds_chosen": 0.04509379714727402,
      "log_odds_ratio": -0.69524747133255,
      "logits/chosen": 1.1726185083389282,
      "logits/rejected": 1.1782230138778687,
      "logps/chosen": -0.13184015452861786,
      "logps/rejected": -0.13727745413780212,
      "loss": 0.1666,
      "nll_loss": 0.13184015452861786,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": -0.006592008285224438,
      "rewards/margins": 0.00027186458464711905,
      "rewards/rejected": -0.006863872520625591,
      "step": 980
    },
    {
      "epoch": 0.2853779474191132,
      "grad_norm": 6.21875,
      "learning_rate": 2.2371868507134142e-07,
      "log_odds_chosen": 0.08361569792032242,
      "log_odds_ratio": -0.6737332344055176,
      "logits/chosen": 1.1739256381988525,
      "logits/rejected": 1.1796576976776123,
      "logps/chosen": -0.12685084342956543,
      "logps/rejected": -0.1371733546257019,
      "loss": 0.1605,
      "nll_loss": 0.12685084342956543,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.006342542357742786,
      "rewards/margins": 0.000516124942805618,
      "rewards/rejected": -0.006858667824417353,
      "step": 1000
    },
    {
      "epoch": 0.2910855063674955,
      "grad_norm": 6.875,
      "learning_rate": 2.2151233248863027e-07,
      "log_odds_chosen": 0.08207818865776062,
      "log_odds_ratio": -0.6765726804733276,
      "logits/chosen": 1.1616705656051636,
      "logits/rejected": 1.1721489429473877,
      "logps/chosen": -0.1313740760087967,
      "logps/rejected": -0.14133639633655548,
      "loss": 0.1652,
      "nll_loss": 0.1313740760087967,
      "rewards/accuracies": 0.5718749761581421,
      "rewards/chosen": -0.006568703800439835,
      "rewards/margins": 0.0004981160163879395,
      "rewards/rejected": -0.007066819816827774,
      "step": 1020
    },
    {
      "epoch": 0.29679306531587774,
      "grad_norm": 6.3125,
      "learning_rate": 2.1936999653592838e-07,
      "log_odds_chosen": 0.1043349951505661,
      "log_odds_ratio": -0.6668421626091003,
      "logits/chosen": 1.1827070713043213,
      "logits/rejected": 1.1864862442016602,
      "logps/chosen": -0.13119131326675415,
      "logps/rejected": -0.14434470236301422,
      "loss": 0.1645,
      "nll_loss": 0.13119131326675415,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.006559564732015133,
      "rewards/margins": 0.0006576695595867932,
      "rewards/rejected": -0.007217234931886196,
      "step": 1040
    },
    {
      "epoch": 0.30250062426426,
      "grad_norm": 6.84375,
      "learning_rate": 2.1728864020781585e-07,
      "log_odds_chosen": 0.09999938309192657,
      "log_odds_ratio": -0.6645081043243408,
      "logits/chosen": 1.1529250144958496,
      "logits/rejected": 1.1585088968276978,
      "logps/chosen": -0.12486504018306732,
      "logps/rejected": -0.137950599193573,
      "loss": 0.1581,
      "nll_loss": 0.12486505508422852,
      "rewards/accuracies": 0.6031249761581421,
      "rewards/chosen": -0.006243253126740456,
      "rewards/margins": 0.0006542770424857736,
      "rewards/rejected": -0.006897530052810907,
      "step": 1060
    },
    {
      "epoch": 0.30820818321264226,
      "grad_norm": 6.5,
      "learning_rate": 2.1526542445480877e-07,
      "log_odds_chosen": 0.0686492770910263,
      "log_odds_ratio": -0.6807483434677124,
      "logits/chosen": 1.171618103981018,
      "logits/rejected": 1.176278829574585,
      "logps/chosen": -0.12466160953044891,
      "logps/rejected": -0.1331419199705124,
      "loss": 0.1587,
      "nll_loss": 0.12466160953044891,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.006233080755919218,
      "rewards/margins": 0.0004240164998918772,
      "rewards/rejected": -0.006657096557319164,
      "step": 1080
    },
    {
      "epoch": 0.3139157421610245,
      "grad_norm": 6.75,
      "learning_rate": 2.1329769189683425e-07,
      "log_odds_chosen": 0.12860627472400665,
      "log_odds_ratio": -0.6553727388381958,
      "logits/chosen": 1.1408389806747437,
      "logits/rejected": 1.1476658582687378,
      "logps/chosen": -0.12643122673034668,
      "logps/rejected": -0.1450921595096588,
      "loss": 0.1592,
      "nll_loss": 0.12643122673034668,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.006321561522781849,
      "rewards/margins": 0.0009330459870398045,
      "rewards/rejected": -0.007254607975482941,
      "step": 1100
    },
    {
      "epoch": 0.3196233011094068,
      "grad_norm": 7.34375,
      "learning_rate": 2.113829521450379e-07,
      "log_odds_chosen": 0.07565739750862122,
      "log_odds_ratio": -0.6793134212493896,
      "logits/chosen": 1.1778619289398193,
      "logits/rejected": 1.1843442916870117,
      "logps/chosen": -0.13324205577373505,
      "logps/rejected": -0.14512678980827332,
      "loss": 0.1672,
      "nll_loss": 0.13324205577373505,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": -0.006662103347480297,
      "rewards/margins": 0.0005942363059148192,
      "rewards/rejected": -0.007256339304149151,
      "step": 1120
    },
    {
      "epoch": 0.32533086005778905,
      "grad_norm": 6.40625,
      "learning_rate": 2.0951886854752806e-07,
      "log_odds_chosen": 0.08736299723386765,
      "log_odds_ratio": -0.6730456948280334,
      "logits/chosen": 1.1625810861587524,
      "logits/rejected": 1.1686502695083618,
      "logps/chosen": -0.12537319958209991,
      "logps/rejected": -0.13687022030353546,
      "loss": 0.159,
      "nll_loss": 0.12537319958209991,
      "rewards/accuracies": 0.578125,
      "rewards/chosen": -0.006268660072237253,
      "rewards/margins": 0.0005748508265241981,
      "rewards/rejected": -0.0068435100838541985,
      "step": 1140
    },
    {
      "epoch": 0.3310384190061713,
      "grad_norm": 6.53125,
      "learning_rate": 2.0770324619863194e-07,
      "log_odds_chosen": 0.073883555829525,
      "log_odds_ratio": -0.6786224246025085,
      "logits/chosen": 1.1667616367340088,
      "logits/rejected": 1.1703726053237915,
      "logps/chosen": -0.13239695131778717,
      "logps/rejected": -0.1426318883895874,
      "loss": 0.1663,
      "nll_loss": 0.13239695131778717,
      "rewards/accuracies": 0.565625011920929,
      "rewards/chosen": -0.006619847379624844,
      "rewards/margins": 0.0005117467371746898,
      "rewards/rejected": -0.007131594233214855,
      "step": 1160
    },
    {
      "epoch": 0.3367459779545536,
      "grad_norm": 6.375,
      "learning_rate": 2.059340210717385e-07,
      "log_odds_chosen": 0.1088346615433693,
      "log_odds_ratio": -0.6648656129837036,
      "logits/chosen": 1.1583200693130493,
      "logits/rejected": 1.165569543838501,
      "logps/chosen": -0.1252637803554535,
      "logps/rejected": -0.13931581377983093,
      "loss": 0.1585,
      "nll_loss": 0.1252637803554535,
      "rewards/accuracies": 0.559374988079071,
      "rewards/chosen": -0.006263189949095249,
      "rewards/margins": 0.0007026006933301687,
      "rewards/rejected": -0.0069657908752560616,
      "step": 1180
    },
    {
      "epoch": 0.34245353690293584,
      "grad_norm": 6.625,
      "learning_rate": 2.042092501533827e-07,
      "log_odds_chosen": 0.10282032191753387,
      "log_odds_ratio": -0.6641172766685486,
      "logits/chosen": 1.1580729484558105,
      "logits/rejected": 1.161696195602417,
      "logps/chosen": -0.12909960746765137,
      "logps/rejected": -0.1406385600566864,
      "loss": 0.1623,
      "nll_loss": 0.12909962236881256,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.006454980466514826,
      "rewards/margins": 0.0005769463605247438,
      "rewards/rejected": -0.007031927350908518,
      "step": 1200
    },
    {
      "epoch": 0.3481610958513181,
      "grad_norm": 6.59375,
      "learning_rate": 2.0252710247134554e-07,
      "log_odds_chosen": 0.07733563333749771,
      "log_odds_ratio": -0.6802065968513489,
      "logits/chosen": 1.1576850414276123,
      "logits/rejected": 1.162522554397583,
      "logps/chosen": -0.12488509714603424,
      "logps/rejected": -0.13480448722839355,
      "loss": 0.1589,
      "nll_loss": 0.12488509714603424,
      "rewards/accuracies": 0.596875011920929,
      "rewards/chosen": -0.006244254298508167,
      "rewards/margins": 0.0004959694342687726,
      "rewards/rejected": -0.006740224547684193,
      "step": 1220
    },
    {
      "epoch": 0.35386865479970037,
      "grad_norm": 6.53125,
      "learning_rate": 2.0088585092257693e-07,
      "log_odds_chosen": 0.1242731437087059,
      "log_odds_ratio": -0.655490517616272,
      "logits/chosen": 1.1623140573501587,
      "logits/rejected": 1.1639635562896729,
      "logps/chosen": -0.12796516716480255,
      "logps/rejected": -0.14206835627555847,
      "loss": 0.1607,
      "nll_loss": 0.12796516716480255,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.00639825826510787,
      "rewards/margins": 0.0007051602005958557,
      "rewards/rejected": -0.0071034180000424385,
      "step": 1240
    },
    {
      "epoch": 0.35957621374808263,
      "grad_norm": 6.9375,
      "learning_rate": 1.9928386481802028e-07,
      "log_odds_chosen": 0.05651208013296127,
      "log_odds_ratio": -0.687849223613739,
      "logits/chosen": 1.1747676134109497,
      "logits/rejected": 1.180385947227478,
      "logps/chosen": -0.12957434356212616,
      "logps/rejected": -0.13640108704566956,
      "loss": 0.164,
      "nll_loss": 0.12957434356212616,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.0064787170849740505,
      "rewards/margins": 0.00034133673761971295,
      "rewards/rejected": -0.00682005425915122,
      "step": 1260
    },
    {
      "epoch": 0.3652837726964649,
      "grad_norm": 6.625,
      "learning_rate": 1.9771960307117736e-07,
      "log_odds_chosen": 0.09077809751033783,
      "log_odds_ratio": -0.6687765717506409,
      "logits/chosen": 1.1625758409500122,
      "logits/rejected": 1.1661226749420166,
      "logps/chosen": -0.12270165979862213,
      "logps/rejected": -0.13382795453071594,
      "loss": 0.1561,
      "nll_loss": 0.12270165979862213,
      "rewards/accuracies": 0.5687500238418579,
      "rewards/chosen": -0.006135082803666592,
      "rewards/margins": 0.0005563151207752526,
      "rewards/rejected": -0.006691398564726114,
      "step": 1280
    },
    {
      "epoch": 0.37099133164484716,
      "grad_norm": 6.25,
      "learning_rate": 1.961916079657376e-07,
      "log_odds_chosen": 0.07469572126865387,
      "log_odds_ratio": -0.6803563833236694,
      "logits/chosen": 1.1645721197128296,
      "logits/rejected": 1.1695630550384521,
      "logps/chosen": -0.12728631496429443,
      "logps/rejected": -0.13893961906433105,
      "loss": 0.1613,
      "nll_loss": 0.12728631496429443,
      "rewards/accuracies": 0.5843750238418579,
      "rewards/chosen": -0.006364315748214722,
      "rewards/margins": 0.0005826650885865092,
      "rewards/rejected": -0.006946981884539127,
      "step": 1300
    },
    {
      "epoch": 0.3766988905932294,
      "grad_norm": 6.5625,
      "learning_rate": 1.9469849944497735e-07,
      "log_odds_chosen": 0.1153842955827713,
      "log_odds_ratio": -0.6580480337142944,
      "logits/chosen": 1.1878401041030884,
      "logits/rejected": 1.195858359336853,
      "logps/chosen": -0.12814465165138245,
      "logps/rejected": -0.1446561962366104,
      "loss": 0.161,
      "nll_loss": 0.12814465165138245,
      "rewards/accuracies": 0.6031249761581421,
      "rewards/chosen": -0.006407232489436865,
      "rewards/margins": 0.0008255770662799478,
      "rewards/rejected": -0.007232809904962778,
      "step": 1320
    },
    {
      "epoch": 0.3824064495416117,
      "grad_norm": 6.46875,
      "learning_rate": 1.9323896987207783e-07,
      "log_odds_chosen": 0.0802692323923111,
      "log_odds_ratio": -0.6752118468284607,
      "logits/chosen": 1.1755759716033936,
      "logits/rejected": 1.1816208362579346,
      "logps/chosen": -0.12898872792720795,
      "logps/rejected": -0.13849465548992157,
      "loss": 0.1627,
      "nll_loss": 0.12898872792720795,
      "rewards/accuracies": 0.5718749761581421,
      "rewards/chosen": -0.006449435837566853,
      "rewards/margins": 0.00047529704170301557,
      "rewards/rejected": -0.006924733519554138,
      "step": 1340
    },
    {
      "epoch": 0.38811400848999394,
      "grad_norm": 6.625,
      "learning_rate": 1.9181177921614533e-07,
      "log_odds_chosen": 0.119959257543087,
      "log_odds_ratio": -0.6586750149726868,
      "logits/chosen": 1.1766316890716553,
      "logits/rejected": 1.1811823844909668,
      "logps/chosen": -0.12140345573425293,
      "logps/rejected": -0.13672715425491333,
      "loss": 0.1543,
      "nll_loss": 0.12140345573425293,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": -0.006070171948522329,
      "rewards/margins": 0.0007661855779588223,
      "rewards/rejected": -0.006836357526481152,
      "step": 1360
    },
    {
      "epoch": 0.3938215674383762,
      "grad_norm": 5.75,
      "learning_rate": 1.9041575062365312e-07,
      "log_odds_chosen": 0.08580292761325836,
      "log_odds_ratio": -0.6745995283126831,
      "logits/chosen": 1.1618746519088745,
      "logits/rejected": 1.1701242923736572,
      "logps/chosen": -0.13371995091438293,
      "logps/rejected": -0.14495250582695007,
      "loss": 0.1674,
      "nll_loss": 0.13371995091438293,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.006685997359454632,
      "rewards/margins": 0.0005616283742710948,
      "rewards/rejected": -0.007247625384479761,
      "step": 1380
    },
    {
      "epoch": 0.39952912638675847,
      "grad_norm": 6.8125,
      "learning_rate": 1.8904976633935942e-07,
      "log_odds_chosen": 0.04863113909959793,
      "log_odds_ratio": -0.6931271553039551,
      "logits/chosen": 1.1948425769805908,
      "logits/rejected": 1.1975750923156738,
      "logps/chosen": -0.1309596598148346,
      "logps/rejected": -0.13705727458000183,
      "loss": 0.1656,
      "nll_loss": 0.1309596598148346,
      "rewards/accuracies": 0.518750011920929,
      "rewards/chosen": -0.0065479823388159275,
      "rewards/margins": 0.0003048806102015078,
      "rewards/rejected": -0.006852862890809774,
      "step": 1400
    }
  ],
  "logging_steps": 20,
  "max_steps": 2000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}